2025 HIGH SCHOOL ABSTRACTS

Use your search or find function on this page to search abstracts by student name, project number or category, title, or keywords.

A1-AS Astronomy and Space Science
SafeSpace
Jasmine Palit

As space exploration and industrialization has accelerated in recent years, especially with the rise of commercial space companies launching large constellations of satellites, so has the subsequent trail of space debris. Space debris can be extremely damaging to spacecraft, and even deadly to exposed astronauts on spacewalks. This necessitates a method of cleaning space of debris. NASA tracks and targets larger debris, but there is no current protocol for small debris (1 cm to 10 cm), which still pose a significant and growing threat due to the high velocities at which they travel (15 to 50 km/s). Modeled estimates from NASA show that the cost of such debris can run into hundreds of millions of dollars, when compounded over the many years such debris stay in orbit. This project aims to develop a method of capturing small debris by determining the most effective materials (filtering ability to weight) to rapidly decelerate and contain such debris. Given that it is infeasible to replicate the actual velocities of debris in space, we will model the problem by reproducing the realistic momentum of such objects, using a much higher mass object to mimic the much lighter, but much higher velocity, debris. The experiment aims to evaluate various materials to determine which are most effective at deaccelerating and capturing small , high momentum objects, while minimizing weight and volume, both critical parameters for space-based devices. We built an experimental apparatus consisting of PVC tubes to attach various test materials used as barriers and a drop tube to guide the projectile to the target. Our experiment evaluated 4 materials – packing bubble wrap, packing box cardboard, cotton fabric cloth and USPS envelope fabric. The later two, both tear resistant fabrics that are flexible but strong, turned out to be the most resilient against impact from our test projectile. We also trained a ML model with transfer learning, using custom images, to help us automate and improve the accuracy of the damage assessment process. A finalized, more resilient design would require several layers (say 2 – 5) of such a material to ensure robustness, redundancy and resilience in case of multiple strikes. A successful design could be launched as a Cube-Sat to maneuver and capture space debris in LEO or GEO. Such a design can also be conceived as a lightweight protective skin on a spacecraft or a spacesuit. This project demonstrates a pathway towards developing cheaper, lower cost designs to create a safer and more sustainable space environment for astronauts and spacecraft alike, an essential step for the progress of the current age of accelerated space development.

G1-AS Astronomy and Space Science
Recipe for an Exoplanet: Using Transmission Spectroscopy to Identify Potentially Livable Exoplanets
Sonia Warrick

Transmission Spectroscopy is a form of data collection that senses dips in light when a planet crosses the path of its
sun. Thanks to the James Webb Space Telescope’s (JWST) new technology surrounding Transmission Spectroscopy, new discoveries of exoplanets and their compositions are being uncovered. This project explores the JWST’s Transmission Spectroscopy data with the goal of creating a potential transmission spectroscopy model of Earth and using it to help find exoplanets that could sustain or maintain human or other forms of life.

G2-AS Astronomy and Space Science
Investigate Solar Activity Through Directional Muon Detectors on a High-Altitude Balloon
Eleanor Chen
Ruby Tran

The experiment is designed to investigate solar activity by building two modified muon detectors based on the MIT CosmicWatch apparatus (Axani). This experiment will focus on the exploration of muons that were produced in the collision between the high energy protons (as well as Gamma rays) and the Earth’s atmosphere. The collison often yields muons with a relatively extended lifespan that covers an average distance of 15 kilometers before decaying or being absorbed. The muon detectors consist of three primary components: an Arduino Nano, a SiPM PCB, and a scintillator. We applied two muon detectors and connected them with a coincidence cable to make the experiment sensitive to the direction of the inbound muons. This arrangement ensures that only signals resulting from muons passing directly through the scintillator are considered. The experiment contains two sub experiments: a flight experiment and a ground experiment. The flight experiment was conducted on a NASA hot air balloon with the TechRise Challenge Team in August 2024 at Utah. The flight data contained a wide range of independent variables such as heading and altitude. In order to further explore the relationship between muon flux and heading, we built a ground experiment which stimulates the experimental process on the balloon. In the ground experiment, we used a rotating disk, a photogate, and a 3D printed platform. The ground experiment is conducted in Worcester, Massachusetts. Our experiment can strengthen people’s understanding of cosmic rays because cosmic rays are closely related to human health, the Earth atmosphere, and global communications. 

G3-AS Astronomy and Space Science
Skyglow: The Ignored Pollution
Angela Dhimal
Ashley Deleon

 

G6-AS Astronomy and Space Science
Iron Group Element Production in Type Ia Supernovae: A Machine Learning Approach to Nucleosynthesis Yields
Aatmaj Nagarkar
Eric Feng
Tanish Parida

Type Ia supernovae (SNe Ia) serve as crucial standard candles in determining cosmic distances, including the Hubble constant and the rate of universal expansion. However, variations in SNe Ia light curves, arising from different progenitor characteristics, introduce significant uncertainties. This study aims to refine the precision of SNe Ia as standard candles by developing a machine learning-based correction model that incorporates progenitor factors and nucleosynthesis yields. Using data on explosion mechanisms, progenitor radii, metallicity, and age, we first establish a model to predict Nickel-56 yields, a key factor driving the light curve of SNe Ia. A neural network is trained on theoretical datasets to capture the relationship between progenitor characteristics and Nickel-56 production. The second phase involves reverse mapping the trained model to constrain progenitor factors from observational Nickel-56 yields derived from Arnett’s Rule on observed light curves with a Physics -Informed Neural Network (PINN) to account for the underlying physics. In Phase 3, a Convolutional Neural Network (CNN) is utilized to analyze how progenitor factors influence the temporal evolution of SNe Ia light curves, and a correction model is created to improve empirical light curve fitting. The final phase involves testing the correction model with observational data, adjusting standard empirical models like SALT2 to improve accuracy. We compare the corrected light curves against existing fitting models and assess their impact on cosmological measurements, particularly the Hubble constant. This innovative approach to integrating nucleosynthesis yields with machine learning could significantly enhance the accuracy of SNe Ia as standard candles, reducing uncertainties in the measurement of universal expansion.

A5-BE Behavioral and Social Sciences
SingSense: Lowering Stress Levels by Singing in a Biometrically Supported VR Platform
Thomas Hong

Teenagers and young adults are experiencing increasing levels of stress, and many are either reluctant to seek help or inaccessible to mental healthcare. Virtual Reality (VR) and singing have both been demonstrated to facilitate stress relief. SingSense, a platform that combines VR, singing, and real-time biometric feedback, offers a novel, entertaining method of stress relief. The SingSense virtual platform contains a smart song recommendation system based on the user’s emotional state input and displays the song lyrics along with biometric data; the physical component consists of a fingertip-pulse-sensor. This study investigates the effect of singing in SingSense on one’s stress level, indicated by biomarkers such as Heart Rate Variability (HRV), which measures the differences between Interbeat Intervals (IBI); a common measure of HRV, the Root Mean Square of Successive Differences (RMSSD) formula, is integrated into the system. As a higher HRV indicates a lower stress level, it was hypothesized that singing would increase HRV and, thus, decrease stress levels. Subjects’ heart rate (BPM), HRV, and self-reported emotional state were recorded before, during, and three minutes after singing on SingSense. Results show that singing on SingSense significantly increases one’s HRV from 44.3Å}7.5ms to 63.2Å}4.9ms (n=21, p<0.05), does not cause significant changes to their BPM, and significantly increases user self-reported emotional state from 6.3Å}1.1 to 7.6Å}0.9 (/10). Future research is needed to increase the accuracy of biometrics calculations and further develop the song recommendation system, but current results demonstrate that SingSense can effectively alleviate users ‘ stress and can be used as an entertaining mechanism to improve mental well-being.

A6-BE Behavioral and Social Sciences
Modeling and Optimizing Firms’ Competition and Collaboration Using Game Theory and Reinforcement Learning
Christopher Yoo

This study explores strategic interactions among firms in modern markets from the perspectives of competition,
cooperation, and balanced (mixed) strategies, using three representative game-theoretic models: Cournot competition, the repeated Prisoner’s Dilemma, and the Stackelberg model. First, Cournot competition assumes simultaneous quantity decisions among firms to illustrate competitive dynamics. In this work, I incorporate reinforcement learning (RL) into the Cournot framework to simulate how firms iteratively learn and converge toward a Nash equilibrium. Next, the repeated Prisoner’s Dilemma captures cooperative settings by modeling repeated interactions where firms can choose between cooperation (C) and defection (D). I use classical game-theoretic analysis to examine conditions under which cooperation can be sustained or broken. Finally, the Stackelberg model describes balanced structures in which a leader’s decision precedes that of followers, mixing elements of both competition and partial collaboration. By comparatively analyzing these three models, the study highlights the diverse and dynamic nature of interfirm strategic behavior and demonstrates how reinforcement learning can effectively complement traditional game -theoretic approaches.

A7-BE Behavioral and Social Sciences
Understanding Social Determinants of Disabilities: A Machine Learning Approach
Anwita Wadekar

Disabilities present profound challenges to over 40 million Americans, including reduced quality of life, economic consequences, and systemic inequities. The leading cause of disabilities is chronic diseases such as cardiovascular disease, cancer, diabetes, and depression. Deficits in social determinants influence the onset and management of these chronic diseases, and living with disabilities further widens the gap in social determinants. This study investigates the associations between social determinants and disabilities by applying machine learning to the CDC PLACES data. Random forest regressor, chosen for its ability to handle non-linearity, correlated predictors, and interpretability, estimates the overall disability rate based on nine social determinants. The results show that social determinants can explain about 82% of the variation in disability rate. SHAP analysis quantifies the relative contribution of the nine social determinants to the disability rate in each county, highlighting the most influential social determinant in each county. Poverty is the most influential social determinant in about 70% of the counties, followed by lack of a high school diploma in around 25% of the counties. A user-friendly app to compare how social determinants contribute to disabilities across different counties within their state makes my research accessible and actionable to public health officials. My research thus highlights the interconnectedness between social determinants and disabilities, and empowers public health officials to find local and national opportunities for targeted interventions to reduce these disparities.

A8-BE Behavioral and Social Sciences
Using Optical Music Recognition and Vector Cosine Similarity to Create an Accessible Learning Tool for Vision -Impaired
Singers
Isaac Lee

For musicians with visual impairments, accessing and learning music is a constant challenge. The World Health Organization reports that more than 285 million people around the world have visual impairments. However, less than 15% of written music is available in Braille. Traditional Braille notation requires extensive memorization and costly transcription software, while other alternatives, such as audio-based tools, lack critical musical details such as dynamics and articulation. Current optical music recognition (OMR) systems limit independent music learning by only prioritizing transcription of flat, one-dimensional sheet music over accessibility. This project focuses on vision-impaired singers (though it is scalable to other instruments), presenting a novel solution that integrates adaptive audio playback, real-time score analysis, and Braille conversion as an accessible outlet for independent music study. The system employs vector cosine similarity to recognize patterns in melody, harmony, and phrasing, allowing visually impaired users to better study the artistic message within each piece. Additionally, the implementation of Natural Language Processing (NLP) in a voice-activated query system enables musicians to interact with the score by asking specific questions about the score, ranging from questioning stylistic choices to the pitches and rhythms of notes. The system integrates Braille translation libraries to accommodate a wide variety of users, guaranteeing compatibility with various musical notations and formats. By translating expressive elements, such as dynamics, diminuendos, and mood, into structured audio playback, this system enables vision-impaired musicians to interpret musical expression independently. With dynamic audio adjustments and an interactive framework, this project removes financial and technological barriers, expanding access to music education and performance for vision-impaired musicians worldwide.

A9-BE Behavioral and Social Sciences
Redefining Inflation Forecasting 2.0: A Novel Approach with Transformer Architecture and Pioneering an Economic Policy Formulation System to Effectively Combat Persistent High Inflation
David Guo

Significant inflation forecasting errors and policy missteps by central banks and central governments are among the
major root causes of persistent high inflation. It has created far-reaching economic challenges for society and significantly increased public health and social welfare costs. Particularly, it has severely and disproportionately deteriorated the quality of life for over 60% of Americans who live paycheck to paycheck and struggle daily to make ends meet . This underscores the urgency for a deeper understanding of inflation in theory and a more accurate inflation forecast in practice.

This Phase 2.0 economic science research achieves a significant advancement over Phase 1.0, making the following key contributions:

• Investigated and created a novel Transformer-based
approach, systematically and significantly improving inflation forecasting accuracy by over 50% beyond an already high
standard while reducing computational costs by a factor of 10.

• Investigated and achieved major improvements in accuracy, robustness and performance for long time horizon forecasting, including 12-month and 24-month forecasts.

• Fundamentally further improved both the theoretical and practical aspects of existing
econometric models related to inflation, including QTM and Phillips Curve.

• Designed and created a
time-aware Transformer architecture for the economic forecasting . The novel attention-based system precisely quantifies the interplay between economic drivers, identifies latent inflationary forces, and uncovers previously unrecognized features and relationships.

• Invented an automated reinforcement learning- based AI agent that optimizes economic policy pathways to achieve target inflation outcomes, empowering policymakers to formulate proactive and effective policies to tackle persistent high inflation.

Keywords: Inflation forecasting, Transformer, economic theory improvement, reinforcement learning, optimal economic policy path, automated policy formulation.

L10-BE Behavioral and Social Sciences
Investigating ECGC and Quercetin Synergistic Effect in Amyloid Beta Modified Drosophila melanogaster
Andrea Lo

Alzheimer’s disease (AD) affects over 50 million patients globally, with projections estimating it to double every 5,
reaching 152 million patients by 2050. Although there has been a great extent of research targeting various aspects of
Alzheimer’s, only three drugs have been approved to treat symptoms of the disease. Throughout the past decade, research studies suggest that flavonoids, particularly quercetin and epigallocatechin 3- gallate may offer potential therapeutic benefits due to their ability to modulate inflammatory and immunological functions. Both have demonstrated promise in reducing amyloid beta plaques, neuroinflammation, and oxidative stress in vitro and in vivo. However, clinical applications have been limited as they have been observed to lack bioavailability and overall therapeutic impact. Interestingly, research has suggested that combination of quercetin and EGCG have a synergetic, producing stronger combination with either compound alone, particularly in conditions like diabetes, protease cancer, and insulin resistance. This study strives to explore whether this synergistic interaction could also enhance the therapeutic potential of flavonoids in the treatment of Alzheimer’s disease through investigation of the behavioral and neuroprotective outcomes by the climbing and lifespan assay.

L11-BE Behavioral and Social Sciences
How Do Influencers vs. Close Relationships Impact Decision Making in Adolescent Girls?
Lane Carroll

 

L12-BE Behavioral and Social Sciences
A Comparison of Mental Health Outcomes Among New and Seasoned Emergency Medical Service Workers
Akul Agarwal

Emergency Medical Services (EMS) professionals operate in high-stress environments, frequently encountering traumatic events, life-threatening situations, and physically demanding workloads. These stressors contribute to significant mental health challenges, including Burnout, Secondary Traumatic Stress (STS), and Resilience. While previous research has examined various factors influencing mental health in EMS, the relationship between years of experience and these psychological outcomes remains underexplored. This study aimed to investigate whether years of EMS experience significantly predict levels of Burnout, STS, and Resilience among providers. Using a cross-sectional survey design, we collected data from EMS professionals using three validated scales: the Copenhagen Burnout Inventory (CBI), the Secondary Traumatic Stress Scale (STSS), and the Brief Resilience Scale (BRS). A total of 30 items assessed respondents’ mental health, with additional demographic questions, including years of EMS experience. Data were analyzed using linear regression models to assess correlations between experience and mental health outcomes. Group comparisons were conducted by splitting respondents into those with less than five years and more than five years of experience. Results indicated that years of EMS experience were not significantly correlated with Burnout, STS, or Resilience, suggesting that other factors—such as workload, organizational support, and individual coping strategies—may play a more critical role. These findings challenge the idea that mental health outcomes improve or worsen with time in the profession and highlight the need for targeted mental health interventions that focus on workplace conditions rather than experience alone. Given these results, EMS agencies should consider early-career resilience training, structured peer support programs, and proactive mental health initiatives for all providers, regardless of tenure. Future research should explore organizational culture, job satisfaction, and coping mechanisms as potential mediators in the relationship between EMS experience and mental health outcomes.

L7-BE Behavioral and Social Sciences
Striking the Right Chord: How Musical Training Amplifies Emotional Reactivity, using Artificial Intelligence (AI)
Gavin Warnakulasooriya

This study investigates the impact of musical training on emotional reactivity through AI-driven sentiment analysis and webcam-based facial expression recognition. Using a Convolutional Neural Network (CNN), participants’ facial expressions were analyzed in real time while they viewed emotionally evocative videos, both with and without audio. Musically trained individuals, defined as those with over 400 hours of cumulative musical training, exhibited significantly higher emotional reactivity than their non-musically trained counterparts. Furthermore, they demonstrated a greater difference in emotional response to auditory -enhanced content compared to non-auditory content, suggesting increased sensitivity to sound-based emotional cues. Statistical analyses, including Mann-Whitney U tests and Cohen’s d effect size calculations, confirmed that these differences were both statistically significant and had a large effect size, reinforcing the hypothesis that musical training enhances emotional processing. These findings contribute to growing research on the cognitive and affective benefits of musical training , providing quantitative evidence that musicians exhibit heightened emotional awareness and empathy. Additionally, this study highlights the potential of AI-powered facial sentiment analysis as an objective tool for psychological research, emotion recognition, and human-computer interaction. Future research could further explore the neural mechanisms underlying these effects and the implications for music education, media analytics, and therapeutic interventions.

Keywords: Musical training, empathy, emotional reactivity, AI, webcam sentiment analysis, facial expression recognition, deep learning, affective computing

L8-BE Behavioral and Social Sciences
The Spread of False Information
Olivia Amaral

In this project I will be attempting to obtain a deeper understanding of how false information can spread and what type of impact it can have on the human brain with repeated exposure. For my experiment I will run 3 trials using 3 groups of 6 people per trial. I will have a control group, a single exposure group, and a repeated exposure group for each trial of this experiment. There will be two stories that participants will be asked to read, a neutral story (the one that contains true information) and a misinformation story (the one that contains false information). These stories are based on fictional characters and events so that there is no risk of actual false information that could cause potential harm. The control group will only be asked to read the neutral story, the single exposure group will be asked to read the neutral story as well as the misinformation story, and the repeated exposure group will be asked to read the neutral story as well as the misinformation story twice. After each group is asked to read the amount of stories that they are supposed to read the number of times that they are supposed to read them, all of the groups will be given the same questionnaire that is based on the stories they had been asked to read. The questionnaire will contain answers from both the neutral story and the misinformation story. Each participant’s ability to differentiate between the correct and incorrect answers will be recorded . The purpose of this experiment and the data collected is to understand how false information can be spread and how the repeated exposure of false information may affect memory and the human brain.

L9-BE Behavioral and Social Sciences
Predicting Poverty in the US Based on Demographic and Socioeconomic Features
Qiaorui Zhang

Poverty is a pervasive issue worldwide, including in the US, where millions face financial hardship and limited access to essential resources. Poverty prediction is crucial for developing interventions that alleviate hardship and reduce cyclical inequality. I used supervised machine learning models to predict poverty in the US, offering a more efficient and accessible approach than traditional methods. Four binary-classification models were trained on publicly-available data from the 2022 American Community Survey Public Use Microdata Sample. The Extreme Gradient Boosting (XGBoost) algorithm performed best, yielding an accuracy of 88.87% and a F1-score of 88.43%. Results identify employment, education, mobility status, relationship to householder, and age as the most indicative factors of poverty. This study demonstrates that classification algorithms are valuable tools for identifying and understanding poverty across America.

M10-BE Behavioral and Social Sciences
Examination of Recovery in Drosophila from Sleep Deprivation Induced by Environmental Cues Changing Circadian Rhythms
Devina Paul

Sleep is essential for maintaining physical and cognitive function, immunity, and development. For most organisms, this is regulated by circadian rhythms and homeostatic processes. Suggested sleep recommendations for humans exist, but many people are unable to follow them due to environmental pressures. This has resulted in sleep deprivation across populations. In this experiment, drosophila hydei were used as model organisms to observe stress and recovery from sleep deprivation. The flies were placed in vials under different lighting conditions. A control group was placed away from intense light, and two other groups were placed under constant light for three days to disrupt their circadian rhythms. Afterward, wall-following, an anxiety-defense behavior involving the flies positioning themselves on or near the wall of an enclosed space was observed and filmed; the flies were put in a petri dish placed over concentric circles to determine how far they were from the walls. Then, the flies were put back into the vials under different conditions; these remained the same for the control group; one group of flies was placed in the same area as the control, and another group of flies was placed under light for limited periods. Wall-following behavior was filmed again. The number of flies in specific areas was counted after reviewing the videos. The average percentages of flies in ten-second intervals were calculated and organized in bar charts to make conclusions on behaviors. While wall-following behavior appeared to increase after changing the conditions, the flies placed under light for limited periods showed a smaller increase
compared to the other flies.

Keywords: sleep, sleep deprivation, drosophila, anxiety, stress, flies, model organisms

M11-BE Behavioral and Social Sciences
Understanding Neurotoxicity: Drosophila Melanogaster as a Model for Memory Loss
Vaagmi Shukla

As the use of CBD and nicotine becomes increasingly prevalent, understanding their effects on memory and cognitive function is critical. This study examines how these substances impact associative memory in Drosophila melanogaster using an olfactory conditioning paradigm. Associative memory, the ability to link a stimulus with an outcome, is fundamental to learning and cognition. To assess memory retention, fruit flies were divided into three groups: a control group, a CBD-exposed group, and a nicotine-exposed group. Each group underwent olfactory conditioning, in which they were exposed to a specific scent (sweet orange or banana) paired with a food source. Following conditioning, the flies were tested in a T-tube test apparatus, a setup designed from repurposed water bottles. One side of the tube contained the conditioned scent, while the other remained odorless. Memory retention was measured by the percentage of flies that moved toward the conditioned scent as well as the time they took to pick a side. 

Statistically significant results indicated that the control group demonstrated the highest memory retention (83.38%), showing strong associative memory. The nicotine-exposed group displayed moderate retention (70.14%), while the CBD-exposed group exhibited significant memory impairment (39.9%), suggesting that CBD disrupts associative learning. 

Beyond the findings, the experiment was conducted using handmade, repurposed materials, reinforcing the idea that meaningful research can be conducted sustainably. Results contribute to the growing body of knowledge on how commonly used substances influence learning and memory, emphasizing the need for further investigation into their long- term effects on cognitive health.

M12-BE Behavioral and Social Sciences
The Effect of Different Energy Drinks on Daphnia’s Heart Rate
Joselyn Majano
Ruth Rodriguez

 

M1-BE Behavioral and Social Sciences
Socioeconomic Bias in Sentencing: Phase 2 — Expanding to Subconscious Associations through Implicit Bias Testing
Grace Bui
Nina Lee
Yurie Lee

In our experiment, we explored how subconscious biases related to socioeconomic status (SES) affect sentencing
decisions. We hypothesized that people from lower SES backgrounds would receive harsher sentences , and that
participants would take less time to assign sentences to low-SES individuals. To test this, we created two surveys using
Inquisit Lab (measures survey responses by the millisecond). Survey A included six crime scenarios, with alternating SES profiles: high-SES offenders were paired with white-collar crimes, and low-SES offenders with blue-collar crimes, reflecting common assumptions about crime and capability. In Survey A, participants assigned longer sentences to the high-SES offenders, assuming they were more capable of committing white-collar crimes. In Survey B, we reversed the SES and crime pairings: high-SES offenders were now paired with blue-collar crimes, and low-SES offenders with white-collar crimes. In this survey, participants assumed that low-SES offenders were less capable of committing white-collar crimes, leading to them receiving lighter sentences. Our findings reveal that implicit biases are context-dependent: high-SES individuals were assumed to be more capable of committing blue-collar crimes, resulting in harsher sentences, while low-SES individuals were perceived as less capable of committing white-collar crimes, leading to lighter sentences. This emphasizes how biases based on SES influence decision-making and underscores the need for standardized practices to reduce bias and ensure fairer sentencing.

M2-BE Behavioral and Social Sciences
The Impact Music Genres Play on the Behavioral Responses of Drosophila melanogaster
Grace Qiu

Autism Spectrum Disorder (ASD) affects the behaviors, communications, and social abilities of millions of people
globally. Recently, music therapy has become a non-invasive treatment for individuals with ASD; however, music therapy often uses classical music as the genre of music during therapy sessions, and little is known about the effects of different music genres on ASD-associated behaviors. This study uses Drosophila to investigate the impacts of music genres on both wild-type and ASD-induced Drosophila. The Drosophila homolog of the human FMR1 protein, dFMR1, is a protein that binds RNA. By dropping the dFMR1 protein, it creates Fragile X syndrome (FXS), a common mental retardation. Then, using two standardized Drosophila behavioral assays – locomotion and geotaxis – wild- type and ASD-modeled Drosophila are exposed to different music genres, such as pop and classical. The geotaxis assays showed that ASD-induced Drosophila exhibited greater improvements in response movement, social interactions, and behaviors when exposed to pop music than those exposed to classical music. Yet the locomotion assay showed that Drosophila exposed to classical music improved significantly. Ultimately, this research hopes to advance the knowledge of the relationship between music and ASD-like behaviors and address the need for effective interventions for individuals with ASD.

M3-BE Behavioral and Social Sciences
Unraveling the Role of GABAergic Dysfunction in Catatonia: A GABAergic Investigation in Drosophila
Varsha Alladi

Catatonia is a neuropsychiatric syndrome characterized by various motor, behavioral, and cognitive differences. Despite it having a treatment, the neurobiological mechanisms underlying catatonia remain poorly understood, and existing models fail to capture the full spectrum of symptoms. Recent studies link the development of catatonia to GABAergic dysfunction, more especially, disturbances in GABA-A receptor signaling. This work seeks to fill important knowledge gaps by using Drosophila melanogaster as a model organism to examine the part that GABAergic disruption plays in catatonia-like symptoms. Using bicuculline, a GABA-A receptor antagonist, the research explores how altered GABAergic neurotransmission affects behaviors fundamental to the syndrome, such as locomotion, sensory responsiveness, and social interaction. The study makes use of Drosophila melanogaster’s short life cycle, behavioral observability, and genetic tractability. To make sure of constant exposure, experimental groups will be given bicuculline through food supplementation, while wild-type flies will act as the control group. The impact of GABAergic dysfunction will be assessed using a set of behavioral tests. Using tracking software, a locomotion test will assess changes in motor function by measuring parameters like total distance traveled and movement frequency. A chemosensory experiment will evaluate sensory responsiveness and determine how the flies respond to chemical stimuli, such as repellents. Lastly, by watching whether flies exhibit increased isolation or clustering tendencies in observation spaces, a clustering assay will assess social behaviors. To compare the experimental and control groups and determine if bicuculline -induced GABA-A receptor antagonism causes catatonia-like symptoms in Drosophila, the data from these assays will be statistically examined. This study will add to our understanding of the GABAergic contributions to catatonia by finding certain behavioral irregularities overall laying the groundwork for future research on the neurochemical basis of the condition.

M4-BE Behavioral and Social Sciences
The Correlation Between Sleep and Cognitive Ability in Teenagers
Samuel Ogbu

This experiment seeks to answer the question of “How sleep influences teenager’s reaction time and memory.” My hypothesis to this question was “if teenagers between the ages of 14 through 16 decrease the amount of sleep they get per night, then their memory and reaction time will decrease. Additionally, if teenagers between the ages of 14 through 16 instead increase the amount of sleep they get per night, then their memory and reaction time will improve.”. The experiment was motivated by discourse surrounding school start time, and more importantly the impact sleep deprivation has on high schoolers. This experiment divides teenagers into two groups, the experimental group can go to sleep and wake up whenever they please but the control group will follow a sleep schedule. Both groups will take online assessments which evaluate reaction time, long term memory, and short term memory. The results of the experiment present a strong correlation between reaction time and sleep, but inconclusive / coincidental results on the topic of memory. Results on short term memory were contradictory to general scientific knowledge and results on long term memory were so marginally significant that its results cannot be considered anything but coincidental. 

M5-BE Behavioral and Social Sciences
Examining Differences in Neural Oscillations via Electroencephalography to Compare Reactions to Vocal and Siren Alarms
Emerson Ohashi
Lana Nguyen

Following a house fire that caused the deaths of the 6 Philpott children, a police investigation concluded that the children had not been restrained and were not woken by the fire alarm in their house. Due to this realization, it became imperative for more effective alarm systems to be put into place so that further tragedies could be prevented. The purpose of our experiment is to determine if a vocal alarm will produce a greater amount of neural activity than a siren. Our hypothesis is that if an alarm uses vocal cues instead of sirens, children would be more aware of the alarm and be more likely to react to it. In this experiment, we had participants ranging from ages 13-15 wear an electroencephalogram (EEG) on their head and draw an image to the best of their ability. Using the EEG, we measured the neural oscillations created when the participants reacted to vocal and siren alarms played during testing. From the data we collected, we found that there was a 36.2% increase in the amplitude and a 11.1% increase in the frequency of the average person’s reaction to a vocal stimulus when compared to a siren . We also observed that subsequent siren alarms produced smaller reactions to the stimuli than the first. Based on the data and observations, we concluded that vocal stimuli create larger reaccompared to sirens, leading to the conclusion that vocal alarms should replace sirens to ensure children’s safety .

M6-BE Behavioral and Social Sciences
Co-use of Tobacco and Cannabis: A Machine Learning Approach to Study the Influence of Social Media on a Public Health Crisis
Vineeth Godavarti

Key Words: Cannabis, Tobacco, Co-use, Social media, Surveillance, Neural network, Machine learning

Co-use of tobacco products and cannabis is becoming increasingly prevalent among young adults aged 18 to 24 years. Co-use has a deleterious effect on both physical and mental health while increasing the risk of addiction to both substances, making quitting harder. Additionally, more states are legalizing recreational and medicinal cannabis use, making it more accessible to young people. Therefore, it is important to take action against this public health concern before it worsens. This research was conducted to analyze the content of social media posts about co-use, from three social media platforms, and investigate their relationship with sentiment (i.e. whether pro-co-use, neutral, or anti co-use). Sentiment analysis results from a manual coding study were compared to those from Mention.com, a social media monitoring platform. Manual coding results showed that over 90% of the relevant posts mentioned ‘smoking’, ‘tobacco’, and ‘cannabis’ generally, suggesting the widespread discussion of these products. Approximately 28% of the posts had a negative sentiment towards co-use while over 50% of the posts were neutral. Additionally, approximately 60% of the posts described experiencing a physical or mental health effect as a result of co-use. However, despite the negative sentiments expressed and the potentially harmful physical and mental health effects, the study revealed that over 90% of the posts had no mention of quitting. These manually coded data then served as training data for a machine learning study. A recurrent neural network (RNN) model was developed that predicted the sentiment of posts with 78% accuracy, which was higher than the 61% accuracy obtained using Mention.com’s model. The RNN model was further optimized by adding more training data to improve the accuracy to 91%. This model can be employed by public health officials and researchers to conduct surveillance of social media influence on young people and identify effective ways to communicate about co-use on social media.

M7-BE Behavioral and Social Sciences
Patient-Centered Radiology Reports Regarding Glioblastomas with Generative Artificial Intelligence
Shreya Jha

 

M8-BE Behavioral and Social Sciences
Reframing Housing Advocacy: The Impact of Informative and Cautious-Based Advocacy On Affordable Housing Support.
Vinay Murugan

The United States is currently in an affordable housing crisis , and more homes need to be built. Research has shown
that multifamily housing can address this, but the movement called NIMBYism (Not in My Backyard) has prevented new home construction through strict zoning. NIMBYists have various concerns about affordable housing, which are shown to be false. I believe, however, that there is an inherent negative perception of what affordable housing is—a low -quality building (also a misconception), leading to opposition. Indeed, the online questionnaire results showed that most people incorrectly associated affordable housing with rundown buildings and were less likely to support new construction if their perceptions of the quality of affordable homes were low. From there, I analyzed ways to change this negative perception of quality to reduce opposition. Through interviews with a housing non-profit, I created new ways these organizations could advocate for housing to change the negative perceptions. One of them is to use visual aids in the form of an informative advocacy approach. In addition, certain words are more stigmatized, so I believe they should be avoided in the form of a cautious advocacy approach. There’ll also be a combined advocacy approach using both the informative and cautious strategies. The t-test results indicate that the new advocacy approaches improve support, but the ANOVA test results indicate that there is no difference between the three of them. This study gives a new method local housing advocates can use to reduce NIMBYism and create more affordable and accessible neighborhoods .

M9-BE Behavioral and Social Sciences
How Noise Affects Eye to Hand Coordination?
Elif Sayar
Lathika Kannan

The main focus of this project was eye-hand coordination. Eye-hand coordination is the ability to use what is seen to guide and control hand movements. This particular skill is extremely crucial and contributes to many activities such as sports, music, and doing homework. Noise is a constant factor in everyday life as well, so that’s how these two components combined into the question of how volume would affect eye-to-hand coordination. Knowing what range of decibels in an environment positively impacts eye-to-hand coordination, if they impact it at all, can help in the future when these two elements need to be tested together. The impact may range on the student but this experiment will calculate the averages to see what may be best or true for most. The experimental procedure consists of two tests that each subject takes, one being a physical test (cornhole) meanwhile the other being a test on the computer (HECOOR). The 25 subjects take each of these tests in an environment of 50db, 70db, and 90db. After recording every participant’s results in the tests, averages are formed out of the data and the answer comes to light that 50db is the best environment for optimal eye-to-hand coordination.

N1-BE Behavioral and Social Sciences
Novel Approach to Uncover the Effects of Social Isolation on Mouse Models Using AI Machine Learning
Daniel Choi

Behavioral patterns in rodents provide critical insights into neurobiological processes. This study investigates whether
significant behavioral differences emerge between socially housed and socially deprived C 57BL/6 mice by leveraging artificial intelligence (AI) and machine learning (ML) techniques for behavioral analysis. Using Keypoint-MoSeq (KPMS), an unsupervised and unbiased AI-driven approach, we identified movement patterns with high reproducibility, minimizing observer bias inherent in traditional manual analysis. Sixteen male mice (8 isolated, 8 grouped) were observed in open-field exploration, where pose-tracking and syllable detection revealed significant differences in movement sequences , including increased rearing down behavior and altered syllable transitions in isolated mice. These findings highlight the advantage of ML-based analysis in uncovering subtle behavioral alterations that may be overlooked by conventional methods. By providing a scalable and reproducible framework for high-throughput behavioral phenotyping, AI and ML have the potential to revolutionize experimental paradigms and accelerate discoveries in neuroscience, pharmacology, and genetics.

N2-BE Behavioral and Social Sciences
Gendered Dimensions: Analyzing Gender Disparities in Spatial Reasoning Variances and Cognitive Processing
Reet Saini

Gender disparities within the realm of spatial reasoning cognitive processing have been a subject of ongoing research for
many decades. While males and females are both equally proficient in active neural functionality, the exact areas in which their skills may be applicable remains a mystery, hence the numerous attempts by scientists to figure out the gender variances in such processes. As such, the experiment aims to examine the biological and environmental factors that may contribute to such variances. This research was conducted by asking sixteen participants between the ages of fifteen and thirty-five to take part in one of three mental tests, evaluating the speed at which they performed as well as their overall accuracy. The three tests included a 2D/3D visualization test focused on asking participants to solve a tangram puzzle with as much time as they needed, a multitasking test focused on asking participants to count various shapes appearing on a blank screen, and an audio localization test focused on asking participants to be blindfolded in the middle of a room and point out where sounds are coming from with their finger. It was hypothesized that the male participants would be more proficient at spatial manipulation while the female participants would demonstrate superior performance in multitasking capabilities. This hypothesis was proven correct as the males demonstrated superior results in the visualization and localization tasks while the females did better in the multitasking task. This variance could be attributed to differences in parahippocampal gyrus activity and memory retrieval.

Q3-BE Behavioral and Social Sciences
Is it harmful or helpful? Examining the causes and consequences of generative AI usage among high school students
Michelle Liu

The findings of this study highlight the growing role of AI tools like ChatGPT in high school education , with most students reporting positive experiences, particularly as a study aid rather than a tool for completing assignments. Many found it helpful for generating practice problems, summarizing complex information, and clarifying difficult concepts when teachers’ explanations were insufficient. However, concerns about dependency and procrastination emerged, as some students admitted to relying on ChatGPT more heavily in certain subjects when there were major workloads . While it can alleviate academic stress by making studying more efficient, some students worry that overuse may weaken their problem-solving skills or increase pressure to absorb excessive amounts of information.

School policies and teacher perspectives on ChatGPT remain divided, with some schools restricting its use while others integrate it as a learning resource. Some teachers encourage students to use AI for brainstorming and practice , while others fear it could replace critical thinking. These findings suggest that AI can be a valuable educational tool when used responsibly , but a balance must be maintained to prevent over-reliance. As AI continues to evolve, schools and educators must establish guidelines that promote ethical and effective use, ensuring that students benefit from its capabilities while developing independent learning skills.

A11-BI Biochemistry
Examining the Effects of Nanoparticles and Reversal Compounds on Drosophila Melanogaster
Andrew Mozinski

Investigating the Effects of Nanoparticles and Reversal-Compounds on Drosophila Melanogaster Andrew Mozinski
Crimson Global Academy Nanoparticles have been reported to cause a detrimental effect on fertility and physical fitness in humans. This is a concern due to an increasing exposure of nanoparticles in our daily environment. Previous studies have also shown that antioxidant compounds increased longevity and fertility in model organisms. In this study, we used a Drosophila melanogaster model to examine the combinatorial effects of nanoparticles and antioxidants . First, using a climbing motility assay and a larvae count fertility assay, we looked at three commonly found nanoparticles: gold, silver and titanium, and established a working concentration where they exerted a negative influence on the physical fitness and fertility of Drosophila adult flies. We then determined the beneficial effect of antioxidants melatonin , N-acetylcysteine (NAC) and quercetin. Subsequently, we examined the combined effects of nanoparticles and antioxidants on Drosophilaphysical fitness and fertility. Furthermore, we examined the transgenerational effect of these chemical combinations on the first generation of progeny flies. We hypothesize that beneficial compounds such as NAC, quercetin and melatonin can reverse the detrimental effects from metal nanoparticles on physical and fertility capabilities on Drosophila melanogaster . Our results showed that antioxidants were unable to completely reverse the detrimental effects caused by the nanoparticles . While most combinations still resulted in a lower physical fitness, the combination of melatonin and gold nanoparticles resulted in some recovery and the flies cultivated in this condition showed physical fitness closest to the control group. In general, nanoparticles also reduced fertility based on larvae count, but melatonin was able to rescue the negative effect on fertility exerted by silver nanoparticles.

A12-BI Biochemistry
Effect of Halogens and Stereochemistry of Hydantoin Derivatives on Antiviral Activities Against SARS -CoV-2 Variants: Atomistic Insight Through Computational Studies
Anushri Pal

A recent report discovered a chloro-hydantoin lead for COVID-19 antiviral therapy, however, effects
of stereochemical changes of the spiro-cycle and halogens were not examined [1]. Herein, minimal structural variations,
such as stereochemistry and atomic size of halogen substituents, had their effect on binding affinity and biological activity examined. Blind docking derived a correlation between binding affinity and IC 50 of various hydantoin analogs. The correlation predicted the fluoro hydantoin derivative (2s,4s) to be more potent than chloro (2s,4s) and bromo derivative (2s,4s). However, experimentally, chloro and bromo analogs exhibited greater in vitro potency against Mpro than the fluoro derivative (8). To rationalize the impact of atomic size and stereo configuration on observed biological activity, the ligand’s volume occupancy and protein-ligand interactions were evaluated. Compounds 9a and 10a occupy 33% and 35% of the protein pocket space, respectively. The bromine atom’s size optimally fits the cavity, conferring higher stability. Molecular Dynamics simulation data of the ligand torsion profile and the solvent accessible surface area also indicated higher conformational stability for the bromo compound compared to the other halogenated derivatives. Chloro and bromo (2r,4r) stereoisomers, had more conformational flexibility and were accordingly less active Mpro inhibitors in vitro. Thus, it is conclusive that larger halogen size increases SARS-CoV-2 inhibitor stability and activity with (2s,4s) being the optimal stereoisomer [2].

Keyword: COVID-19

[1] Luttens et al., Journal of the American
Chemical Society 2022, 144 (7), p.2905

[2] Pal et al., IEEE Transactions on Computational Biology and
Bioinformatics, vol. 22, p. 240, Jan. 2025 https://ieeexplore.ieee.org/document/10819964

O11-BI Biochemistry
The Role of Endocrine-Disrupting Pesticides in Decline of Bee Populations
Shea Ryan

Pollinator populations, particularly honeybees (Apis mellifera), are declining at alarming rates, raising concerns about
food security and ecosystem stability. This study investigates the role of endocrine-disrupting pesticides in pollinator decline by examining their effects on bee reproductive success and behavior . Field sites were selected based on pesticide exposure levels, and behavioral observations, including foraging efficiency and communication. Biological samples such as pollen, nectar, and deceased bees were collected and analyzed for pesticide residues and hormonal disruptions. Statistical comparisons between exposed and control groups revealed significant differences in hormone levels , brood development, and hive activity. This supports the hypothesis that endocrine-disrupting pesticides negatively impact bee populations. Thesem findings suggest a need for stricter pesticide regulations and alternative pest management strategies to protect pollinators and maintain ecological balance.

O12-BI Biochemistry
Effects of Allura Red on Algae
Asha Patel
Grey Courchesne
Tess Armstrong

Recently the use of Red 40 (Allura Red) has been linked to colon cancer, hyperactivity and damage to DNA strands. It has not been banned by the US or EU yet, but the use of Red 40 is majorly controversial. This study collected data from algae (Chlorella) submerged in different concentrations of Red 40 to measure the impact of Allura Red on the effectiveness of photosynthesis within an algae plant. Photosynthesis is a light dependent reaction which releases oxygen and occurs in plants while respiration, which consumes oxygen, doesn’t require light. Chlorella algae beads were put into test tubes with varying concentrations of Allura Red under a light lamp . A respirometer was then used to measure the gas produced from each test tube. The control group had zero traces of Allura Red. Red 40 had a statistically significant effect on the algae (P=.00001). Using data from graphs (R2= 0.912) and our own qualitative observations showed that there was a statistically measurable effect on the Chlorella algae beads from Allura Red. In another experimental control, test tubes for each concentration of Allura Red were covered in Aluminum foil to limit light access to the algae and test light as a factor. In this study it became evident that Red 40 has significant effects on the effectiveness and efficiency of photosynthesis. While plants are not closely related to humans biologically. One can only wonder what are the serious consequences that Red40 has on our own ability to maintain homeostasis. 

P10-BI Biochemistry
Advanced Study of Cell-Free DNA Methylation and Circulating Serum Protein Markers as Non-Invasive Predictors of Lung Cancer Risk in Non-Smoking Asian Women with Pulmonary Nodules
Claire Tong

It has been demonstrated that over half of Asian American females diagnosed with lung cancer are non -smokers. Despite various factors, such as secondhand smoke exposure, unique cooking practices, or genetic backgrounds, contributing to the development of lung cancer among non-smoking Asian women, the underlying mechanisms remain largely unknown. Additionally, there is a lack of efficient early-diagnosis methods for assessing lung cancer risk in this demographic. In this study, blood samples were collected from non-smoking Asian women patients with 5-10 mm pulmonary nodules detected by Low-Dose Computed Tomography (LDCT). Histological analysis was conducted on these nodules to distinguish between malignant and benign cases. Subsequently, a 16-Plex Tandem Mass Tag (TMT)-Based Mass Spectrometry was used to discern differentially expressed serum proteins, revealing 63 up-regulated and 48 down-regulated proteins in malignant cases. Among them, 14 up-regulated proteins were identified as candidate protein biomarkers. Through a comprehensive literature review, it was discovered that 12 proteins were established lung cancer biomarkers, and 2 proteins were not identified in any other studies, suggesting that they may be potentially novel non-invasive predictors. To further enhance the study’s diagnostic potential, cell-free DNA (cfDNA) methylation analysis using Next-Generation Sequencing (NGS) was integrated. cfDNA was isolated from the collected blood samples, and NGS was performed to identify methylation patterns associated with lung cancer. The methylation profiles of cfDNA were compared between malignant and benign cases revealing specific differentially methylated regions (DMRs) that could serve as additional biomarkers for lung cancer detection. The combined approach of serum protein biomarkers and cfDNA methylation analysis aims to provide a more comprehensive and sensitive method for early detection of lung cancer in non- smoking Asian women. The following study would extend its scope by evaluating the lung cancer predicting performance of these biomarkers and methylation patterns in a larger pool of at-risk Asian female patients. The ultimate objective of this study is to develop a panel of non-invasive serum protein and cfDNA methylation biomarkers that can serve as auxiliary tools for the early -stage diagnosis of lung cancer risk in this population.

P11-BI Biochemistry
Cyanide Toxicity of Apple Seeds vs. Cherry Pits
Katherine Smith

Cherry pits and apple seeds in very small amounts and their concentrations are compared. Cyanide testing and Kirby
Bauer will be done. Calculations are used to determine how many seeds or pits the average person would need to consume to inflict harm. The diluted solution of cyanide will also be used on E. Coli to determine if hydrogen cyanide has any impact on bacterial growth.

P12-BI Biochemistry
Waste Not, Fuel More: Extracting Ethanol from Disregarded Leaves
Olivia Wan
Sophia Wan

Every year, over 100 billion trees in the US lose their leaves during the fall season, resulting in about 20 trillion leaves left as yard waste. While cleanup efforts have been made by local officials , there is still an abundance of leaves unaccounted for on the streets, sidewalks, parks, etc. Due to this, billions of leaves are left to wither, providing no benefit to the ecosystem, yet there is so much potential for them in the world of biofuels. Although we consider leaves as dead once they have fallen off the trees, they still store an abundance of potential energy in their chloroplasts and can be utilized to provide a source of ethanol. Our project delves into the process of extracting the cellulose of disregarded leaves that hold the potential energy we can use through the processes of enzymatic hydrolysis and microbial fermentation. We conducted various experiments on the most efficient ways to produce this ethanol using four major steps . We first begin by pretreating our leaves using different methods to break down the cell wall, so that we can increase the contact the added enzymes will have with the cellulose during hydrolysis. By doing so, we can gain a higher number of polysaccharides that can undergo microbial fermentation and transform into ethanol. As a result, we have a higher yield of ethanol, which is a substance used in many everyday products, such as fuels and disinfectants. We demonstrate the efficiency of ethanol by calculating the heat energy produced by burning ethanol to boil water, as well as how effective it is at killing bacteria. By showcasing the successful results using discarded leaves, our research can advance further on a larger scale, and potentially use the trillions of wasted leaves for our society’s benefit.

P1-BI Biochemistry
In a State of Crisis, Which Type of Water Is Optimal When Filtered Water Is Not Available?
Arianna Hopkins
Bradie Barrett
Melissa Rodriguez

 

P2-BI Biochemistry
The Synergistic Effect of Combining Natural Compounds 1,8-cineole (Eucalyptol) and Naringenin with 11 Antibiotics of
Different Drug Classes
Ishana Saroha

The use of synergistic combinations between natural compounds and commercial antibiotics is a promising solution for
antibiotic resistance, as increasingly prevalent bacterial infections have become more difficult to treat with existing
antibiotics. This study evaluates two natural compounds, 1,8-Cineole (Eucalyptol) and Naringenin, as potential antibiotic adjuvants to enhance efficacy of eleven commercial antibiotics from different drug classes against Escherichia Coli. Individually, Eucalyptol demonstrates a mechanism of action similar to efflux pump inhibitors , as it displays antimicrobial properties through membrane disruption, and Naringenin damages bacterial membranes, thus performing similar to membrane permeabilizers. In vitro methods were applied across 22 compound-antibiotic combinations, including broth microdilution assays and a derivative of checkerboard assays. Minimum Inhibitory Concentration values, Area Under Curve values from growth curves, and Inhibitory Concentration 50% (IC50) values were obtained from these assays. IC50 values were compared to identify any potential synergistic interactions between the natural compounds and antibiotics. Results from kinetic studies helped identify multiple novel compound-drug combinations. Eucalyptol significantly enhanced the effectiveness of Azithromycin. Antagonistic interactions were observed when Eucalyptol was combined with Tetracyclineand Kanamycin. Similarly, an antagonistic interaction was found when Naringenin was combined with Trimethoprim . This research suggests that Eucalyptol and Azithromycin have a synergistic relationship , which offers a promising strategy to combat antibiotic resistance and provide new insights into potential therapeutic applications.

P3-BI Biochemistry
Computational Modeling of circCOPA-Antibody Complex for Novel Glioblastoma Detection and Therapeutic Applications
Kushal Patil

Glioblastoma Multiforme is an aggressive, late-stage form of brain cancer that originates in the brain or spinal cord.
circCOPA99aa is a novel protein linked to the early stages of Gliomas and can be considered a biomarker that would aid in detecting this aggressive brain cancer. An antibody is a protein that our immune system uses to defend against specific antigens(bacteria, viruses) inside us. It binds to them, allowing our T-cells and other immune system cells to attack the antigen. Recently, the role of the circCOPA protein in the regression of glioblastoma disease has been identified and can be used as a biomarker to detect glioblastoma. I hypothesize that the antibody interacts with the binding site of the circCOPA protein and thus can be used in a detection process. Therefore, I have performed computational modeling of the circCOPA-antibody system. The amino acid sequence I obtained from UniProt was uploaded on AlphaFold 3 to get the 3D structure and better understand its tertiary structure. Then, I used HDOCK to determine the potential binding sites for antibodies on the CircCOPA99aa protein. Based on the visual analysis and binding energy calculations, antibody 1A5F was identified as the most appropriate antibody. This research could be very influential to the scientific community as this aids in detecting gliomas in patients in highly early stages and acts as a way to make treatments more effective , as finding antibodies that would bind to this would help us control cell division in the tumor.

Keywords: CircCOPA, NONO-SFPQ, Antibody, Binding Energy

P4-BI Biochemistry
From Worms to Humans: Developing Acid-Resistant Casing for Novel Alzheimer’s Drug
David Priefer
Kais Guessab

Recently, the first animal studies of novel potential Alzheimer’s Disease maintenance drugs were performed. These extended chalcone compounds, inspired by curcumin, worked by inhibiting the formation of beta-amyloid plaques in the brain. We and others were the first to demonstrate that these compounds displayed physiological benefits on both C. elegans and fruit flies. However, it was observed that these compounds are acid-labile, making oral administration in humans a challenge, as a result of the low pH of the stomach. Since the ultimate goal of this type of drug is for daily maintenance use (i.e. the beta-amyloid plaques have either not started to grow or have been removed by one of the recently FDA-approved biologic drugs) use of more invasive delivery options, such as intravenous, is both inconvenient and have low compliance issues. Thus, it would be necessary to formulate a pill of the extended chalcone that is “protected” from the high acidity of the stomach and be liberated in the more basic or neutral pH environment of the small intestine.

In order to enable a drug to be ingested as a pill and bypass the acidity of the stomach, a protective layer must be created. One approach that is employed to create this coating is multi-layering. In this experiment, two naturally occurring polymers, hyaluronic acid (HA) and poly-l-lysine (PLL) were sequentially layered onto a clean microscope slide. Each layer was designed to be a single monolayer of the chosen polyelectrolyte. Due to these polymers’ degree of ionization being affected by solution acidity , altering the solution pH when creating the films inherently affected the thickness. Subsequently, after loading the compound at a variety of solution pHs, 25 unique combinations (delineated as [assembly pH, loading pH]) were created. Each of these were then investigated to determine the rate and degree of release when immersed in pH 2.5 (i.e. mimicking the stomach) and pH 7.4 (i.e. mimicking the small intestine).

It was found that two systems, specifically [5,7] and [9,7], were able to simultaneously protect the drug from release in low pH (stomach), while being able to quickly release in higher pH (small intestine). Thus, coating a pill of the extended chalcone with either of these combinations could allow, following ingestion, for an effective protection from the caustic environment of the stomach while still releasing within the small intestine. Once released, the compound could be absorbed into the bloodstream and travel to the brain, where it could potentially be utilized to stop the aggregation of the debilitated beta- amyloid plaques of Alzheimer’s Disease.

P5-BI Biochemistry
Designing a Novel Peptide Nucleic Acid Disrupting mTORC1-RHEB Interaction
Gia Bao Le

Ras homolog enriched in brain (RHEB) is the protein that plays a pivotal role in controlling the activity of the mTOR
pathway. By binding to the mammalian target of rapamycin complex 1 (mTORC1) allosteric site, RHEB indirectly controls the activity of protein synthesis, cell proliferation, and glucose metabolism in all species’ cellular bodies exhibiting mTOR pathway. However, a mutation in gene expression of inhibitory proteins against RHEB may lead to the elevated activity of RHEB, causing cells to grow uncontrollably and lead to the initiation of cell disorienting diseases such as cancers or premature aging. Herein, I have performed in silico virtual scanning tests to develop novel right-handed helical heterogeneous 1:1 α/Sulfono-γ-AA peptides as potential inhibitors toward the interaction between RHEB and mTORC1. Keywords: RHEB, mTORC1, protein synthesis, cancers, aging, right-handed helical heterogeneous 1:1 α/Sulfono-γ- AA peptides.

P6-BI Biochemistry
In Silico, In Vitro, and In Vivo Evaluation of Natural Compounds for Neuroprotection Against Oxidative Stress and Protein Aggregation
Mary Deng

One of the primary causes of neurodegeneration is the accumulation of toxic protein aggregates, including tau (phosphorylated by glycogen synthase kinase-3 beta, GSK-3β) and amyloid-β. Neurodegeneration is characterized by the increased presence of reactive oxygen species (ROS) that result in oxidative stress, inflammation, and mitochondrial dysfunction. While synthetic amyloid-β and GSK-3β inhibitors exist, many pose toxicity risks, and the potential of naturally derived compounds as safer alternatives has not been systematically evaluated. Using Pass prediction software, I prescreened potential neuroprotective compounds from PubChem, with compounds displaying a probability of activity (PA) ≥ 0.8 isolated for further analysis. Molecular docking simulations using Autodock Vina revealed strong binding affinities (≤ -7 kcal/mol) for Anthocyanins, Epigallocatechin Gallate (EGCG), Ginsenoside Rg1, and Quercetin. Next, I performed an in-vitro assay to evaluate water and alcohol-based extraction of these selected compound sources and found that green tea (containing EGCG) and Onion (containing Quercetin) significantly neutralized the ROS hydrogen peroxide. Additionally, both green tea and onion significantly increased embryo viability in an in-vivo cell viability assay. These results suggest the inhibitory effect of Green Tea and Onion against neurodegeneration . Modification of EGCG functional groups may increase their efficacy and binding affinity with GSK-3β, which could lead to novel treatments for neurodegenerative diseases.

P7-BI Biochemistry
Ozempic On My Mind: Examining the Effects of Semaglutide on Addictive Behavior in Caenorhabditis Elegans
Pradnya Cowlagi

Despite the widespread use of semaglutide, its neurological effects, particularly in addiction and mood regulation, remain underexplored. Semaglutide is a GLP-1 receptor agonist that interacts with brain regions involved in emotional regulation, but preliminary case studies suggest it may induce symptoms resembling major depressive episodes. To investigate these potential effects, we used C. elegans, a nematode model with conserved neuromechanical pathways relevant to addiction and mood regulation in humans. Our hypothesis was that semaglutide exposure would reduce the response of nicotine-addicted C. elegans to stimuli, as assessed by thrashing frequency (a measure of locomotion and mood) and brood size (a proxy for reproductive health and potential long-term effects on fertility).

We established four experimental groups: two control groups (no treatment and nicotine-only) and two experimental groups (semaglutide-only and semaglutide with nicotine). Following exposure, we measured body thrashing frequency as an indicator of mood and locomotion, and we assessed brood size to evaluate potential long-term effects on fertility. Statistical analysis using ANOVA confirmed that semaglutide significantly reduced thrashing frequency, suggesting a depressive-like state. The combination of semaglutide and nicotine further dampened locomotor activity, indicating a reduced response to addictive stimuli. Additionally, brood size was significantly reduced in semaglutide-treated worms, pointing to potential long- term effects on fertility and health.

These findings support our hypothesis that semaglutide may influence mood regulation and addiction pathways in C. elegans. While these results suggest potential implications for addiction treatment, they also raise important
questions about the side effects of semaglutide, particularly regarding mood disorders and fertility. Further research is needed to explore the long-term neurological effects of semaglutide, its mechanisms of action, and its potential as a therapeutic for addiction and mood disorders in humans.

P8-BI Biochemistry
The Impact of Hydroxyl Radicals on Drosophila Melanogaster using the Fenton Reaction to Mimic Oxidative Stress
Aarush Naik
Madhav Warrier

This project investigates the effects of hydroxyl radicals , highly reactive free radicals normally produced during fuel
combustion, on Drosophila Melanogaster (fruit flies). Using the Fenton reaction to generate hydroxyl radicals, we aim to mimic to oxidative stress caused by fuel emissions. The experiment will assess the impact of these radicals on fruit fly
health by monitoring mortality, behavioral changes, and cellular damage. The results will provide insight into how oxidative stress from fuel combustion can harm organisms, helping to understand its broader implications on living systems.

P9-BI Biochemistry
Effect of Citrus Juice on Apple Browning As a Model for Hyperpigmentation in Human Skin
Michaela Oppong

Once cut open and exposed to air, apples undergo a process where the enzyme polyphenol oxidase (PPO) is produced and brown patches start to appear on its insides. The rate at which browning occurs in apples when exposed to oxygen, can be compared to the melanin production in human skin. Because the sun contains UV radiation, it can react on human skin, causing an increase of an enzyme called tyrosinase to further cause an increase in melanin production in human skin. When too much melanin is produced in human skin, hyperpigmentation occurs which can cause the skin to darken. While there is no dangerous aspect of this condition, hyperpigmentation can cause emotional distress. When it comes to skin care products, many vitamin-c and citrus rich fruits are present in the ingredients to combat further melanin production on skin. In this experiment, apples will be used to model human skin, to test which citrus fruit juice is most effective in reducing browning in apples. It was hypothesized that if apples were treated with different fruit juices , then the fruit containing the most vitamin C would be more effective at delaying browning in apples. Apple slices were treated with juice extracted from lemon, lime, orange and grapefruit, excluding one untreated apple slice which acted as a control for this experiment. Pictures of the slices were taken every 60 minutes for up to 3 hours or longer, and image analysis software was used to measure the color change in the apple slices as they turned brown over time. The experiment will determine the fruit most efficient in limiting how quickly apples turn brown and can be used to estimate which fruit would have the greatest possibility of reducing excessive melanin production in human skin. The effective fruit in this experiment can help dermatologists create more effective hyperpigmentation products , while using natural ingredients.

A2-BCMM Biology: Cellular, Molecular, or Microbiology
Using Herbal Antimicrobial Properties to Mitigate Antibiotic Resistance
David Letelier
Naser Abdelhadi

This experiment dives into the antimicrobial efficiency of different herbal extracts compared to the pharmaceutical antibiotic Bacitracin against non-pathogenic E. coli. The main objective was to see whether or not herbal remedies could serve as viable alternatives to help fight against antimicrobial resistance (AMR). The materials that were used through this project include, but are not limited to, garlic (Allium Sativum), rosemary ( Rosmarinus officinalis), thyme (Thymus vulgaris), sage (Salvia officinalis), ginger (Zingiber officinale), and honey extracts, as well as Bacitracin for comparison. Inhibition zones were measured, serving as the quantitative data displaying the effectiveness of each treatment. The results showed that garlic (Allium Sativum) and sage ( Salvia officinalis) had the greatest inhibition zones in terms of size, while the other herbs, including rosemary (Rosmarinus officinalis), thyme (Thymus vulgaris), ginger (Zingiber officinale), and honey, displayed little or no significant antibacterial activity. These findings hint at the fact that specific herbs may propose a potential alternative to pharmaceutical antibiotics in the everlasting fight against AMR. In parts of the world that lack the infrastructure for pharmacies, pharmacists, or pharmaceutical products, such as many rural or developing regions, these herbal alternatives could provide a sustainable, cost-effective solution for preventing bacterial infections and reducing the overreliance on antibiotics . Additionally, many of these regions already grow these plants/herbs that could work even better, meaning they could use existing resources to their advantage. Further research is needed to evaluate the broader applicability and safety of these herbal remedies on a much larger scale.

A3-BCMM Biology: Cellular, Molecular, or Microbiology
Optimization of HRT Imaging Cell Counting Protocol in Study of UVA Effects on Fuchs Endothelial Corneal Dystrophy
Zimon Li

Fuchs Endothelial Corneal Dystrophy (FECD) is a degenerative eye condition where corneal endothelial cells deteriorate, leading to corneal swelling, vision impairment, and blindness. This research examines the role of ultraviolet A (UVA) light. One of the key methodologies used is HRT (Heidelberg Retinal Tomograph), a quantitative imaging tool that analyzes corneal cell density and morphology. This study investigates how UVA light and certain medicines alter corneal endothelial cells in mice and how HRT counting enhances data collection precision . Using mouse corneal models, we applied HRT to quantify cell numbers in corneal tissue before and after exposing the corneas to UVA light . Our findings demonstrate that HRT counting provides critical insights into endothelial cell changes, offering high-resolution data on cell density and morphology. This approach allows the accurate monitoring of disease progression and the assessment of a pill’s efficacy . The study underscores the importance of integrating advanced imaging techniques in ophthalmologic research and contributes to the broader understanding of FECD preventative methods. These results pave the way for refining diagnostic methods and developing novel treatments for corneal health.

A4-BCMM Biology: Cellular, Molecular, or Microbiology
In Silico Usage of Cordyceps Militaris and its Effectiveness on NSCLC Cellsx
Jianquan Zhao
Juns Ye

Globally, cancer treatments have been the primary research direction in the past decades, with innovation in treatment
options increasing alongside costs. Due to the dominancy of cancer cases being non-small cell lung cancer (NSCLC),
Cordyceps Militaris can be used as a potential drug to inhibit NSCLC activity with its unique bioactive metabolite,
cordycepin, a key component to the fungus’ immunotherapeutic properties . Through in silico methods utilizing molecular docking software, observations of the interactions between the cordycepin ligand and target proteins for NSCLC were made. Proteins 8EXL, 4JPS, 4AGD, and 1M17 were targets of experimentation because they activate pro-cancer pathways. This study also considers the toxicity of the usage of cordycepin. Creation of cordycepin analogs was done to address toxicity issues while improving the binding affinity score of cordycepin . A machine learning model was used to determine the toxicity, using known toxic and non-toxic molecules to train the model to differentiate between the two . After docking cordycepin onto each of the proteins respectively, we concluded that it had an average binding affinity of -6.51 kcal/mol, moderate binding affinity however toxic. Through minimal to moderate modifications, a cordycepin analog [2- (3-Hydroxymethylphenylamino)benzamide] was created and has a stronger binding affinity score of –8.175 kcal/mol that is non-toxic, with considerations of ligand efficiency when compared to other analog candidates. Future additions to this study include in vivo or in vitro experimentation for additional validation, as well as finding an effective analog possible through genetic engineering rather than chemical synthesis .

G10-BCMM Biology: Cellular, Molecular, or Microbiology
5 Second Rule
Madeline Neary

The purpose of this project is to test the “five-second rule” and determine if pathogens can transfer onto food dropped
onto the floor in under five seconds. The hypothesis was that if cookies were dropped for less than five seconds, then
bacteria and other pathogens would still transfer on to them. To test this hypothesis, the scientist separately dropped five chocolate chip cookies in five spots in the same area on the Bishop Feehan High School cafeteria floor. Each cookie was dropped for a different amount of time. They were dropped for one, two, three, four, five, and six seconds. The control cookie was not dropped. Next, Sterile swabs were used to swab each cookie. They were then streaked on separate agar plates. The agar plates were incubated for four days, and observed once each day to measure growth and record data. The expected outcome was that all the plates except for the control would show growth. The outcome was that only the cookie dropped for three seconds showed significant growth, growing two colonies of bacteria. This outcome supported the hypothesis and disproved the five-second rule. This knowledge can be used to prevent the spread of disease and food-borne illness. If people refrain from eating food that has fallen on the floor, even for less than five seconds, they can protect themselves from ingesting these harmful pathogens.

G11-BCMM Biology: Cellular, Molecular, or Microbiology
Molecular Profiling of Stem Cell-Derived Neurons: Evaluating Their Potential to Replace Huntington’s Disease-Affected
Neurons
Pranav Addanki

Huntington’s disease (HD) is a progressive neurodegenerative disorder caused by a CAG repeat expansion in the HTT
gene, leading to widespread neuronal dysfunction and cell death, particularly in the striatum and cortex. Currently, there are no disease-modifying treatments, making stem cell-derived neurons a promising therapeutic option. However, their effectiveness depends on how closely they replicate the molecular and functional properties of healthy neurons . This study analyzed publicly available RNA sequencing (RNA-seq) and proteomics datasets to compare the molecular characteristics of healthy neurons, HD-affected neurons, and stem cell-derived neurons. Differential gene expression and pathway enrichment analyses were performed to examine key neuronal functions, including synaptic signaling, neuroprotection, and mitochondrial activity. Additionally, machine learning models, including random forest classifiers and support vector machines, were trained to classify neurons into these groups and identify the most important genetic and protein markers driving classification. The results showed that while stem cell-derived neurons exhibited molecular profiles more similar to healthy neurons than HD-affected neurons, critical differences remained. Synaptic signaling pathways, including the expression of key neurotransmitter receptors (GRIN1, GABRA1), were downregulated in stem cell-derived neurons, potentially affecting their ability to integrate into functional neural circuits . Additionally, mitochondrial function, particularly genes involved in oxidative phosphorylation (NDUFS3, ATP5F1), showed incomplete restoration, suggesting potential metabolic vulnerabilities. Neuroprotective mechanisms, such as brain-derived neurotrophic factor (BDNF) signaling, were also weaker in stem cell-derived neurons, which may reduce their resilience against oxidative stress and excitotoxicity. Machine learning models achieved high classification accuracy and identified these key pathways as major distinguishing factors between neuron types. These findings suggest that while stem cell-derived neurons represent a promising approach for HD treatment, further optimization is needed to enhance their functional properties. Future work will focus on refining differentiation protocols, improving synaptic connectivity, and enhancing mitochondrial and neuroprotective pathways to increase their therapeutic potential.

G12-BCMM Biology: Cellular, Molecular, or Microbiology
Understanding GBM Relapse Mechanisms though scRNAseq Analysis and Experimental Validation of Functional Genes
Wenna Mao

Glioblastoma Multiforme (GBM), the most common and lethal brain tumor, exhibits a high relapse rate of 90% within 2
years. Traditional therapies involve surgical resection followed by chemo /radiotherapy; however, nearly all patients shortly develop resistance to treatment. In this study, we analyzed the Single-Cell RNA-Seq data of 70 patients diagnosed with primary or relapsed GBM. We detected significant upregulation of genes associated with tumor microenvironment and cell migration in relapsed patients, among which CTNNA3 and MBP are among the top 5 significant genes. These genes were predominantly expressed in neural-oligo precursor cells (OPCs) – cells that share characteristics with Glioblastoma Stem Cells (GSCs) in their capability for self renewal and differentiation – suggesting their critical role in tumor recurrence by “regressing” the tumor to a more developmentally primitive state that enhances their survival under treatment pressure. To functionally validate these findings, we introduced CTNNA3 and MBP recombinant proteins at different concentrations into LN229 and T98G GBM cell lines treated with temozolomide (TMZ). Wound-healing and CCK8 assay results showcase how CTNNA3 and MBP enhance cell migration and viability in a dose-dependent manner. Immunofluorescence analysis showed increased collagen expression in the extracellular matrix, suggesting the genes’ role in engaging fibroblasts to reinforce a protective OPC/GSC niche, likely contributing to persistent treatment challenges. These findings indicate that CTNNA3 and MBP contribute to tumor plasticity, invasion, and chemoresistance, highlighting their potential as therapeutic targets.

G4-BCMM Biology: Cellular, Molecular, or Microbiology
Hormonal Diseases in Black Women: A Statistical, Computational and Experimental Analysis of the Impacts and Influences of Common Practices and Solutions
Camilla Royal

Black women have historically been subjected to unethical research practices, and their health concerns often remain overlooked. They face a higher risk of developing uterine cancer, endometriosis, and related complications, such as infertility, weight fluctuations, and chronic pain. Despite this heightened risk, Black women are less likely to be diagnosed with these conditions compared to white women. One potential contributing factor to these health disparities is the widespread use of cosmetic products, including chemical hair straighteners and skin lightening agents, which are often influenced by societal pressures to conform to Eurocentric beauty standards. These products expose users to harmful chemicals, such as formaldehyde, calcium hydroxide, and propyl paraben. This study investigates the intersection of social and chemical influences on Black women’s health through three analytical approaches: statistical, and experimental. The results reveal that Black women are more likely than white women to use chemical hair straighteners and skin bleaches, with nearly 55% of Black women reporting the use of hair straighteners and 1 in 40 using skin bleaching agents. Hormonal disruptions were noted in three participants who had been exposed to these chemical products. Research indicates that exposure to these chemicals disrupts estrogen receptor pathways, potentially leading to infertility, cellular damage, and other endocrine-related issues. Experimental wet lab findings confirm that exposure to chemical hair straighteners results in ovarian damage, including mutations, cell death, and genomic instability. Future research will expand on these findings by incorporating E. coli testing and genomic sequencing, as well as further examining the effects of these chemicals on earthworm ovaries using advanced imaging and staining techniques. This study aims to contribute to better diagnostic practices for conditions like endometriosis and uterine cancer, ultimately improving healthcare outcomes for Black women.

G5-BCMM Biology: Cellular, Molecular, or Microbiology
The Neuroprotective Role of Vitamin C in Mitigating Hypoxia-Induced Oxidative Stress and Enhancing Viability in C.elegans: Insights into a Model for Post-Stroke Recovery
Jacklyn Omere-Okundaye

Each year strokes take the lives of over 5 million people and forever change the lives of nearly 50% of survivors. Strokes are caused by restricted blood flow to the brain which in turn leads to a condition called hypoxia. This condition occurs when oxygen levels drop to 1-2%. When oxygen is reintroduced through a process called reperfusion reactive oxygen species (ROS) are generated at great amounts. An excess amount of ROS leads to what is known as oxidative stress. At the cellular level oxidative stress can destroy the lipid and proteins that make up a cell, essentially leading to its death. The cells that are often restricted of oxygen during a stroke are neurons which control the way an organism can move, cognition, breathe, all other daily functions. The aim of the current study was to test if vitamin C can be used to protect neurons and other somatic cells within C.elegans from oxidative stress that forms after exposure to hypoxic conditions. To test this C.elegans were exposed to vitamin C at two concentrations (0.1mmol and 1.0mmol) and they were exposed to hypoxic conditions for three different durations of time (2, 24, and 72 hours). C.elegans viability and behavior were tested in this study. All data were gathered using light microscopy methods. The results of this study indicated that vitamin C can be used to protect C.elegan viability and their neuromuscular functions when given at a high dose (1.0mmol) after a 72 hour exposure to hypoxic conditions. The results of this study also showed that vitamin C can display pro-oxidant behaviors when vitamin C is given under certain conditions. The results from this study can be used to further research vitamin C’s role in protecting neurons from the oxidative stress that occurs after a stroke.

Keywords: Strokes, Oxidative Stress, Antioxidants, Vitamin C, Neurons, Hypoxia, Reactive Oxygen
Species, Reperfusion injury

G7-BCMM Biology: Cellular, Molecular, or Microbiology
Bacteria behind the Beauty: What Lip Gloss Applicator Holds the Most Bacteria?
Emily Carter

This experiment evaluates the bacteria levels when testing three lip gloss applicators with E.coli. The study will compare
the presence and quantity of bacteria on each applicator type: a doe-foot applicator, an angled plastic applicator, and a
metal roller ball. Each applicator will be tested for bacterial growth after simulating the application. The bacterial colonies will be put into liquid cultures measured on the nanodrop on the OD600 setting to determine the degree of contamination. This research seeks to identify potential hygiene concerns and inform professionals in the cosmetic industry about the risks involved with different applicators. The results will provide valuable insight into improving designs and minimizing bacterial contamination throughout the community.

G8-BCMM Biology: Cellular, Molecular, or Microbiology
The Effectiveness of Natural Cleaners Against Bacteria Compared to Generic Cleaners
Emma Giovenelli

This project seeks to compare the effectiveness of cleaners marketed as natural to generic cleaners and to figure out
what components of the cleaners contributed to their sanitization ability. This was tested by running Kirby-bauer on E.coli on LB agar plates, incubating them at 37 C for 24 hours, and recording the zone of inhibition of the cleaners on each plate with working positive and negative controls in millimeters. Each sample was tested 30 times.

G9-BCMM Biology: Cellular, Molecular, or Microbiology
The Effects of Oxidative Stress on Huntington’s Disease
Lilya Fouda

This experiment investigates the effects of oxidative stress on Huntington ‘s disease using yoghurt, specifically
Lactobacillus. This experiment also allows one to understand the cellular impact of neurological diseases such as
Huntington’s disease. Yogurt that had Lactobacillus present was used and varying concentrations of hydrogen peroxide were added to produce oxidative stress. 6 hours later, after measuring the initial pH of the yoghurt, the pH was measured and the thickness of the yoghurt was observed. This allows an individual to be enlightened on the relationship between oxidative stress and its neurological impact. It could also help lessen the symptoms of Huntingtons’disease based on this experiment.

H10-BCMM Biology: Cellular, Molecular, or Microbiology
Fermentation Time and pH-Driven Antimicrobial Effects of Kombucha on E. coli and S. epidermidis
Joseph Yoon

This study investigates the antimicrobial properties of black tea and green tea kombucha by examining how pH and
fermentation duration influence the growth inhibition of Escherichia coli (Gram-negative) and Staphylococcus epidermidis (Gram-positive). Using liquid and solid media assays, bacterial susceptibility was analyzed across different fermentation stages to assess the time-dependent and pH-mediated antimicrobial effects of each kombucha type. The results showed that black tea kombucha generally exhibited stronger antimicrobial activity against E. coli and S. epidermidis compared to green tea kombucha, with inhibition varying over time and correlating with pH changes. These findings provide valuable insights into the potential of kombucha, particularly its tea base, as a natural antimicrobial agent, contributing to microbiology, food science, and the development of alternative antimicrobial strategies.

H11-BCMM Biology: Cellular, Molecular, or Microbiology
How Different Hand Soaps Affect Bacterial Growth
Elizabeth Coach
Hannah Callini

Hand washing removes bacteria and debris from hands. This experiment focuses on finding the hand soap that best
prevents bacterial growth and what active ingredients are the most effective . This research is important especially during flu season and because of COVID 19 so people can be educated on what brand best prevents bacterial growth. The methods began by growing bacteria in sterile broth for 24-48 hours. Prepare a petri dish with agar and let it solidify. Once solidified, swab grown bacteria onto agar surface using the lawn method, sterilize tweezers, use them to dip filter paper into hand soap and then place it in the center of the agar plates. Incubate agar plate (agar pointing down) for 48 hours. After incubation measure the zone of inhibition around filter paper in millimeters and repeat 2 more times for each bacteria. It was hypothesized that if different hand soaps are tested then bacterial growth will vary because of different active ingredients and concentrations of active ingredients. The data resulted in soft soap being the most effective for Staphylococcus epidermidis with the highest average measurements on the zone of inhibition with about 8.666667 mm. With E Coli it was found that Purell was the most effective with an average measurement of 9.333333 and overall JR. Watkins was the most effective on average between the two bacterias tested. Based on the data, it was concluded that Benzalkonium chloride was the most effective active ingredient for killing Staphylococcus epidermidis and Sodium Lauryl Sulfate , Cocamidopropyl Betaine, and Sodium Chloride were the most effective together as active ingredients for killing Escherichia coli bacteria . More research will be done on the idea of accessibility, cheap vs expensive hand soaps would be tested to see which has the most benefits while using a price point. Other bacteria or organisms could be substituted along with also testing natural hand soaps.

H12-BCMM Biology: Cellular, Molecular, or Microbiology
Identification of Potential Genomic Biomarkers for the Personalized Treatment of Colorectal Cancer Utilizing Bioinformatics Approach
Anvi Sarmah

Colorectal cancer (CRC) is the third most common cancer and the second leading cause of cancer-related deaths globally. CRC development is primarily driven by two mechanisms of genomic instability: chromosomal instability (CIN) and microsatellite instability (MSI). MSI is observed in approximately 13-16% of CRC cases, with MSI-high (MSI-H) tumors linked to defective DNA mismatch repair (MMR) and tumor suppressor gene silencing. While MSI-H tumors often respond to immunotherapy, the variability in patient responses signifies the need for more reliable biomarkers to predict therapeutic outcomes. This study hypothesizes that identifying mutations in genes frequently altered in MSI-H CRC and their association with distinct immune cell infiltration patterns and gene expression profiles—indicative of a heightened immune activation state—may reveal potential predictive biomarkers for improved immunotherapy outcomes. Conversely, frequently mutated genes in MSI-H CRC that do not share these characteristics would be better biomarkers for alternative therapy. MSI-H CRC patient data from cBioPortal was examined to test this hypothesis, and the top two frequently mutated genes, KMT2D and ARID1A, were identified. Co-occurrence analysis of these mutations with classical MSI-H marker genes, such as MLH1, MSH6, MSH2, and MSH3, revealed frequent co-mutation of KMT2D and ARID1A with these genes. To determine whether gene expression patterns in CRC with KMT2D or ARID1A mutations resemble those in MSI-H, particularly for immune checkpoint genes (PD-1, PD-L1, and PD-L2), mRNA expressions in tumors harboring these mutations were analyzed. The results showed that MSI-H tumors with these mutations exhibit elevated expressions of immune checkpoint genes, suggesting a potential immune evasion mechanism with implications for immunotherapy. Pathway analysis using Reactome further revealed enriched expression of genes associated with an activated immune system in CRC tumors harboring KMT2D or ARID1A mutations. Moreover, immune cell infiltration analysis using TIMER2.0 showed increased infiltration of CD8+ and CD4+ T cells along with macrophage cells, suggesting a more immunogenic tumor microenvironment in CRC with ARID1A or KMT2D mutations. In contrast, mutations in RIT1 and GNAS, which are also associated with MSI-H CRC but less prevalent and mutually exclusive to KMT2D and ARID1A mutations, do not exhibit an immune cell infiltration phenotype, which indicates that MSI-H CRC harboring these mutations may not respond to immunotherapy. These findings suggest that mutations in genes such as ARID1A and KMT2D, along with their associated immune signatures, could serve as biomarkers to inform personalized treatment strategies for MSI-H CRC, potentially enhancing immunotherapy efficacy.

H1-BCMM Biology: Cellular, Molecular, or Microbiology
STING Inhibits Migration and Invasion in Triple-Negative Breast Cancer Independently of Immune Cells
Jason Xie
Neal Tandon

Triple-negative breast cancer (TNBC), negative for estrogen receptor, progesterone receptor, and HER2, is the most
aggressive subtype of breast cancer, with high metastatic potential and limited treatment options. Recently, STING
(Stimulator of Interferon Genes) agonists have shown therapeutic potential to stimulate anti-tumor immunity via activation of the cGAS-STING pathway in immune cells. However, STING’s involvement in metastatic processes and tumor cells remains poorly explored and thus needs immediate investigation to ensure drug safety and efficacy . Here, we investigated whether STING would affect migration and invasion, key drivers of metastasis, in an immune-independent role. We demonstrate that STING in tumor cells significantly suppresses TNBC migration and invasion , independently of its immunomodulatory functions by assessing in vitro migration and invasion capabilities without immune cell co-culture. Across repeated transwell assays, there were statistically significant differences between STING-overexpressing and control cell lines. Overexpression was confirmed by Western Blot analysis. Cell proliferation was unaffected by STING overexpression as determined by cell counting and doubling time assay. Additionally, we employed and trained an image segmentation deep-learning algorithm with Cellpose AI tool to overcome image analysis challenges. Our findings strongly indicate the existence of non-immune STING pathways that critically influence metastasis, with STING in tumor cells playing a more direct role in tumor metastasis. Our data stressed the need to further understand STING biology for its use as a therapeutic target in triple-negative breast cancer.

H2-BCMM Biology: Cellular, Molecular, or Microbiology
Exploring the Role of PI3K_ in Cell Growth
Angelina Fang

Phosphatidylinositol 3-kinase (PI3Ks) cadence pathways not only translate extracellular triggers into intracellular signals, but also regulate cell phenotype and many other metabolic processes. PIK3CB, which encodes the protein PI3Kβ, is a gene in the PI3K family that has not yet been researched extensively. In this procedure, we built and utilized pHAGE-PIK3CB plasmids to produce transiently and stably overexpressing PIK3CB cells. We further used these two systems to examine PI3Kβ function under the treatment of either a general growth hormone (insulin), PI3Kβ agonist (lysophosphatidic acid (LPA)), or inhibitor (GSK2636771). Our findings showed that PI3Kβ is able to induce AKT (protein kinase B) activation, resulting in increased cell growth.

H3-BCMM Biology: Cellular, Molecular, or Microbiology
Biophysical Determinants of T-Cell Receptor Signaling in Danio rerio
David Aubrey

Cancer immunotherapy is an emerging tumor-fighting solution that uses T-cells in the body’s natural immune system to
target and kill cancer cells. T-cells, a variation of white blood cells developed in the thymus, have cell-specific killing
properties which cancer immunotherapy solutions use to eradicate cancerous cells in the body without deleterious effects to unaffected tissues. This project investigates the role of the plexinD1 gene in T-cell functionality and migration using the model organism Danio rerio to better understand the role of genetic engineering in T -cell cancer immunotherapies. Three groups of cxcr4b:Tract-mCherry (cxcr4b transgenic fluorescent reporter line) zebrafish were studied: a group of mutant zebrafish with a CRISPR-Cas9 plexinD1 gene knockout (induced through in vivo embryo microinjection), a mock dye-injected control group, and an uninjected control group. These groups were grown in the same conditions until 6 days post-fertilization, with confocal imaging occurring on days 5 and 6 to examine reporter-generated fluorescence shown in lymphatics, circulatory vessels, and the thymus to determine the possible phenotypes produced by the plexinD1 knockout. The plexinD1 knockout was expected to negatively affect T-cell progenitor migration to and within the thymus during development and cause leukocyte retention in circulatory vessels from migratory issues. Results showed signs of improper leukocyte migratory behavior in circulatory vessels and the thymus through fluorescence measurements. Data also showed aberrant lymphatic branches in plexinD1 mutants compared to controls, suggesting that plexinD1 is involved in lymphatic vessel regulation. This study suggests an emerging role of plexinD1 in leukocyte motility and vascular development.

H4-BCMM Biology: Cellular, Molecular, or Microbiology
The Effects of Terpene, a Plant Compound, on Reducing Lobaplatin-Induced Neurotoxicity
Miyu Hasegawa
Piyusha Majgaonkar

Chemo brain also known as chemotherapy-induced neurotoxicity is an often overlooked side effect of chemotherapy that is experienced by 75% of cancer patients after their treatments. Chemo brain causes brain fog which leads to severe memory problems, lack of mental sharpness, and other cognitive impairment issues. Chemotherapeutic agents trigger apoptosis and interfere with mitochondrial DNA which causes an increase in reactive oxygen species causing oxidative stress. This can damage DNA, RNA, and proteins, and can cause a reduction in ATP synthesis which leads to an overall sense of brain fog . Reducing this brain fog requires a treatment that reduces the presence of apoptotic proteins and promotes the production of anti-apoptotic proteins. Terpene is an organic plant compound that can be used as a treatment as it specifically targets the platinum-based compounds in chemotherapy agents like lobaplatin, a widely used chemotherapy drug. Terpenes can reduce the side effects of chemo brain without impacting treatment efficiency. C. elegans were used as a model to measure neurotoxicity. This project aims to determine if terpene, specifically alpha-pinene terpene can be used as a treatment to reduce the symptoms of lobaplatin-induced neurotoxicity. For experimentation, the independent variable was the terpene treatment and the dependent variable was the
number of thrashes in the C. elegans indicating corresponding neurotoxicity levels. In the experimental groups, the terpene counteracted the effects of the lobaplatin with the C. elegans having healthy thrashing levels. After experimentation, it was determined that terpene can be used as a treatment to reduce lobaplatin-induced neurotoxicity.

H5-BCMM Biology: Cellular, Molecular, or Microbiology
Power of Poop: Identification and Analysis of Gut Microbiome Biomarkers in Diarrhea
Chenxi Hu

The human gut microbiome, a vast collection of billions of bacteria in the gastrointestinal tract, plays a crucial role in maintaining host health through homeostasis. Diarrhea, which affects approximately 179 million people annually, significantly alters the gut microbiome. While infectious diarrhea is characterized by an increase in pathogens such as Escherichia coli, the changes in gut microbiome following non-infectious and chronic diarrhea remain unclear. Recent advances in metagenomic sequencing have enabled us to investigate the changes in gut microbiome. To identify biomarkers in diarrhea and characterize the gut microbiome , we integrated datasets from 12 published studies into a comprehensive dataset, comprising 1015 healthy samples and 2302 diarrhea-related disease samples, among which approximately 90% are non-infectious. Using linear regression model and machine learning, we identified microbial taxa enriched in healthy samples, including Bacteroides uniformis, Bacteroides vulgatus, Akkermansia muciniphila, Firmicutes bacterium, and Alistipes. In contrast, Flavonifractor plautii, E. coli, and Prevotella were more abundant and prevalent in diarrhea samples. For microbial function analysis, we found that pathways such as phytol degradation and aerobic respiration I (cytochrome C) were enriched in healthy samples, whereas other pathways including glycerol
degradation, methanogenesis, isoprene biosynthesis, and taxadiene biosynthesis, were more commonly detected in
diarrhea samples. We employed graph theory and inferred a denser but less stable network from diarrhea samples compared to the network of healthy controls. These results indicate that healthy and diarrhea gut microbiomes have distinct features characterizing diarrhea alterations. We anticipate that these biomarkers can be used for personalized dietary interventions and targeted, cost-effective diagnosis and treatments of diarrhea that consider complex gut microbiome interactions.

Keywords: Gut microbiome, Diarrhea, Taxonomic biomarkers, Functional pathway, Machine learning

H6-BCMM Biology: Cellular, Molecular, or Microbiology
The Creation of Mutant GJB2 Knockout Xenopus Utilizing CRISPR Cas9
Oleksandr Kaplenko

About 430 million people in the world experience genetic hearing loss, approximately 15% of American adults (37.5
million) ages 18 and over have significant hearing loss and must utilize hearing aids. This problem could be solved by using CRISPR Cas9 technology. CRISPR is a gene-editing technology that allows modifying DNA by cutting and replacing specific genetic sequences. The GJB2 gene provides instructions for making connexin 26, a protein essential for cell communication in the inner ear, and mutations in this gene are a common cause of genetic hearing loss. The first two steps of the experiment were to design sgRNA to guide Cas9 to the gene target and to design primers to amplify regions of DNA. The next step was to template guide synthesis, making RNA from DNA. Then Targets one and two were microinjected into Xenopus embryos. After microinjection, the frogs grew for about three weeks before the tail clipping procedure was done. Lastly, the PCR procedure was performed, which included extracted DNA, amplified DNA, and cleaned DNA. If Xenopus tropicalis was injected with CRISPR Cas9, targeting frameshift mutations in the gene GJB2, then the homozygous frog would experience hearing loss. Results were measured in two categories, indel percentage, and knockout score percentage (KO). From Target one, seven samples were successfully knocked out, and from Target two, two samples were successfully knocked out. In conclusion samples # 10, 2, and 4 from Target one, and sample #2 from Target two are most likely to have GJB2 knockout.

H7-BCMM Biology: Cellular, Molecular, or Microbiology
Defining the Role of abcg2a on the Hematopoietic Processes in D. rerio
Taleena Gonneea

The emergence of hematopoietic stem cell (HSC) transplantation as a therapeutic technique for hematological disorders
has made the clarification of relevant genes necessary. The ABC transporter family, proteins that regulate cellular material by transmembrane translocation, is theorized to be involved in hematopoiesis. The human gene ABCG2, and corresponding zebrafish gene abcg2a, are members of the ABC family that have been suggested to influence hematopoietic processes . The aim of this project was to determine the role of abcg2a coded proteins in hematopoiesis, specifically in HSC differentiation and proliferation. It was hypothesized that the role of abcg2a would correspond with cellular proliferation that would decrease as chemical dosage increased. The chemical CAY10719, a beta-carboline derivative, was used to inhibit the gene. Three day old zebrafish embryos were exposed to dimethyl sulfoxide, 0.1 uL, and 1.0 uL of CAY10719. At 4 days, post fertilization images were taken. After isolating the ribonucleic acids, samples were assessed using quantitative polymerase chain reaction (qPCR). The qPCR analysis validated the assertion that an increase of drug concentration would yield decreased gene expression. The highest dosage group exhibited larger macrophages, indicating that the abcg2a gene has a stronger correlation with HSC self-renewal properties than with HSC differentiation. While drug treatment reduced macrophage population, macrophages continued to display routine activity, supporting the hypothesis that cell population size relates to the expression of the abcg2a gene. While a function of abcg2a was specified, there are still major developments needed to improve HSC transplantation and substantiate theories concerning hematopoiesis.

H8-BCMM Biology: Cellular, Molecular, or Microbiology
Demographic and Genetic Insights into ARID1A Mutations in Hepatocellular Carcinoma: Implications for Prognosis and Treatment
Agastya Sarmah

Liver cancer ranks as the sixth most common cancer and the third leading cause of cancer-related deaths globally,
posing significant challenges for patients and healthcare providers. Hepatocellular carcinoma (HCC) accounts for majority of the primary liver cancer cases and has a disproportionate impact across genders, races, and ethnicities. One of the primary challenges in developing effective therapies for HCC is tumor heterogeneity . Understanding biological diversity of these tumors, particularly across different demographics, is crucial to understand the tumor biology and to develop therapeutics. In this study, I analyzed data from 2,351 HCC patients available on cBioPortal, focusing specifically on HCC with ARID1A mutations. ARID1A mutations are common in HCC, and they may facilitate tumor progression from adenoma to carcinoma, likely through dysregulation of ARID1A functions in chromatin remodeling, transcription, replication and DNA repair. I observed that HCC patients with ARID1A mutations had a significantly shorter median survival of 23.46 months compared to 53.29 months for those without ARID1A mutations. HCC patients with ARID1A mutations exhibited a higher frequency of mutations in genes related to the cell cycle, DNA repair, gene expression, metabolism, cellular stress responses, with significant enrichment in WNT signaling, mRNA processing, DNA damage repair, collagen degradation, and glucose metabolism. Notably, there appears gender and racial disparities in HCC patients carrying ARID1A mutations. ARID1A mutations were more prevalent in males than females and highest in Black patients, who also had the lowest median survival compared to White and Asian patients. Comparison of genomic alterations between genders revealed that females had significantly higher mutation in genes involved in inflammatory cytokines production, glucose and mRNA metabolism, carbohydrate digestion, small molecule transport, DNA damage repair, and senescence. Black patients had significantly more mutations in genes associated with gene expression, immune system, signal transduction, DNA repair, and metabolism of lipid, protein, and RNA compared to Whites or Asians. These demographic disparities and distinct genomic alterations identified may bring insights into underlying HCC biology and developing personalized treatments. Finally, as an attempt to identify potential therapeutic targets, I compared the genomic alterations of disease- free ARID1A-mutated HCC patients with deceased ARID1A-mutated HCC. The analysis identified 26 unique gene mutations present only in disease-free patients, predicted to form synthetic lethal pairs with ARID1A leading to cell death when both genes are inactivated. Notably, among them was the gene encoding POLQ, involved in DNA damage response and repair, was mutated in 15.38% of the disease-free individuals. Survival analysis revealed that no events observed in cases with combined ARID 1A and POLQ mutations, suggesting that the combined mutations may confer a survival advantage to patients with HCC. ARID1A/POLQ co-expression correlation analysis showed their strong positive correlation in the normal liver tissue and altered relationship in liver cancer. Together, this study highlights POLQ as a novel target for ARID1A-mutated HCC and provides a foundation for further exploration of ARID1A/POLQ interactions as a therapeutic strategy for ARID1A mutated HCC.

H9-BCMM Biology: Cellular, Molecular, or Microbiology
An ABC Transporter Mutant of Mycobacterium smegmatis Resistant to a Membrane-Fluidizing Drug
Jaelyn Luo

Mycobacteria cause many diseases including tuberculosis and leprosy, and the plasma membrane is the final barrier of
the sugary and greasy armor of mycobacterial cells. Research has shown that dynamic organization of mycobacterial
plasma membrane is critical for the creation of the cell envelope. Mycobacterium smegmatis, a non-pathogenic cousin of dangerous pathogens like Mycobacterium tuberculosis, is a model used for tuberculosis research. Since different genes in Mycobacterium smegmatis could potentially contribute to the plasma membrane dynamics, our research focused on investigating the function of a gene MSMEG_3119.

The MSMEG_3119 gene was previously implicated in resistance to a membrane fluidizing drug known as dibucaine. MSMEG_3119 encodes a putative ATP-binding cassette (ABC) transporter that move substances across cellular membranes, but its function is unknown. Using knockout techniques, the mutant Δ3119, which lacks the MSMEG_3119 gene, was generated. We first confirmed previous research prediction that Δ3119 recovers faster than the wild type when treated with dibucaine. We further found that Δ3119 was less permeable to a membrane-impermeable dye than the wild type when the cells are treated with dibucaine. Using Mass Spectrometry for protein analysis, we showed that the fatty acid chains of phosphatidylinositol (PI) could be composed of either C16:0 palmitic acid / C18:1 oleic acid or C16:0 palmitic acid / C19:0 tuberculostearic acid (TBSA). The higher quantity of saturated fatty acid C16:0 / C19:0 PI accumulated in cell membrane of Δ3119 compared to the wild type is a possible reason for its less permeable membrane. Altogether, our research results suggest that the ABC transporter encoded by MSMEG_3119 may be involved in transporting oleic acid (C18:1)-containing phospholipids across the plasma membrane from the inner leaflet to the outer leaflet.

KEYWORDS: mycobacteria; plasma membrane; fatty acid; cell envelope; phospholipids

I10-BCMM Biology: Cellular, Molecular, or Microbiology
Effects of GABA Treatment on Locomotive Behavior Recovery in Sleep-Deprived Drosophila Melanogaster
Hannah Cha

The prevalence of sleep disorders and chronic sleep deprivation in modern society has become a significant public health concern as the consequences of inadequate sleep extend beyond mere fatigue, impacting varying aspects of human health and function. This study investigates the effects of gamma-aminobutyric acid (GABA) on the recovery rate of locomotive behavior in sleep-deprived Drosophila melanogaster, utilizing the open-field assay to observe their movements. Wild-type flies were categorized into eight groups: female and male, put in an 18-6 LD cycle or 12-12 LD cycle, observed under GABA treatment, and no GABA treatment. Results revealed that the substitution of GABA in wild-type flies results in an increased rate of activity, whereas flies fed a traditional diet showed a continuous decrease in locomotor activity. Although the data collected supports the hypothesis, further research is needed to confirm the relationship between GABA application and locomotive behavior.

Keywords: Sleep, GABA, locomotive behavior, open-field assay

I11-BCMM Biology: Cellular, Molecular, or Microbiology
Effect of Nigella Sativa Oil on Reducing Oxidative Stress
Anjali Daga

Oxidative stress, which occurs at the cellular level, can be extremely detrimental to humans’ health. Oxidative stress
contributes to cancer, Alzheimer’s, Parkinson’s, heart disease, strokes and more. The project being carried out is aiming to provide information on whether nigella sativa oil is able to decrease oxidative stress, as measured via heart rate.
I will be looking at the effects of nigella sativa oil on reducing oxidative stress in daphnia magna.

I12-BCMM Biology: Cellular, Molecular, or Microbiology
Synergistic Effects of Plant Extracts and Silver Nanoparticles in Combatting Environmental Bacteria
Dua Rafaqat

Antibiotic resistance is a growing global concern, requiring alternative solutions to combat bacterial infections. This
experiment hypothesized that combining silver nanoparticles with plant extracts (honey, garlic, ajwain, and clover) would enhance their antibacterial activity against environmental bacteria. To test this, plant extracts were prepared, filtered, dried, and redissolved, then mixed with silver nanoparticles. Bacterial cultures in LB Broth were exposed to these solutions, and OD600 measurements were recorded at 16 and 24 hours. The results showed that garlic extract with silver nanoparticles had the strongest antibacterial effect, significantly reducing bacterial growth compared to plant extracts or silver nanoparticles alone. These findings support that silver nanoparticles can enhance the antibacterial properties of natural extracts, offering a potential approach for developing antimicrobial treatments.

I1-BCMM Biology: Cellular, Molecular, or Microbiology
Effects of Probiotic Supplementation on Migraine-Related Reversals in C. elegans
Adelaide Mims

Migraine is a common neurological disorder that creates symptoms like intense head pressure, nausea, blurred vision, vomiting, and extreme sensitivity to light, sound, touch, and smell. Migraines disrupt normal neurotransmission leading to intense symptoms that impair daily life. Current treatments are hard to navigate and can be ineffective and expensive , highlighting the necessity for new and innovative approaches. It has recently been discovered that the gut brain axis, which connects gut health and brain function, may play a role in migraines. This study explored the link between the gut and migraines using C. elegans as a model. The UNC-2 genotype in C. elegans mimics human migraines, with abrupt directional changes representing migraine episodes. Since C. elegans feed on bacteria, two strains, wildtype and UNC-2, were given one of three bacterial diets, and their movements were observed for one minute. Two of the bacterial diets consisted of probiotic bacteria, which are known to improve gut health. Results showed that UNC-2 C. elegans fed the probiotic diets of Lactobacillus casei or Bifidobacterium lactis had fewer migraine-like episodes than those on an E. coli diet, suggesting a possible link between gut health and migraines.

I2-BCMM Biology: Cellular, Molecular, or Microbiology
Investigating the Neuroprotective Potential of T. Ammi Extract on Parkinson’s Disease via Gut-Brain Axis Modulation
Vyshnavi Donthabhaktuni

Parkinson’s disease (PD) is a common neurodegenerative disease (NDD) that is characterized by the gradual loss of dopaminergic neurons and progressive motor impairment. Recent studies suggest that PD may originate in the gut, highlighting the gut-brain axis (GBA) as a critical area for research. This study investigated the neuroprotective potential of Trachyspermum ammi (T. ammi) oil on PD-associated symptoms using Caenorhabditis elegans (C. elegans) as a model organism. This study will assess various PD symptoms including impaired locomotion, dopaminergic neuron degeneration, elevated reactive oxygen species (ROS) levels, and gut permeability. PD was induced in C. elegans using genetic models, and varying concentrations of T. ammi oil were incorporated into the worms’ food. Behavioral assays (locomotion, thrashing) as well as physiological assays (oxidative stress, alpha-synuclein levels, and gut permeability), were conducted to assess the impact of T. ammi oil on these parameters. Results from these assays suggested a positive effect of T. ammi on Parkinson’s, with the ajwain-treated groups showing improved locomotion, thrashing, improved survival under oxidative stress and even lowered alpha-synuclein levels. The findings support the hypothesis that T. ammi oil mitigates PD-like symptoms in C. elegans. The results of this study could be further applied to develop an efficient , cost-effective, and widely available treatment for mitigating PD symptoms in humans.

Keywords: Parkinson’s, C. elegans, ajwain, gut-brain axis (GBA), treatment.

I3-BCMM Biology: Cellular, Molecular, or Microbiology
Ayurvedic Treatment as Oral Hygiene Alternatives for the Prevention of Dental Caries
Divi Bhaireddy

Dental caries, primarily caused by Streptococcus mutans (S. mutans), are a major public health concern. 3.5 billion people worldwide are affected by oral disease and approximately 2 billion people struggle with dental caries. Traditional antimicrobial agents have limitations, prompting interest in natural alternatives. This study investigates the effect of bay leaf (Laurus nobilis) extract on S. mutans biofilm formation using a crystal violet assay. Biofilms were grown in a 12-well plate with sucrose to enhance formation, and treated with bay leaf concentrations of 0.25%, 0.5%, 1%, 2%, and 5%. Transmittance levels were measured to assess biofilm inhibition . Results indicate a positive correlation between bay leaf concentration and biofilm density when treatments are higher concentrated with bay leaf, but a negative correlation with lower concentrations leading to increased transmittance. However, greater variability was observed at higher concentrations, as indicated by larger error bars. A one-way ANOVA test confirmed statistical significance in the effect of bay leaf on biofilm formation . These findings suggest that bay leaf extract may have potential as a natural antimicrobial agent against S. mutans, though further research is needed to optimize its application.

I4-BCMM Biology: Cellular, Molecular, or Microbiology
Synergistic and Individual Optimization of Reproductive Health and Chronic Stress Resistance in Caenorhabditis elegans Using Resveratrol and Astaxanthin
Jayaratna Deshamouni

Globally, infertility affects 186 million people. It is caused by mitochondrial dysfunction, genomic instability, and chronic inflammatory dysregulation due to oxidative stress (OS). These pathologies accelerate reproductive decline, gametogenic disturbance, and embryonic developmental disruptions, and they are the root cause of neurodegenerative diseases like Parkinson’s and Alzheimer’s. Furthermore, they even cause metabolic syndromes like PCOS and Type 2 Diabetes, and cardiovascular degeneration. Interventions that focus on the bioenergetic, proteostatic, and epigenetic aspects of reproductive aging are necessary because conventional assisted reproductive technologies (ARTs) are still not as efficient, failing to address the molecular causes of reproductive decline. In Caenorhabditis elegans, a nematode with genetically tractable model with conserved reproductive, metabolic, and stress-response pathways, the cumulative effects of resveratrol and astaxanthin—strong redox regulators and mitochondrial bioenergetic enhancers—on reproductive longevity and stress resistance are examined in this work. A SIRT1 activator, resveratrol coordinates mitochondrial biogenesis, FOXO3-mediated antioxidant reactions, and p53-controlled genomic stability, whereas astaxanthin triggers Nrf2-Keap1 signaling, which alters glutathione metabolism and inhibits cascades of inflammatory cytokines. The phenotypic evaluations were carried out using microscopy-based tracking, AI-assisted longevity analytics, and statistical modeling using One-Way ANOVA, Tukey HSD post-hoc analyses, and machine learning-derived nonlinear regression. The well-executed experimental design combined antioxidant-enriched E. coli preconditioning, nematode growth medium (NGM) plate-based interventions, and H₂O₂-induced OS paradigms. Beyond the effects of monotherapy, the combined antioxidants produced significant benefits, significantly increasing lifespan efficiency, gametogenic throughput, and reproductive span (p < 0.00001). Epigenetic recalibration and metabolic reprogramming were identified as key mechanisms in computational simulations that demonstrated exponential increases in reproductive resilience and systemic stress adaptation. With practical implications for clinical infertility treatments, metabolic homeostasis, and neuroprotective longevity, these findings mark a quantum leap in antioxidant-based reproductive therapeutics. For instance, translational implications for metabolic homeostasis, neuroprotective longevity paradigms, and clinical infertility interventions. The next generation of bioinformatics-driven fertility restoration and systemic age-reversal strategies will be made possible by future research that uses multi-omic profiling, mammalian validation models, and precision-dosed antioxidant solutions.

I5-BCMM Biology: Cellular, Molecular, or Microbiology
Post-Transcriptional Regulation of Huntington’s Disease: An Antisense Oligonucleotide Therapeutic Strategy for Inhibiting Mutant Exon 1 Expression
Aryan Shah
Simrit Kukreja

Huntington’s Disease (HD) is a genetic neurodegenerative condition characterized by the loss of locomotive control. 41,000 Americans suffer from HD, most patients passing within 8-10 years of diagnosis. The HTT gene codes for huntington protein, in which the first exon’s poly-glutamine tract is expanded when mutated. Misfolded proteins form aggregates and cause toxicity. The traditional pharmaceutical approach has been to prescribe patients with dopamine suppressors and schizophrenia medications to solely treat symptoms, which come with several side effects, namely depression. Antisense oligonucleotide (ASO) sequences have the potential to act as inhibitory molecules and prevent the expression of HD by inducing enzymatic cleavage and blockage of mRNA translation. This project aimed to engineer and test a particular ASO sequence’s ability to inhibit the expression of mutant exon 1 in a bacterial model. Two trials were performed to confirm the impact of intracellular glutamine concentrations as the ASO inhibitor concentration increased. Trial 1 demonstrated that increasing ASO in bacterial samples induced with standard mRNA concentrations correlated with a decline in intracellular glutamine levels , however, samples with excess mRNA saw no change. These results indicated that homozygous dominant individuals may require increased ASO concentrations to offset malignant protein expression , whereas heterozygous individuals require lower concentrations. 10% error bars between the highest and lowest concentration groups of ASO (0.0 nM and 10.0 nM) confirmed significance. The ASO strategy opens avenues for the treatment of niche diseases, bypassing drug side effects and more effectively regulating HD promoting prognoses.

I6-BCMM Biology: Cellular, Molecular, or Microbiology
The Effects of Moringa Oleifera and Nigella sativa on Influenza A Virus
Meena Lakshmanan

Influenza A is a common respiratory virus that is highly contagious and can range from being mild to deadly . In the United States of America from 2023-2024, about 34 million people are ill from Influenza and about 17,000 of them died (CDC, 2024). This study aims to find an alternative natural substance to help reduce the replication of Influenza A. When someone has the flu in the U.S., they go to the pharmacy and buy over the counter drug called tamiflu (oseltamivir). This is a great drug to reduce the days that someone experiences flu symptoms; however, tamiflu is not accessible to everyone, especially those who live in other countries. For example, underdeveloped countries have limited access to oseltamivir and healthcare. Looking at the accessibility of Moringa oleifera and Nigella sativa, they are easily accessible and widespread across these countries. The present study employs the MDCK cell culture as a model that was infected with the WSN/33 strain of Influenza A and tested the antiviral effects of M. oleifera and N. sativa. There were five different concentrations of M. oleifera leaf powder, N. sativa seed powder, M. oleifera seed oil, and N. sativa seed oil in order to record cell death. It was hypothesized that M. oleifera and N. sativa have an antiviral effect on Influenza A minimizing Influenza A replication, which would ultimately lower cell death. Through this experiment, the hypothesis was not well-supported by the data as all but one of the concentrations had no significant difference in cell death in comparison to the cell culture with the virus infected. The N. sativa oil at a 1:40 concentration shows some promise for an antiviral effect because the amount of cell death was not nearly as low as all the other concentrations, and it was significantly higher than the cell treated with the virus, but still lower than oseltamivir.

Keywords: Moringa oleifera, Nigella sativa, Influenza A, oseltamivir

I7-BCMM Biology: Cellular, Molecular, or Microbiology
Pathophysiological Disruption of the Endothelial Glycocalyx by Microplastic Contamination
Avani Jain

Endothelial cells are a vital group of cells found in both blood and lymphatic vessels and are responsible for many physiological functions in the human body, the most significant being production of the glycocalyx. The glycocalyx is a
carbohydrate-rich fibrous layer that lines the surface of endothelial cells and plays several key roles in the human body.
However, small plastic particles, known as microplastics, can be found in the blood vessels, potentially posing a threat to the endothelial cells and the glycocalyx. These microplastics enter the body through various methods, including inhalation and ingestion. They are newly discovered and thus there is little research on them – but they are known to be very toxic.

The goal of this study is to investigate the effects these microplastics can have on endothelial cells , and consequently, the glycocalyx. This study utilizes two of the most common types of microplastics – polyethylene and polypropylene – to investigate the effects of microplastics on the viability of human lung endothelial cells . A trypan blue exclusion assay, ImageJ, and Cell Profiler were used to analyze cell viability and morphology. As shown through the study, microplastics do impact both cell viability and morphology. This study aims to showcase the dangers of microplastics in hopes of conveying the true threat these small plastic particles pose. Through this, the study hopes to show the need for further research into microplastics and for an alternative form of plastic.

I8-BCMM Biology: Cellular, Molecular, or Microbiology
Does Diet Supplementation with Fructooligosaccharides Reduce Motor Symptom Severity in a D. melanogaster Model of Parkinson’s Disease?
Mia Solomon

Parkinson’s Disease (PD) is a neurodegenerative disorder affecting approximately 10 million people worldwide. With no known cure, current treatments focus on symptom management. However, they are often expensive and come with side effects requiring additional medication . This study investigates the potential of prebiotics in modulating the severity of PD symptoms using Drosophila melanogaster as a model organism. Prebiotics were administered from embryos to assess their impact on PD symptom progression. Despite being arthropods, D. melanogaster share 50-60% of human genes and exhibit complex nervous system functions, making them a valuable model for neurodegenerative disease research. Prior studies suggest that prebiotics, such as fructooligosaccharides, enhance short-chain fatty acid (SCFA) production, which may reduce α-synuclein accumulation, a hallmark of PD. It was hypothesized that prebiotics would improve locomotive performance in PD- affected D. melanogaster compared to untreated counterparts. While the results showed a trend toward enhanced locomotion in treated flies, they did not reach statistical significance, indicating that further research is needed to establish a definitive correlation.

I9-BCMM Biology: Cellular, Molecular, or Microbiology
The Effect of Curcumin on Beta-Amyloid Plaques in the Brain.
Adil Siddiqui

Alzheimer’s disease, the leading cause of dementia, is characterized by the accumulation of beta-amyloid plaques that impair neuronal communication and accelerate cognitive decline. Despite extensive research, preventive therapies effective in curbing progression remain challenging to establish. Curcumin, a bioactive compound derived from turmeric (Curcuma longa), has exhibited anti-inflammatory and neuroprotective properties. However, its effectiveness in preventing beta-amyloid aggregation requires further investigation. This research employed an in vitro model to mimic the brain environment through the use of agarose gel infused with beta-amyloid protein and Congo red dye for fluorescence-based analysis. Curcumin solutions (10-60 μM) were introduced to beta-amyloid samples, and plaque formation was measured by fluorescence microscopy and image analysis software. The results showed a dose responsive reduction in the formation of plaque, and the highest concentration of curcumin (60 μM) caused an 89.3% decrease in plaque number and 87.8% diminution of fluorescence intensity compared with the control group. Statistical analysis confirmed the relevance of these reductions (p < 0.05), and the hypothesis was supported that curcumin performs in an inhibitive manner in plaque aggregation. These findings point to curcumin’s potential as a preventative measure for Alzheimer’s disease through its inhibition of beta-amyloid plaque deposition. Due to curcumin’s low bioavailability, follow-up research should examine new delivery systems in order to optimize its therapeutic effects. This research closes the distance between traditional medicine and modern neuroscience, providing a foundation for the exploration of natural products as neuroprotective agents.

J10-BCMM Biology: Cellular, Molecular, or Microbiology
Efficacy and Mechanisms of Reprograming Factors on ex Vivo HDF Rejuvenation.
Kejia Ni

Aging is a natural process that significantly impacts global health, vitality, and quality of life for people. Here, we report several exciting recent advances in the field of partial cell reprogramming, a potential anti-aging approach that rejuvenates cells without converting cellular identity as in reprogramming. We have developed a novel mRNA-based method that utilizes a cocktail of 6 mRNA-coded reprogramming factors and an in vitro treatment regimen that robustly and reproducibly improves well-known cellular biometrics in a stepwise fashion.

Using human skin-derived fibroblasts as a model for aged cells, we demonstrate that our 6F-mRNA method significantly improved numerous cell parameters: enhances mitochondrial function by increasing membrane potential and reducing dysfunctional mitochondria, improving metabolic state without altering mitochondrial content, facilitates cell cycle progression by increasing S-phase entry, reducing G2/M accumulation, and modulating senescence pathways, lowering p53 and p21 expression without activating p16^INK4A, promotes telomere extension, and restores embryonic fibroblast-like properties, increasing Nestin expression, clustered growth, and migration across scratch assay.

Overall, these findings reinforce and add to current evidence suggesting that mRNA-directed cell rejuvenation is feasible. Studies on an extensive panel of human iPSC-derived cells are currently underway in research and cGMP settings that can be used to support future preclinical and clinical studies.

J11-BCMM Biology: Cellular, Molecular, or Microbiology
Investigating the effects of Novel-Transcription Regulator X in HEK293 Cells on Kidney Fibrosis
Ayanna Rohil

 

J12-BCMM Biology: Cellular, Molecular, or Microbiology
The Effects of Nutmilk on UTI-Causing Bacteria
Isabella Scarnici

The susceptibility of E. coli in response to Almond Milk, Walnut Milk, and Cashew Milk was tested based on the size of the zone of inhibition (the area around an antibiotic where bacteria colonies could not grow) when the bacteria were exposed to each of these kinds of milk. The exposure of Almond Milk to E. coli was expected to obtain an average 15% higher than the untreated E. coli. To test this, petri dishes were filled with Nutrient Agar, and left to set overnight. After streaking each dish, 5 μL of each milk was dropped onto a 6 mm content disc and placed onto a cultured plate of Escherichia Coli K-12. After being placed in a 37Åã incubator overnight, the zone of inhibition was measured from top to bottom and subtracted by 6 mm to determine the size of the zone surrounding the disc. Results showed E. coli is not susceptible (resistant or stronger than) to any of the conditions. To be considered susceptible, the average diameter needed to be 15 mm or more. Averages, including and excluding numerous zeros throughout the data, displayed 0-3 mm. The characteristics of almonds, walnuts, and cashews were responsible for E. coli’s resistance to each condition. By understanding the characteristics of these nuts, further experimentation can be done. This further experimentation could contribute to the relief of people across the world suffering from E. coli-related infections. 

J1-BCMM Biology: Cellular, Molecular, or Microbiology
Investigating the Anticancer Potential of Herbal Extracts on Saccharomyces Cerevisiae Models of Cancer Cell Proliferation.
Ananya Mathur
Gianna Nalumansi
Mariyam Patel

Cancer remains one of the leading causes of mortality globally, with traditional treatments like chemotherapy and radiation often resulting in severe side effects and emerging treatments such as immunotherapy resulting in limited effectiveness . This research investigates the anticancer potential of herbal extracts from Licorice Root, Ginger Root, Turmeric, and Garlic, evaluating their impact on the proliferation and viability of Saccharomyces cerevisiae, a yeast model of cancer. It was hypothesized that these herbal extracts would reduce cell proliferation and increase cell death due to the bioactive compounds believed to have anticancer properties. 

The study involved treating yeast cells with varying concentrations of herbal extracts (1%, 5%, and 25%) and monitoring cell growth. Initially, the experiment yielded no results, prompting a modification in methodology. Fresh produce herbs were used, and a traditional extraction method involving boiling the herbs in water was implemented. The growth medium was also changed from nutrient agar to Sabouraud Dextrose agar. Despite these changes, the experiment has yet to yield significant results regarding the reduction in cell proliferation or increased cell death. These findings suggest that further optimization of experimental conditions, such as concentration levels or experimental models, may be necessary. 

While no conclusive results were obtained, this study contributes valuable insights into the challenges of using herbal extracts in cancer research, providing a foundation for future investigations to refine experimental approaches and explore potential anticancer properties in more complex models.

J2-BCMM Biology: Cellular, Molecular, or Microbiology
Testing Ice from Local Fast Food Restaurants for Microbial Contamination.
Muhammad Chaudhry

Foodborne illnesses are typically thought of being contained to raw meat or uncleaned vegetables but it extends to items like ice as well. While most of these other foods are checked regularly by fast food restraunts to ensure quality, some fast food restaurants are notorious for not cleaning some of their machines, resulting in those foods either not being served, or being served regardless of the possible contamination. Therefore, it is important to identify the possible contamination in the local area. For this reason, ice from a fast food restaurant will be tested by plating the samples to see if they contain microbial contamination. Ice will be gathered from a fast food restaurant in a sterilized jar before being plated. It will be plated on a sample of regular nutrient agar, nutrient agar with bromothymol blue, and lastly a combination of bromothymol blue and ampicillin within the nutrient agar. The results will likely come back with heavy contamination for the regular nutrient Agar plates, and will come back with light contamination for the plate with ampicillin in the medium. The alternative result would likely be the nutrient Agar plates containing very little to no contamination , and the plate with ampicillin in its medium would have little contamination.

J3-BCMM Biology: Cellular, Molecular, or Microbiology
Unmasking The Data: Evaluating the Effectiveness of Different Materials in Preventing Bacterial Growth
Clara Hubbard

Masks are a great breeding ground for bacteria. This experiment focuses on finding different materials of masks , and how this can prevent or provoke bacteria growth on the inside of the mask due to the bacteria from your mouth. Evaluating different materials and their ability to prevent bacteria from adhering and growing on masks can greatly impact the decision of what masks someone chooses to wear. Three masks, a Cambridge Co Basic mask (PM10 filter and a three ply-micro particulate layer), Old Navy non-medical grade Mask (100% cotton), Virtue Code second skin cloth mask (100% polyester) were sprayed with an E-coli and water mixture, then stored in a room temperature environment. Over the next two weeks, data was collected on the amount of colony forming units found in the petri dishes. It was hypothesized that the Cambridge Co Basic mask would be found to have the most bacteria in it, due to filtered masks being unable to be worn in hospitals, and the three ply-micro particulate layer. The data resulted in a high amount of bacteria ranging from 28 – 31 colonies in the Cambridge mask, while the cloth and polyester mask had very little to no bacteria. Based on the data, the Cambridge mask promotes the most bacteria growth of all three masks. More research will be done with a more advanced variety of masks, as well as adjusting the pH of the water used in the E-coli spray, and the temperature in which the masks are stored.

J4-BCMM Biology: Cellular, Molecular, or Microbiology
Terminating Cancerous Cells through the Mitochondrial Pathway of Apoptosis with the Insertion of the Bim Protein
Haley Bilodeau

 

J5-BCMM Biology: Cellular, Molecular, or Microbiology
Increasing Accessibility of Methodology for Diagnosing Chronic Traumatic Encephalopathy During Life
Kyra Moody

Chronic Traumatic Encephalopathy (CTE), a neurodegenerative disease caused by repetitive traumatic brain injuries (TBI), is currently only able to be diagnosed after death. Research in the CTE field is growing exponentially, but the methods being used are costly and not easily accessible. This experiment was conducted to explore the possibility of more accessible and cost-efficient areas of methodology for CTE. Human TBI protein expression was modeled through Drosophila melanogaster in which TBI were delivered. Protein analysis through column chromatography, using isolated proteins from Drosophila melanogaster, is hypothesized to demonstrate detecting human TBI proteins. The drosophila protein solution was run through the column chromatography, and proteins were detected. The findings from this experiment are promising and future application will hopefully allow for during-life diagnostic research to be done for CTE cost effectively.

J6-BCMM Biology: Cellular, Molecular, or Microbiology
Investigating the Role of HSF-1, SKN-1, and DAF-16 in Regulating Oxidative Stress Response in C. elegans
Saketh Madhusudhan

 

J7-BCMM Biology: Cellular, Molecular, or Microbiology
Assessing the Influence of ALAN (Artificial Light At Night) on Soil Microbial Diversity and Invertebrae Ecosystems
Tanya Wongchaisuwat

This project investigates the effects of artificial light at night (ALAN) on soil microbiomes, which are essential for nutrient cycling and ecosystem health. Given the intricate relationships between soil, plant, and human microbiomes, understanding how ALAN disrupts microbial diversity is crucial for promoting sustainable agricultural practices and protecting soil health as urbanization increases.

J8-BCMM Biology: Cellular, Molecular, or Microbiology
Effect of Vitamin Oils on the Growth of Saccharomyces Cerevisiae Cells
Cayah Maciaszek

 

J9-BCMM Biology: Cellular, Molecular, or Microbiology
Investigating Generational Trauma in Caenorhabditis elegans
Anoushka Ksheersagar
Arushi Vora

This study aimed to better understand the impact of generational trauma in Caenorhabditis elegans. It was hypothesized that generational trauma would increase linoleic acid levels, consequently decreasing the lifespan of C. elegans. As a form of trauma, the C. elegans were heat shocked at 37ÅãC. Initially, lipids were going to be measured directly by extracting them through the Bligh and Dyer method and thin layer chromatography, to measure the levels of linoleic and oleic acid. However, due to issues with accessing a laboratory space, a different approach was utilized. Instead of lipid extraction, photos were taken to observe general changes in C. elegan behavior. These photos were analyzed by a convolutional neural network machine learning program to analyze changes in worm morphology following a heat shock. It was observed that worms straightened and formed clumps after being heat shocked. Clumps of worms may have been formed due to rapid aging, which would indirectly indicate that the heat shock caused a shortened lifespan. Since oleic acid correlates to C. elegan longevity, and has an inverse relationship to linoleic acid, this also means that the levels of linoleic acid increased, supporting the initial hypothesis. Furthermore, statistical analysis suggested that the percentage of worms with a normal level curvature decreased sharply after the heat shock and never fully recovered to their original levels. Inversely, the percentage of worms with a lower curvature increased after the heat shock and stayed abnormally high. This suggests that the heat shock may have induced muscular neurodegeneration and neurodegeneration, possibly resulting in the worms being straighter. Ultimately, the goal of this research is to use the collected data to develop treatment options for health conditions associated with generational trauma, such as specialized linoleic acid or oleic acid supplements to mitigate some of the health risks. Additionally, this will increase understanding towards the effects that generational trauma may have on other organisms such as humans.

K10-BCMM Biology: Cellular, Molecular, or Microbiology
Nutrition and Alzheimer’s Disease: Analyzing the Effect of Fenugreek Extract with and without Turmeric on Amyloid Beta in Alzheimer’s Disease Using Caenorhabditis elegans
Medha Sri

Alzheimer’s disease (AD) is a progressive neurodegenerative disorder characterized by the accumulation of amyloid-beta plaques and tau tangles, leading to cognitive decline. Recent studies have linked Type 2 diabetes (T2D) to an increased risk of AD, suggesting shared metabolic dysfunctions. Natural compounds such as fenugreek and turmeric have been investigated for their neuroprotective properties, yet their combined effect on AD pathology remains unclear. This study aimed to evaluate the impact of fenugreek and fenugreek with turmeric on the longevity of Caenorhabditis elegans (C. elegans) expressing human amyloid-beta. Using CL2355 (AD model strain), CL2122 (control strain), and AB1 (wild-type), worms were exposed to treatment solutions and monitored for survival over five days. Initial trials encountered contamination and methodological challenges, necessitating multiple refinements. The final experiment showed that fenugreek and fenugreek with turmeric did not improve survival but instead worsened mortality, with both treatment groups experiencing 100% decline compared to the 50.91% decrease in the untreated control. The findings suggest that these treatments may exert toxic effects rather than neuroprotective benefits under the tested conditions . Further research, including optimized dosage concentrations and alternative biomarkers such as amyloid aggregation, is needed to assess the true therapeutic potential of these natural compounds in AD models.

K11-BCMM Biology: Cellular, Molecular, or Microbiology
Improving Early Lung Cancer Detection Through Use of Neural Networks in Conjunction with LDCT Imaging
Brian Xu

 

K12-BCMM Biology: Cellular, Molecular, or Microbiology
Bacterial Resistance
Ryan Thornberry
Xavier Reynolds-Zuckerman

This experiment is being done to document DH5 Alpha E. coli mutations as it becomes resistant to antibiotics. Given the ever-increasing prevalence of antibiotic resistance, an issue exacerbated in accordance with misadministration and overuse of antibiotics impactful especially in hospitals worldwide, both awareness and understanding of the problem are critical. Understanding the mutation/evolution in antibiotics will play a role in assisting doctors to create informed decisions about treatments. To do this, various antibiotics such as kanamycin and ampicillin will be utilized. Multiple petri dishes with Agar will contain a gradient of antibiotic . It is expected that the bacteria will mutate and develop the ability to exist in areas with high concentrations of antibiotics. To document the mutagenic growth of the E. coli, pictures will be taken to document daily growth. As an added experiment, bacteria will be taken from the edge of their growth. That bacteria should have the most antibiotic resistance out of all the bacteria. These bacteria will then be grown in an agar plate and antibiotic disks will be placed upon it, this is being done to determine of antibiotic resistance bacteria will have increased resistance against other antibiotics.

K1-BCMM Biology: Cellular, Molecular, or Microbiology
Effects of Non-Nicotinic E-Liquid and Nicotinic E-Liquid Exposure and Ameliorating Properties of Hesperidin on Locomotion in C. elegans
Deepika Bhardwaj
Nandita Ganesh
Yoyo Wu

E-cigarettes are most commonly used by teenagers and young adults. In 2024, around 1.63 million American underage students reported present e-cigarette use, a concerning figure. While nicotine consumption through e-cigarettes is seen as safer due to their lack of tobacco, a carcinogen, the side effects of their usage, especially those related to non-nicotinic ingredients of e-liquid, have yet to be fully investigated. Non-nicotinic components include vegetable glycerin (VG) and propylene glycol (PG), alongside a selection of flavorings; exposure of C. elegans to these compounds has been implicated to increase oxidative stress, decrease lifespan, and inhibit development. Nicotine is a drug known to induce addiction pathways and nicotine dependence; it is also a known agonist of nicotinic acetylcholine receptors (nAChRs). Nicotine dependence in C. elegans is characterized by increased immediate movement as a withdrawal symptom due to over-expression of nAChRs. Hesperidin is a flavone glycoside commonly found in citrus fruits with ameliorative effects such as antioxidant, anti-inflammatory, anticarcinogenic, and neuroprotective properties. C. elegans lifespan increased significantly when treated with hesperidin, which may reduce oxidative stress associated with aging. We investigated the effects of hesperidin on components of e-liquid—VG, PG, Euro Gold flavoring, and nicotine—to promote healthier lifestyles and mitigate the harms of e-cigarette usage through the benefits hesperidin and other flavonoids may provide. Euro Gold e-liquid, equivalent in effect to PG/VG, was found to significantly decrease all strains of C. elegans’ body bends per second; hesperidin’s tandem treatment, however, alleviated the decrease significantly. Nicotine dependence and nicotine-mediated increase in locomotion were also compounded by hesperidin, suggesting hesperidin upregulates acr-16 and likely promotes organism-wide stress response through α7-nAChR pathways. Further, hesperidin ameliorated locomotion in pink-1-mutant C. elegans, implicating its potential autophagy- inducing and therapeutic effects in the treatment of Parkinson’s Disease, Alzheimer’s Disease, and other neurodegenerative disorders.

K2-BCMM Biology: Cellular, Molecular, or Microbiology
Inference of Cell-Cell Communication to Understand the Molecular Mechanisms of Alzheimer’s Disease
Adrita Samanta

Alzheimer’s disease (AD) is a neurodegenerative disorder that affects around 55 million adults worldwide. The molecular and cellular mechanisms of AD are still poorly understood due to its complex etiology. The brain relies on complex cellular interactions between the neural and non-neural cells to maintain homeostasis. This communication is disrupted in Alzheimer’s disease. In AD, microglial cells detect β-amyloid plaques and activate astrocytes which causes neuroinflammation and neuronal damage. However, the effect of disrupted intercellular communication remains largely unexplored . I hypothesized that identifying these dysregulated interactions, and the cell types involved, will reveal key mechanisms driving AD progression. In my study, I investigated perturbations in cell-cell communication across different stages of the disease as well as between male and female patients, to uncover these mechanisms and identify potential therapeutic targets. Using single-nuclei RNA-sequencing data from over 2.3 million nuclei in the prefrontal cortex of 427 patients across control, intermediate-AD, and late-AD stages, I predicted cell-cell communication between the 29 cell types using a LIgand- receptor ANalysis frAmework (LIANA) and developed a unique approach to compare signaling patterns between patients. My results show that disruptions during the earlier stages of AD affect insulin and calcium signaling, while changes during the later stages contribute to β-amyloid accumulation and blood-brain barrier dysfunction. My research also revealed sex- specific variations in AD, with female and male patients exhibiting differences in neuroinflammation, neurovascular dysfunction, blood-brain barrier integrity, and synaptic function. This is the first study to provide a detailed understanding of how dysregulated interactions affect the onset and progression of AD by comparing cell-cell communication across multiple stages of AD, highlighting key differences between male and female AD patients. My findings emphasize the need for personalized therapeutic strategies and offer future directions for targeted treatment and drug development in AD.

K3-BCMM Biology: Cellular, Molecular, or Microbiology
Assessing Goldenrod Extract as a Potential Antibiotic: The Effects on Bacteria and Eukaryotic Cell Viability
Ayaan Garg

Antibiotic resistance presents a global health crisis, with the overuse and misuse of antibiotics driving the emergence of multidrug-resistant bacterial strains. This research investigates the potential of goldenrod (Solidago virgaurea) extract as a natural alternative to traditional antibiotics by assessing its effects on Escherichia coli (E. coli) and yeast cells. E. coli was used to evaluate the extract’s antibacterial efficacy, while yeast viability was tested to assess its safety for eukaryotic cells. The E. coli experiment measured inhibition zones following exposure to goldenrod extract, methanol (negative control), and streptomycin (positive control). Goldenrod extract produced an average zone of inhibition of 16.38 mm, demonstrating significant antibacterial activity compared to methanol (mean: 4 mm) but less than streptomycin (mean: 24.35 mm). One-way ANOVA analysis (p < 0.01) confirmed statistically significant differences among treatment groups, highlighting goldenrod’s potential as an antibacterial agent . In the yeast experiment, cells were exposed to two concentrations of goldenrod extract (0.5 mL and 1 mL), with viability assessed using methylene blue staining. Both concentrations preserved high viability (mean: 98.82% for 0.5 mL and 98.60% for 1 mL), comparable to the control group (96.17%). Methanol significantly reduced viability (mean: 67.37%), reinforcing the extract’s safety and underscoring the importance of distinguishing between antimicrobial efficacy and cytotoxicity . These findings support goldenrod extract as a promising candidate for further development as a plant-based antibiotic, offering an effective and safe solution to address antibiotic resistance. Future studies should focus on identifying active compounds, optimizing extraction methods, and expanding its application to other microbial species.

K4-BCMM Biology: Cellular, Molecular, or Microbiology
Reengineering the Fluorescence Spectra of GFP with CRISPR Adenine Base Editing
Cayden Bairos
Luira Leite
Ricardo Contreras Chacon

Using Adenine Base Editor (ABE) to perform a single nucleotide mutation to covert super folder GFP to super folder BFP inserted into Escherichia coli.

 

K5-BCMM Biology: Cellular, Molecular, or Microbiology
Developing CRISPRa As a Potential Therapeutic for SYNGAP1 Haploinsufficiency in Syngap1- Encephalopath
Daniel Chen

Synaptic Ras GTPase-activating protein (SYNGAP) is found at the junctions (synapses) between nerve cells in the brain and is critical for regulating synaptic plasticity, which is essential for learning, memory, and cognitive functions. Mutations in the SYNGAP1 gene often result in a partial loss of SYNGAP expression, leading to haploinsufficiency and, consequently, a range of neurodevelopmental disorders and conditions, including intellectual disability, autism spectrum disorder, and sometimes epilepsy. This project aims to develop a CRISPR activation therapy for treating encephalopathy caused by SYNGAP1 haploinsufficiency by increasing SYNGAP expression from the remaining functional copy of the SYNGAP 1 gene.

K6-BCMM Biology: Cellular, Molecular, or Microbiology
Can Slime Mold Be Used To Kill Bacteria as a Disinfectant
Nevin Lebron

This project explores the use of slime molds, specifically Dictyostelium discoideum, as a sustainable disinfectant against K12 E. coli. Slime molds, which act as decomposers and reproduce via spores, play a crucial role in nutrient recycling. Given the health and environmental risks associated with traditional disinfectants, this study hypothesizes that slime molds can effectively reduce E. coli populations on contaminated surfaces by consuming bacteria. The experimental procedure involved cultivating slime molds alongside E. coli under controlled conditions, using aseptic techniques. Initial results showed successful growth and interaction, supporting the hypothesis of their antimicrobial potential. Future research will assess the stability of bioactive compounds and regulatory considerations, indicating that slime molds could offer a cost-effective and eco-friendly alternative to conventional disinfectants.

K7-BCMM Biology: Cellular, Molecular, or Microbiology
A Study Utilizing Planarians: Identifying an Optimal Range of Tomato Lycopene Consumption to Reduce Prostate Cancer
Hafsa Naim
Hemam Henok
Maria Alves

This scientific study investigates the effects of lycopene on cancerous tumor growth using planarians , or flatworms, as a model instead of traditional tumor cells. Lycopene, an antioxidant, is known for its ability to prevent cellular damage. By testing various concentrations of lycopene on planarians, the research aims to translate these findings to human cancer treatment. Given the lack of a definitive cure for cancer, this study seeks to identify a temporary solution that can reduce tumor effects. The experimental procedure involved measuring the length of the planarians and assessing the Blastema area, which are regenerative cells of organs of body parts. After initial measurements (cm), planarians were treated with lycopene concentrations of 2.5%, 5%, and 10% of their body weight, while a control group received no lycopene to compare normal growth patterns. Each planarian was allowed to sit in the lycopene solution for 24 hours before measurements were repeated. The results indicated that the average Blastema area for the different groups was as follows : 2.5% lycopene had an average area of 165,395.08, 5.0% had 183,546.57, and 10% had 169,898.31, while the control group averaged 222,920.65. These findings provide insights into the potential effects of lycopene on tumor growth , laying the groundwork for further research in cancer treatment.

K8-BCMM Biology: Cellular, Molecular, or Microbiology
Decoding Aging: Multi-Omics Insights into Oxidative Stress, Mitochondrial Dysfunction, and Cellular Senescence in
Fibroblasts
Rajarshi Mandal

This research investigates the complex biochemical mechanisms underlying aging by analyzing primary human fibroblasts using a longitudinal multi-omics dataset. This dataset includes cytology, DNA methylation and epigenetic clocks, bioenergetics, mitochondrial DNA sequencing, RNA sequencing, and cytokine profiling. Key findings indicate that mitochondrial efficiency declines with age, while glycolysis becomes more prevalent to compensate for energy demands. Epigenetic clocks, such as Hannum and PhenoAge, showed strong correlations with biological age (ρ > 0.650, p < 1e-6), validating the experimental setup and confirming that the cultured fibroblasts were aging appropriately. Fibroblasts with SURF1 mutations exhibited accelerated aging, marked by bioenergetic deficits, increased cell volume, and reduced proliferative capacity, underscoring the pivotal role of mitochondrial dysfunction in cellular senescence. Novel insights were gained from analyzing cytokines like IL18 and PCSK9, some of which were linked to age-related diseases such as Alzheimer’s and cardiovascular disorders. Experimental treatments revealed distinct effects on cellular aging. Dexamethasone reduced inflammation but also increased DNA methylation , induced metabolic inefficiencies, and shortened cellular lifespan. Oligomycin heightened oxidative stress and RNA degradation, emphasizing how such treatments contribute to cellular stress and metabolic imbalance while shedding light on aging mechanisms. By uncovering connections between mitochondrial dysfunction, epigenetic biomarkers, and immune dysregulation, this study identifies potential therapeutic targets for age-related diseases. Future research could validate the most promising biomarkers across diverse cell types and experimental treatments to build a more comprehensive understanding of aging.

Keywords: Aging, Bioenergetics, Cellular Senescence, Cytokine, Cytology, Dexamethasone, DNA Methylation, Epigenetic Clock, Fibroblast, Mitochondrial DNA Sequencing, Multi-Omics, Oligomycin, Oxidative Stress, RNA Sequencing, SURF1 Mutation.

K9-BCMM Biology: Cellular, Molecular, or Microbiology
Sugar and Ice: Sweet Survival Strategy for Brine Shrimp in the Deep Freeze
Alyssa Yasuhara
Hongyi Ni

This project explores the implementation of different cryopreservatives and their effects on invertebrate species such as the Brine Shrimp (Artemia spp.). Cryopreservation success depends mainly on using cryoprotectants, which prevent ice crystals from forming under fragile tissues. Most commonly found as glucose or urea in species such as the wood frog (Rana Sylvatica), cryoprotectants prevent the cellular structures from collapsing. Studies in this project will investigate the roles of different sugars such as trehalose, glucose, glycerol, saccharin, and sucrose, in cryopreservation. These sugars are known to create highly concentrated solutions which lowers the freezing point of water, thus protecting cellular structures and potentially increasing the survival rates during processes of extreme freezing and thawing. Nauplii (Larvae of Brine Shrimp) were hatched and treated with different sugars in their solutions . Frozen with trehalose, glucose, glycerol, saccharin, and sucrose, Nauplii were kept at -20.C for 24 hours. Viability was measured by counting the survival of individuals after 1 hour of thawing. The results that followed, presented a 0% success rate of survival in any solution of sugar. Therefore, suggesting that the passive absorption of sugars from the solutions was not sufficient for protection from extreme measures such as freezing. To further explore the effects of these same sugars, however, in less extreme conditions, a second experiment was conducted. For 36 hours, Nauplii were stored at 4.C. The results of this experiment represented the success of trehalose, with a survival rate being 95%, compared to glucose which had a a survival rate of 72%. Although not extreme, the experimentation of a stress-induced environment (low temperatures) results in observable data on the effects the sugars have on the viability of Brine Shrimp. These findings present that trehalose is the most effective cryopreservation sugar for Brine Shrimp under temperature stress, demonstrating the potential future uses of this sugar under increased environmental stress. Trehalose and further investigations regarding the mechanisms of cryopreservation , as well as observation of organisms that withstand extreme environmental stress will increase the ability to conduct better experiments through deeper understanding. And, understanding the stabilization of cellular structures from this sugar will enhance the future knowledge of cryopreservation and its limits. Such insights could lead to future improvements in cryopreservation and its techniques possibly leading to the ability of application to broader subjects.

L1-BCMM Biology: Cellular, Molecular, or Microbiology
A Self-Regulating Gene Therapy Vector for Rett Syndrome
Yifan Ding

Rett syndrome is a devastating neurodevelopmental disorder caused by loss-of-function mutations in the MeCP2 gene. Gene therapy represents one of the most promising strategies for treating or potentially curing this genetic disease. However, a key challenge in developing an effective gene replacement therapy for Rett syndrome is achieving sufficient MeCP2 expression in affected cells without risking harmful overexpression . My study aims to engineer a gene self regulation system that maintains proper MeCP2 expression levels necessary for therapeutic benefit, while avoiding the detrimental effects associated with MeCP2 overexpression. Building on my previous research, I introduce a novel gene self-regulation system, “TREAD-Dimmer,” which leverages mechanisms of termination readthrough of the stop codon UGA and Cre-mediated recombination. In the TREAD-Dimmer system, high target gene expression triggers increased readthrough of a stop codon-containing sequence, leading to the translation of Cre recombinase fused downstream of the target gene. This results in Cre-mediated DNA recombination and self-deletion of the target gene, effectively downregulating its expression. Through cell culture experiments, I validated that TREAD-Dimmer can be applied to self-regulating gene therapy vectors, preventing overexpression of both green fluorescent protein and human MeCP2. Our ongoing collaborative efforts seek to further translate the TREAD-Dimmer system and its associated MeCP2 gene therapy vector into clinical applications. In conclusion, TREAD-Dimmer offers a promising solution to prevent overexpression in gene therapy , including MeCP2 gene replacement therapy for Rett syndrome, potentially improving therapeutic outcomes and minimizing risks associated with gene therapy treatment.

L2-BCMM Biology: Cellular, Molecular, or Microbiology
Optimizing T-Cell-Mediated Cancer Killing: An Agent-Based Model
Andrew Yu

T Cells are known to be potent killers of cancer cells, but the underlying mechanisms of exactly how the T cells mediate cell-killing, as well as how best to orient T cells in three-dimensional space are unclear. I developed an agent-based model to explore the behavior of T Cells and their killing of cancer when exposed to varying types of tumors by various cell growth and interaction parameters. Through this project, I hope to explore the influence of cytokines like IL-2 on T cell interactions and optimize the configurations of T cells to most efficiently kill cancer .

L3-BCMM Biology: Cellular, Molecular, or Microbiology
Fluorescent Bacteria
Taijah Beard

The purpose of my project is to know if the protein’s expression will affect the bacteria’s antibiotic resistance. I want to study this project because I want to become a Veterinarian, and I am interested in learning more about bacteria. Specifically, I want to determine what concentration of ampicillin would result in the most or least amount of bacterial death.

L4-BCMM Biology: Cellular, Molecular, or Microbiology
Acidic Foods’ Effect on Gut Microbiome
Angel Morales
Anh Nguyen

The focus of our project is to determine the effects of acidic foods on E coli as a representative of the gut microbiome . Using a handful of selected foods under a pH of 7, we analyze which foods promote E coli growth compared to others. The overall premise of the experiment is to determine how one’s diet may affect their gut microbiome, and by that their overall health.

L5-BCMM Biology: Cellular, Molecular, or Microbiology
Testing the Effects of Facial Cleansers Against E. Coli
Raina Ferreira

Lots of bacteria grow and thrive around the eyes. All sorts of makeup removers are advertised as being capable of fully removing all bacteria and product from the face. Testing to see how certain makeup removers affect E. Coli can show which ones are worth buying. In this test, a more ‘aggressive’ facial cleanser (CeraVe Salicylic Acid Cleanser) is being compared against the performance of a more ‘gentle’ facial cleanser (Garnier Micellar Water), and the performance of water which seems to have no damaging effects on the skin.

L6-BCMM Biology: Cellular, Molecular, or Microbiology
Efficacy of Antimicrobial Surfaces: A 16S rRNA Sequencing Comparison of Potential Pathogens in Hospital vs. School Water Sources
Chelsea Bateman
Olivia Chen

Water fountains have long served as essential providers of drinking water in public places , whether that be in school or in a shopping mall. Though companies often assert that these fountains are clean and equipped with antimicrobial features to inhibit bacteria growth, doubts still persist about the validity of such claims. This project investigates the purity of water fountains by swabbing and testing areas for bacteria growth. The presence of bacteria would suggest that these surfaces are not fully antimicrobial, raising concerns about the hygiene of this prominent water source. In addition, distinct environments harbor different bacteria. In order to explore a wider spectrum of bacteria presence, water fountains in both a school and a hospital were swabbed for comparison. We hypothesize that bacteria will not be present on any of the school or the hospital water fountains, as their surfaces have antimicrobial properties.

A10-BEPA Biology: Evolution, Plant and Animal Science
Do Orange Peels Improve Soil Water Retention Compared to Regular Soil?
Sumeyyah Laher

This project explores the use of orange peels to improve soil water retention as a sustainable alternative to synthetic polymers. By testing different concentrations of orange peels in the soil (0%, 5%, 10%, and 20%), results show that higher concentrations, especially 20%, significantly enhance water retention. This research highlights the potential of repurposing organic waste for sustainable agriculture in drought-prone areas.

N10-BEPA Biology: Evolution, Plant and Animal Science
A New Mechanism of Harmful Plastics: The Effect of G-15 on the Heart Rate of Daphnia magna Exposed to BPA or BPS
Sierra Kelch

Bisphenol A (BPA), a chemical used in hard plastics for food containers, is a xenoestrogen often replaced with Bisphenol S (BPS) to create “BPA-free” products (Seltenrich, 2015; Wagner, 2014). It has commonly been thought that BPA/BPS act through estrogen receptors α and β (ERα/β), (LaPlante et al., 2017); however, typical BPA exposure levels are much lower than levels needed to bind ER α/β (Hall & Filardo, 2023). Building off of my data from last year, which surprisingly found that BPA and BPS increase heart rate in Daphnia magna, this study aims to test my central hypothesis that BPA and BPS increase heart rate by binding a different receptor: G protein-coupled estrogen receptor (GPER). 

To test this hypothesis, Daphnia magna were exposed to BPA, BPS, G-15 (a specific GPER antagonist), BPA with G-15, and BPS with G-15. After six days, their heart rate was measured. I hypothesized that as the concentration of BPA or BPS increases, the heart rate of D. magna would also increase. I further hypothesized that as a GPER antagonist, G-15 would suppress the effects of BPA or BPS on D. magna heart rate, since GPER regulates calcium flow, and therefore heart rate (Luo & Liu, 2020). I also predicted that G-15 alone would decrease heart rate.

As hypothesized, BPA and BPS caused a similar increase in D. magna heart rate, suggesting BPS may be just as harmful as BPA. Additionally, the heart rate of D. magna exhibits a hyperbolic relationship with BPA or BPS concentration, suggesting that BPA or BPS saturate a receptor at those concentrations . I calculated EC50’s, which suggest BPA and BPS cause half the effect on heart rate at normal human exposure levels (Bao et al., 2020). More importantly, inclusion of G-15 successfully suppressed this effect, suggesting that GPER may be the receptor for BPA and BPS at these concentrations. Contrary to what I predicted, G-15 alone increased heart rate. This result can be explained by a prior study showing that the GPER agonist G1 decreases heart rate (Kardec et al., 2024). 

Overall, this study provides an alternative method for improving the safety of plastics: Instead of banning individual bisphenols like BPA or BPS, perhaps plastics could be screened for chemicals that bind GPER. This will prevent the continued use of BPA alternatives that are just as harmful as BPA. For future research, I propose genetically blocking GPER function using CRISPR or RNA interference to confirm if BPA and BPS affect heart rate by binding GPER. The system used in this study can also be used to test if BPA or BPS also affect reproduction through GPER.

N11-BEPA Biology: Evolution, Plant and Animal Science
Understanding the Immunomodulatory Effects of Echinacea purpurea on Caenorhabditis elegans
Emilia Graham

In the United States, approximately 8% of the population has been diagnosed with some form of an autoimmune disease. These diseases present with various symptoms; for many with these conditions, just prescription medications may not soothe all their symptoms. At the same time, social media encourages diet culture more than ever, and all sorts of fads are constantly promoting and romanticizing different diets and supplements . But could these supplements have a significant harmful impact on those with autoimmune conditions? Especially supplements that claim to be “immune boosting”. This project investigates the immunomodulatory effects of the immune boosting supplement echinacea on the immune system experiencing an autoimmune flare-up. Using a C. elegans model and an external bacterial pathogens. Marcescens to simulate an autoimmune inflammatory response in an invertebrate’s innate immune system , an echinacea treatment was implemented on this model and through a killing assay, the death rate of the nematodes over time was tracked. After the assay was concluded, it can be deduced that there was a significant increase in the death rate of C. elegans who were exposed to the echinacea following the stimulation of an inflammatory response. Therefore, those who have an autoimmune disease should avoid consuming supplements that claim to be “immune boosting”. This research encourages people to be more curious about the supplements that they consume and be skeptical when it comes to using natural remedies to soothe symptoms.

N12-BEPA Biology: Evolution, Plant and Animal Science
Creating Desiccation-Tolerant (DT) Koshihikari Rice Plants
Emma Toohey

Rice makes up a significant portion of the human diet with millions of hectares of it being cultivated across the globe. Consequently, it is imperative that its production rates remain steady. Unfortunately, climate change and droughts pose a threat to its cultivation, meaning that it is crucial to investigate methods to increase the desiccation-tolerance (DT) of rice plants. Rice is not the only crop adversely affected by droughts , however. Tobacco is another desiccation-sensitive plant, and one upon which a 2013 study performed by Sun et al. increased the salt tolerance of through the ectopic expression of the Arabidopsis glycosyltransferase gene UGT85A5 using the pBI121 vector. This study seeks to use ectopic expression to transfer UGT85A5 to Koshihikari rice, a lowland short-grain used for sushi. Prior to the creation of the transgenic rice plants, data on wild-type (WT) rice plants’ response to salt stress was obtained, UGT85A5 was PCR amplified and cloned into the pBI121 vector. It was found that the plants grown under less salt stress grew taller than those under higher salt stress. Given the success UGT85A5 had in increasing the salt tolerance of tobacco plants, the transgenic rice plants are expected to grow taller than their WT counterparts when treated with higher salt concentrations.

N6-BEPA Biology: Evolution, Plant and Animal Science
The Effects of CYP450 Knockout on APAP Toxicity in Zebrafish Embryos
Siana Solarazza

Enzymes catalyze biological reactions resulting in chemicals that can be toxic (Cooper, 2000). The metabolization of acetaminophen (APAP) is catalyzed by the enzyme cyp2y3 (Lane et al,. 2002). This experiment aimed to see if low doses of APAP affected zebrafish larvae movement, exploring the necessity of cyp2Y3 in the metabolism of APAP. Four treatments of zebrafish (Danio rerio) were created. Wild-type (WT APAP) zebrafish and cyp2y3 knockout (2y3 APAP) were dosed with APAP and Wild-type (WT control) and cyp2y3 knockout (2y3 control) were left as controls. The embryos were then placed in 10-minute dark and light periods for 50 minutes and data of distance traveled was then collected. It was hypothesized that the 2y3 APAP would move less than the WT APAP although both would move less than their control counterparts . Not supporting the hypothesis, the WT APAP moved less than the 2y3 APAP and the WT control. The 2y3 APAP had little statistical difference from its control. The products of APAP metabolization may create a chemical that assists in pain management in addition to a chemical that damages the liver and the rest of the organism.

N7-BEPA Biology: Evolution, Plant and Animal Science
Sustainable Carbon Capture: Increasing Food Security in a Changing World
Emily Ward

As of 2022, agriculture contributes 16-27% of all anthropogenic emissions annually, and while ammonia fertilizer is what enables the sustenance for the global population the emissions are unsustainable. One solution to maintain current agricultural output while reducing greenhouse gasses is to use fertilizer alternatives to conventional ammonia production which do not require the large amount of power to produce. Microalgal replacement of conventional fertilizer is a multifaceted solution, since growing microalgae directly removes carbon dioxide from the atmosphere, and utilizing species such as Chlorella vulgaris has the potential to replace some or all of the fertilizer in food crop production. Growing C. vulgaris resulted in a large and rapid decrease in CO2, from an ambient level to a six times decrease within two hours at a rate of 8.6 ppm CO2 /mg/min, and both alfalfa and radish plants receiving C. vulgaris as a nutrient source resulted in a significantly greater final height compared to those receiving no microalgae-extract nutrient treatment. Radish plants receiving 0.15 mg/ml resulted in a 223% percent increase in total height over the 0 mg/ml treatment and radishes in the 0.6 mg/ml treatment resulted in a 239% percent increase in growth over those in the 0.0 mg/ml. Microalgae fix approximately 183 tons of CO2 per 100 tons of dried microalgal biomass produced, which means that if an area slightly larger than the state of Texas was utilized for growing C. vulgaris, then annual emissions for the entire planet could be mitigated. Large scale microalgae farming is a step into a world of a financially and environmentally sustainable future, the only question left is; what are we waiting for?

N8-BEPA Biology: Evolution, Plant and Animal Science
Optimizing Vitamin C Production in Hydroponically-Grown Common Bean Seedlings (Phaseolus vulgaris) Through Controlled
Sunlight Exposure
Maggie Chen

Common bean seedlings (Phaseolus vulgaris) were grown at 4 different lighting conditions under a 2000 lumen plant grow light to determine the correlation between sunlight duration and vitamin C accumulation in plants in order to explore light as a biofortification strategy against reactive oxygen species. It was hypothesized that if common bean seedlings (Phaseolus vulgaris) were grown hydroponically under 12 hours of sunlight daily, then vitamin C synthesis was maximized. Bush bean seeds were inserted into the prepared rock wool growing mediums to germinate for 6 days on a 20-30ÅãC heating mat under a humidity dome. After the 6-day germination period, the common bean sprouts were transferred into mason jars with liquid plant food solution. Each Phaseolus vulgaris plant was placed under a 2000 lumen plant grow light for a duration of time based on their control or experimental group (e.g. 6 hours, 3 hours, 12 hours, or 24 hours) for a 7-day growth period, then seedlings were blended and filtered into plant extract and titrated with iodine solution for vitamin C content determination. Experimental Group 3, common bean seedlings under 24 hour sunlight conditions, were found to yield the highest vitamin C concentrations with 4.75 x 10-4M of vitamin C. Experimental Group 2 (12 hours of sunlight) produced the second-highest concentration of 2.47 x 10-4M. Control Group (6 hours of sunlight) was found to have a concentration of 2.36 x 10-4M, and Experimental Group 1 (3 hours of sunlight) obtained the lowest vitamin C concentration of 1.54 x 10-4M. The data indicated that a greater duration of light exposure to the Phaseolus vulgaris plant was found to have a correlation with vitamin C content and increased the antioxidant levels in the seedlings. The results suggest radiation on plants from a 2000 lumen plant grow light, specifically hours of daily light exposure, can be utilized as a biofortification strategy for ascorbic acid accumulation in Phaseolus vulgaris seedlings.

N9-BEPA Biology: Evolution, Plant and Animal Science
Analyzing the Effects of Different Lighting Types on Vanessa cardui Butterflies
Kaitlynn Goulette

Light pollution is an issue across the entire world and can potentially be very dangerous. The purpose of this science fair experiment is to better understand the physical side effects of light pollution . To do this, research was done on the characteristics of light and vertebrates. Painted lady butterflies were chosen for this experiment, divided into four groups of ten. Raised from larvae, the groups were exposed to different types of lighting – LED, fluorescent, incandescent, and natural light (control group). Every two days they were fed crushed bananas and as their life cycle progressed, their consumption rates were measured. After about a month of data collection, the data was analyzed. The group that was exposed to incandescent light did the best. In reference to the spectrum of the different light (LED, fluorescent, incandescent), the incandescent light had the most natural spectrum, which can be the reason as to why the butterflies thrived. The control group with the natural light didn’t do as well, which can be explained by my reliance of the sun instead of consistent light by bulbs. In the future, I want to further test what light butterflies prefer. Infrared red would be interesting as its spectrum is heavy in the most abundant wavelength of the incandescent light (which is what did the best). I would also like to make my control more reliable. There are light bulbs that mimic the spectrum of natural light completely and I would like to test them next year.

O10-BEPA Biology: Evolution, Plant and Animal Science
Microorganism Based Aquaculture Meal for Farmed Shrimp through the Use of Fermentation-Grown Marine Yeast and Fungi
AJ Shaw

Farmed carnivorous fish require a lot of fish meal and fish oil, which harms our ocean environment. Farmed species such as salmon require small pelagic fish to be turned into fish meal and fish oil to meet their needs for a high protein diet with omega-3 fatty acids. Even though farm-raised fish do not directly impact the ocean ecosystem, this requirement for wild-caught bait fish has resulted in over-fishing of these species. There is a need for a sustainable alternative to wild-caught bait fish that can satisfy aquacultural needs while maintaining healthy oceans. During this research, we were able to isolate and genetically identify various marine fungi and yeasts from five locations in the greater Boston harbor area that could be used alongside with other microorganisms as alternatives for fish meal and fish oil in an aquaculture feed. In our experiments, we prepared small test amounts of aquaculture feeds based on these microorganisms and determined Penaeus monodon ’s (tiger prawn’s) preference for the feeds. No significant preferences were seen between the yeast, fungi and a fish meal control, suggesting feasibility for their long-term incorporation into an aquaculture feed.

O1-BEPA Biology: Evolution, Plant and Animal Science
How Nutrient Outputs From Title 5 Septic Systems and IA Systems Compare Based On Their Affects to Pond Ecosystems on Cape Cod
Nina Hill

 

O2-BEPA Biology: Evolution, Plant and Animal Science
Evaluating the Effects of Noise Pollution on Cognitive Function in Drosophila Melanogaster
Carey Huang

Environmental noise pollution is linked to cognitive impairment, yet its sensory mechanisms are not fully understood. This study used Drosophila melanogaster to explore the effects of loud noise during feeding on cognitive performance , using food choice behavior as a proxy for decision-making. We hypothesized that wild-type flies would exhibit impaired cognitive function under noise, while deafened flies would remain unaffected. Flies were assigned to four groups: wild type and deafened, each tested under noise and no-noise conditions. Our results revealed that wild-type flies exposed to noise showed a significant shift in food choice distribution compared to silent conditions, suggesting impaired cognitive performance. In contrast, deafened flies did not display significant changes between conditions, underscoring the role of auditory input in mediating these effects . Although these findings support our hypothesis, further research is necessary to confirm the link between noise-induced sensory disruption and cognitive deficits, and to assess the broader implications for human cognition.

O3-BEPA Biology: Evolution, Plant and Animal Science
The Lead Factor in Fertility
Jeevitha Gnanavel

The projects tests the effects of lead in fruit fly egg production . It connects to the real world issue in which, recently, there were metals found in tampons and lead was found as the highest concentration in non-organic tampons. Based on research, lead negatively affects the reproductive system.

O4-BEPA Biology: Evolution, Plant and Animal Science
The Effect of B12 on Silkworms Fed a High Glucose Diet
Gigi Malvi
Lillian Graham

The project’s primary objective was to determine whether administering vitamin B-12 influences hemolymph glucose levels in Bombyx mori given a high-glucose diet. As Bombyx mori larvae advanced to the fourth instar stage, they were divided into our groups. Group 1 was given a control diet of just silkworm chow, Group 2 was fed silkworm chow with added B-12, Group 3 was fed silkworm chow with added glucose, and Group 4 received silkworm chow with both vitamin B-12 and glucose. Over eight days, the number of living larvae was counted in each group, the lengths of the living individuals were measured, and the OneTouch Lancing device was used to puncture the foot of the Bombyx mori, facilitating hemolymph extraction for glucose assessment. The findings indicated that vitamin B-12 diminished the level of hemolymph glucose in the Bombyx mori, and also increased the lifespan of the larvae. The larvae fed glucose along with B-12 exhibited approximately half the hemolymph glucose levels compared to those fed only added glucose. This experiment corroborated the hypothesis, demonstrating that the B-12 supplement effectively suppressed glucose levels in the hemolymph. In addition, although all of the larvae in the glucose-only group died by Day 2, all of the individuals in the other groups survived past the last day of the study. The addition of glucose and vitamin B12 did not seem to affect growth. The experiment’s outcome resulted in the possibility that B-12 could protect individuals from the effects of high levels of glucose .

O5-BEPA Biology: Evolution, Plant and Animal Science
Regeneration of Stem Cells in Planaria
Samhita Gowda

The purpose of this project was to study the effects of different concentrations of Aloe Vera, VitaminB12, Vitamin C and bFGF on the regeneration rate of Dugesia dorotocephala. Stem Cells are very important for organisms as they are what let them grow and heal. However a problem with this is that as humans grow, the amount of stem cells in their body decreases, causing there to be a more difficult time locating enough stem cells to harvest . It was hypothesized that these molecules, (Aloe Vera, VitaminB12, Vitamin C and bFGF) would create a more rapid regeneration rate. For each of these molecules there were three different concentrations, totaling in 13 different solutions. The four groups were aloe vera 10μg/mL, 50μg/mL, 100μg/mL, Vitamin C, 0.1%, 1%, and 2% and later 600μM, 700μM and 800μM,  Vitamin B12 0.1%, 1%, and 2%, and bFGF 10μg/mL, 50μg/mL, 100μg/mL, and a control. The organisms were observed over a course of two weeks. All the planaria were fed and kept under similar conditions, however only the planaria in the control group grew back to their original sizes, while all of the others were smaller than what they were full size. Additionally all planaria in Vitamin C solution died within or before the first week . Future experimentation could be conducted on more groups of planaria to get a better average, or on another organism with less stem cells to better replicate humans.  

O6-BEPA Biology: Evolution, Plant and Animal Science
Oh Snap! Venus Flytraps: A Look into Alternative Solutions to Protect Our Unique Plants
Anabelle Castoreno

My project, titled “Oh Snap, Venus Flytraps: A Look into Making Unique Organisms Resistant to Rising
Temperatures,” investigates the impact of rising global temperatures on Venus Flytraps , a species threatened by climate change. With projections indicating a significant increase in average temperatures by 2100, this research examines whether treatment with Azospirillum brasilense, a bacterium that may enhance root and chemical responses, can improve the resilience of Venus Flytraps under drought conditions . My experiment involves three experimental groups: a control group with normal watering, a group under drought without the bacterium, and a group under drought with the bacterium. Action potentials will be measured to analyze plant responses, aiming to develop strategies for preserving this unique organism in the face of climate challenges.

O7-BEPA Biology: Evolution, Plant and Animal Science
The Effect of Antioxidant in the Speed of Regeneration in Earthworms
Sharvi Sharma

The title of my project is The Effect of Antioxidant in the Speed of Regeneration in Earthworms . The purpose of this experiment was to find a simple substance found in human’s everyday diet to supplement existing drugs and medical procedures regarding regeneration. To test this 18 worms were used, six with antioxidants, six without antioxidants, and six control. The six with antioxidants and six without were cut in half after six days of adapting to their environment. The six with antioxidants were then given an antioxidant mixture. The control group was left untouched. They were all observed over a period of eight days to determine if Vitamin A had an effect on regeneration speed . After this trial, the results were inconclusive so a second trial is currently underway.

O8-BEPA Biology: Evolution, Plant and Animal Science
Smart Habitat Design and Implementation for Rabbits with Non-Invasive EEG
Rich Zhou

The project aims to design a custom rabbit cage that gives rabbits a bigger, more comfortable, more organized, and more interactive space to live in. A review of academic literature was conducted and determined that enrichment of a rabbit ‘s environment, especially with foraging opportunities, benefited rabbits. EEG data collection of a lab rabbit was used to gauge its reaction to different stimuli. The assumption was made that it is beneficial to organize a rabbit’s activity areas based on whether each activity excites or relaxes the rabbit. A rabbit backpack, consisting of a battery, sensors, and a Bluetooth unit, was built and put on the rabbit. The Bluetooth unit sends sensor data to the main control board, which uses a SIM card to InfluxDB, an online time series data storage platform. Gamma and beta waves are treated as signs of excitement, while alpha, delta, and theta waves are treated as signs of relaxation. The rabbit’s brain waves showed relaxation in the exposure of dimmed lights and soothing music, and it reacted with alertness to food and toys. Based on these observations, a three-story cage was designed, with the first floor being the “alert” floor with food, drinks and toys, the second floor a digging area that mixes exploration with ease, and the third floor a rest area with a small nest. The rabbit proved able and willing to use the stairs provided, and bodily wastes proved easy to dispose of. The project aims to take a step towards integrating pet owning into future smart homes and towards the welfare of pet rabbits in comparison with more common pets, such as cats.

O9-BEPA Biology: Evolution, Plant and Animal Science
Testing for Wolbachia in Insects
Jake Riendeau
Nathan Thornberry

My project is about testing for Wolbachia in different insects . In order to test for Wolbachia, we have to collect different types of insects, like beetles, bees, and other types of insects. We then have to run the samples through PCR to amplify the Wolbachia genes. This will allow us to run them through gel electrophoresis, where we can observe if Wolbachia is present. If Wolbachia is not present, we should not see anything on the gel. If Wolbachia is present, we should see a band. At the moment, we are planning on testing honeybees and yellow jackets. We will test more insects if we are able to find them. We expect to see Wolbachia present in some insects, but not others. It may be hard to get more insects, as it is the winter. Some insects in one location may have Wolbachia, while insects of the same species in another location may not have Wolbachia present. Once we have tested the insects, we will be entering them into a database that shows all Wolbachia data for insects collected and observed in a certain area. My project will extend off this and will hopefully be able to test for Wolbachia in bees around the state.

B1-CH Chemistry
The Shocking Effects of Different Liquids on Electrolysis
Nicholas Franco

Electrolysis is the chemical reaction where an electrical current separates the atoms of a molecule and releases them in
pure form. The most common form of electrolysis is that of water, where electricity is run through water to separate the
oxygen and hydrogen atoms and releases them as pure gasses. This occurs because the moving electrons from the
electrical current cause ions, or charged molecules, to form and react by separating from their molecules and exiting the
liquid at the anode and cathode, or the negative and positive entrances of the electrical current into the liquid. If I electrolyze different liquid electrolytes, then a different amount of hydrogen and oxygen will be produced because the conductivity and molecular characteristics and distinctions of different liquid electrolytes would affect the electrolysis process . I will test my hypothesis by running electricity through a container filled with different liquids and measure the amount of hydrogen gas released at a stainless steel nail by placing a test tube filled with water that will be displaced as the hydrogen bubbles rise to the top. After completing the experiment and analyzing the resulting data, my hypothesis was proven correct. I found that different liquids did produce more hydrogen gas than other liquids . I found that liquids with ionic bond substances electrolyzed the most efficiently, while those with covalent bond substances electrolyzed not as efficiently , which shows that electrolysis efficiency is based on efficiency of electrical conductivity since ionic bonds conduct electricity better than covalent bonds do, and the electrons can move faster to separate molecules faster.

B2-CH Chemistry
Capturing Carbon Utilizing Portable Units
Arjun Mohandass

The increasing concentration of carbon dioxide in the atmosphere is a significant contributor to climate change, causing major environmental issues. To mitigate this, portable carbon capture units offer a promising solution. These compact systems, inspired by submarine CO₂ scrubbers, can absorb CO₂ from the atmosphere without requiring enormous amounts of energy used by traditional carbon capture facilities. Unlike large-scale systems that rely on filters and complex compounds, these portable units utilize potassium hydroxide to chemically sequester CO₂ into potassium bicarbonate, lowering operating costs. This process is much simpler, requiring low electricity demands, which can be met by solar panels creating a more self-sufficient system. Furthermore, the potassium bicarbonate byproduct can be reutilized for various applications such as neutralizing soil and producing fire retardants. The potential of these portable units is significant, as they could be deployed in areas with high CO₂ concentrations, such as urban areas or near forest fires, effectively reducing atmospheric carbon and contributing to the mitigation of global warming.

B3-CH Chemistry
Electrolysis for Pollutant Removal in Simulated Wastewater
Rohan Amin

This project investigates the use of electrolysis as a method for removing organic pollutants from water. To simulate contaminated wastewater, a solution is prepared using food dyes, salts, and caffeine, which represent common pollutants. Electrolysis is conducted using electrodes and a sodium sulfate electrolyte to create oxidizing agents such as hydroxyl radicals and oxygen species. These agents help break down and degrade the simulated pollutants. To evaluate the effectiveness of the process, a spectrophotometer is used to measure the color intensity and absorption of the solution before and after electrolysis. This provides quantitative data on the reduction of pollutants in the water. The study explores how changes in experimental conditions such as electrolyte concentration, applied voltage, and treatment time affect the removal efficiency. This experiment reveals the potential of electrolysis as a sustainable method for water purification. The results could have practical applications in addressing environmental challenges, such as treating wastewater and reducing chemical contaminants in water supplies. By simulating real-world pollutants and using simple lab techniques, the project demonstrates how innovative solutions can contribute to cleaner water and a healthier environment.

Q1-CH Chemistry
Ionic Interactions in Ice
Ayaan Moriya
Faizaan Mohammed
Zayan Anwar

This experiment investigated the effects of different salts—sodium chloride (NaCl), magnesium chloride (MgCl₂), and a calcium chloride (CaCl₂) blend—on the melting rates of ice by measuring the amount of water produced over time. The independent variables were the different types of salt applied and the dependent variables were the model road’s mass change after the melted ice was removed from the system. A control group with no salt was used for comparison. Each salt was tested over a 30-minute time period, with the mass change recorded every five minutes. The results indicated that the CaCl₂ melted the most ice, followed by MgCl₂, NaCl, and the control (no salt). This suggests that CaCl₂ is the most effective in lowering the freezing point of ice. The experiment highlights the potential applications of salts in melting ice and the possible novel use of the optimal salts on roads.

Q2-CH Chemistry
Photonic Nanoglass Stress Sensors
Hayim Sims

Uniform nanoparticles can serve as building blocks to engineer designer materials with unique optical and electronic properties. However, engineering nanoparticles with precise sizes and shapes poses many challenges. This study investigates a novel synthetic route for liquid colloidal particles, utilizing metallic nanoparticles as seeds to control the growth of a liquid silicone phase. The seed serves to control the process of nucleation, which enables exquisite control over the size of the liquid droplets, while the liquid enables robust functionalization with polymer. Through fine control of the robust synthesis process, this study has generated a new family of colloidal particles ranging in size and composition, which assemble into crystals with photonic properties. Additionally, these crystals have been embedded into elastomeric hydrogels, which allow the crystals to respond to stress by changing color during deform. These materials have a tangible use as a sensor for biomedical and electronic applications.

Q4-CH Chemistry
Unlocking the Grain: Comparative Analysis of Glucose Release in Healthy Grain Varieties
Mahima Ganesan

Whole grains are increasingly recommended as healthier alternatives to refined grains due to their higher fiber content, better nutrient profile, and purported benefits for blood sugar management. However, many people struggle with properly preparing whole grains, potentially leading to reduced nutritional benefits. A key challenge is understanding how different preparation methods – particularly soaking and cooking durations – affect the way these grains release glucose when digested. While nutritionists universally recommend whole grains, there is surprisingly little systematic research on how preparation methods affect their glucose release patterns. Our study seeks to bridge this gap, with a comprehensive study across seven whole grain types (Oats, Millet, Quinoa, Farro, Brown Rice, Bulgur, and Barley), as well as comparing their glucose release to a low glycemic white rice variety. We tested these grains across four soaking times (0 hours, 4 hours, 8 hours, and 12 hours) and cooked them, testing the glucose release in five minute increments as well as taking the final cooked grain, while mixing with the enzyme amylase that is found in saliva. In total, our study involved 192 hours of cumulative soaking time, nearly 14 hours of total cooking observation, and generated 246 distinct glucose measurements across the eight grain varieties. Surprisingly, we found that rather than establishing a clear ranking of grains from best to worst in terms of glucose release, all the grains tested—including white rice—demonstrated a ‘universal convergence’ point of approximately 2% glucose release when soaked for 12 hours and fully cooked, regardless of their starting values. Our findings have significant implications for nutrition recommendations that can be provided to individuals based on their needs.

Key Words: Whole grains, Glucose Release, Soaking Duration, Cooking Time, Amylase Activity, Food Preparation Methods, Dietary Recommendations

Q5-CH Chemistry
Comparative Analysis of Sunscreen SPF Ratings and UV Protection
Abigail Jemiolo
Makenzie Watt

This project is the result of experimenting with and testing different sunscreens based on SPF ratings and analyzing the effectiveness of the UV protection. The goal of this was to find the most effective sunscreen. After trial and error with a few different kinds of paper we used “sun paper” to analyze the UV blockage and then recorded our data. When conducting this experiment we made sure to use proper materials like gloves to keep our experimentation as even as possible and UV protective eye wear to prevent serious damage to our eyes. We measured sunscreen effectiveness by the color of the sun paper after being under a UV light for 30 minutes checking them after every 5 and documenting changes. We concluded that surprisingly the SPF 100 was not the most effective sunscreen. The most effective sunscreen was the SPF 70. This sunscreen blocked the most UV rays for the longest time. While SPF 100 blocked the UV rays effectively at first, You have to reapply it sooner making the SPF 70 last longer. While we were successful in our experimentation we were surprised by the results. When testing sunscreen it seems like a fact that the higher SPF rated sunscreens would be more effective. We expected this to be the case when approaching experimentation however it was not. In the future we would love to take a further look at testing all the same brand of sunscreen with different SPF ratings or the same SPF rating and all different brands to further research what brand of sunscreen is the most effective.

Q6-CH Chemistry
Design of Next-Generation Dendrite-Free Electrolyte for Aqueous Zn-Metal Batteries
Ginny Ke

Aqueous zinc-ion batteries present a promising and safer alternative to lithium-ion batteries, leveraging zinc’s low cost, high abundance, and excellent stability in water. Despite their potential, challenges such as dendrite formation and zinc hydroxide side reactions hinder their performance and longevity. This research proposes the design of next-generation dendrite-free electrolytes by optimizing the water-in-salt ratio, specifically utilizing a composition of 1 M Zn(OTF)₂ and 20 M LiTFSI. This tailored electrolyte aims to minimize free water content, thereby mitigating dendrite growth and enhancing battery efficiency. Comparative studies with traditional KOH electrolytes indicate that the proposed ratio significantly improves cycling performance and Coulombic efficiency. The study employs comprehensive testing methodologies, including Raman spectroscopy, X-ray diffraction, electrochemical impedance spectroscopy, and linear sweep voltammetry, to analyze electrolyte composition and monitor dendrite formation. By refining the electrolyte composition, the research seeks to enhance the charge capacity, safety, and overall performance of zinc-ion batteries. The expected outcomes include identifying an optimal electrolyte formulation that effectively prevents dendrite growth, thereby advancing aqueous zinc-ion batteries as a viable, cost-effective, and environmentally friendly energy storage solution with superior safety profiles compared to lithium-ion counterparts.

Q7-CH Chemistry
Don’t Be Salty: Temperature-Responsive Coating for Controlled Release of Road Salt to Minimize Environmental Impact
Melissa Panaligan
Zayaan Mamun

Road salt is used as a common de-icing agent, but excessive application poses several environmental risks. This includes water contamination, soil degradation, and harm to aquatic life. This research investigates temperature-dependent coatings for road salt optimization, reducing unnecessary application, and as a result, minimizing ecological damage. Various coating materials—starch-based polymers, cellulose-based polymers, paraffin, and polyvinyl alcohol (PVA) with chitosan—were considered for their chemical properties and former applications as coatings, then synthesized and applied to salt grains. These coatings were tested for structural integrity under freeze-thaw cycles involving trials in the refrigerator and freezer, their effects on plant and aquatic life using wheat grass and water fleas, as well as dissolution rates in water, ice, and snow. Experimental analysis involved multiple trials and block testing to evaluate coating effectiveness in controlling salt release. The findings aim to improve de-icing efficiency while reducing the negative environmental impact of road salt.

Keywords: chemistry, environment, encapsulation, cross-linking, salinity, road salt, statistics, starch, cellulose, paraffin, chloride, optimization.

B10-CS Computer Science and Technology
CivicLens: AI Municipal Budget Chatbots for Enhancing Civic Engagement
Jerry Xu
Justin Wang

There are a growing number of AI applications. Still, none are explicitly tailored to help residents answer their questions about municipal budget, a topic most are interested in, but few have a solid comprehension. This research paper proposes GRASP, a custom AI chatbot framework for Generation with Retrieval and Action System for Prompts. GRASP provides more truthful and grounded responses to user budget queries than traditional information retrieval systems like general Large Language Models (LLMs) or web searches. These improvements come from the novel combination of a Retrieval-Augmented Generation (RAG) framework (“Generation with Retrieval”) and an agentic workflow (“Action System”), as well as prompt engineering techniques, the incorporation of municipal budget domain knowledge and collaboration with local town officials to ensure response truthfulness. During testing, we found that our GRASP chatbot provided precise and accurate responses for local municipal budget queries 78% of the time. In comparison, GPT-4o and Gemini were precise only 60% and 35% of the time, respectively. GRASP chatbots significantly reduce the time and effort needed for the general public to get an intuitive and correct understanding of their town’s budget, thus fostering more communal discourse, improving government transparency, and allowing citizens to make more informed decisions.

B11-CS Computer Science and Technology
Calculating Risk Percentage for Sjogren’s Syndrome Using Neural Networks
Jay Bhatia

Sjogren’s Syndrome is a immune system condition that attacks salivary glands and other organs within the body, affecting ~4 million Americans. In cases, Sjogren’s Syndrome can lead to detrimental complications such as organ failure, lymphomas, and neuropathy. The aim of this project is to utilize machine learning to create a risk percentage calculator for Sjogren’s Syndrome by utilizing existing patient and population demographic information. Specifically, a machine learning algorithm called neural networks will be utilized. Once trained with sufficient data, the network will be applicable to new cases, allowing any person to calculate a risk percentage for Sjogren’s Syndrome.

B12-CS Computer Science and Technology
Using an AI Equipped Drone for Protection of Keystone Species
Ansh Hiranandani
Yujiang Zhu

Keystone species maintain critical ecosystem functions, providing habitats for various animals, stabilizing soil, and mitigating natural disasters. In the Sonoran Desert, the Saguaro Cactus plays a vital role in preventing dust storms by anchoring soil and reducing erosion. However, illegal removal and habitat destruction threaten this iconic species, destabilizing the environment and increasing the risk of desertification and extreme weather events. Current methods of identifying and monitoring keystone species are limited and labor-intensive, highlighting the need for an efficient, scalable solution. An autonomous drone equipped with computer vision and machine learning can significantly speed up and improve the accuracy of Saguaro Cactus detection and monitoring, enabling timely intervention against illegal poaching and environmental threats that could lead to more frequent dust storms and ecosystem collapse. As the device is actively deployed in Arizona, it sets the new standard in conservation technology, promising substantial improvements in environmental protection and ecosystem management.

B5-CS Computer Science and Technology
TheraBee
Finnegan Sommer
Olutayo Oyewusi

The global mental health crisis highlights the urgent need for more accessible and accurate ways to diagnose mental health conditions, especially as demand increases and professional resources remain limited. This project aims to develop an intelligent support tool powered by advanced language technology that aligns with established mental health guidelines. This tool will provide clinicians, researchers, and individuals with a precise, easy-to-understand, and fair platform for mental health screening and personalized help. By incorporating sophisticated methods to explain how the tool makes its predictions, the prototype will demonstrate clear reasoning behind its assessments, thereby building trust among users and healthcare professionals. A strong fairness framework will be implemented to identify and reduce any biases related to different demographic groups, ensuring that the tool works equally well for everyone. From a technical perspective, the system will be designed to handle challenging inputs through thorough testing, scale efficiently to manage large amounts of data and protect sensitive information with secure data storage and strict access controls. The user interface will be created to be simple and inclusive, providing customized, evidence-based recommendations in formats that are easy to understand and apply. Through comprehensive testing, using metrics such as fairness across different groups , clarity of explanations, and stress testing, the project aims to reliably detect signs of mental health conditions, decrease the rates of missed diagnoses, and broaden access to quality mental health care. Ultimately, this solution advances the vision of using artificial intelligence responsibly to improve mental health management, encouraging forward-thinking and ongoing engagement while maintaining high ethical, clinical, and regulatory standards.

B6-CS Computer Science and Technology
AsthmaAssist: An Application that Better Detects and Monitors Asthma Management Through ML and Post-Diagnostic Tools
Neha Nagireddy

Asthma affects over 339 million people worldwide, yet many go undiagnosed due to the cost and inaccessibility of traditional tools like spirometers. Additionally, existing non-invasive methods often lack practicality, accessibility, or fail to support long-term management. AsthmaAssist is a point of care iOS application that bridges this gap by using a Neural Network model trained on audio features to classify and diagnose asthma from an audio file of the user coughing. The Neural Network model achieved an average validation accuracy of 87.77%, a test accuracy of 88.01%, a precision of 90.00, a recall of 85.00, and an AUC of 92.80. The model was then integrated into a SwiftUI-based iOS application, AsthmaAssist. This application also offers a simple, non-invasive way to assess asthma control for users to help enhance asthma management through the use of the ACT (Asthma Control Test) and Asthma Education. Future updates will use environmental and wearable data to predict asthma attacks, improving early intervention and long-term care.

B7-CS Computer Science and Technology
ADHD Brain Activity Based on Transformer and MAIA Analysis
Guanbo Wang

This research investigates ADHD diagnosis using advanced machine learning techniques , specifically focusing on Transformer models and MAIA analysis of brain activity. The study aims to address current limitations in ADHD diagnosis, including insufficient data volume, limited model interpretability, and inadequate temporal information utilization. The project proposes using Transformer’s attention mechanism to analyze fMRI data and combining it with MAIA for enhanced model interpretability. Key hypotheses include that Transformer-based models will more effectively capture spatiotemporal features in ADHD patients’ fMRI data compared to traditional CNNs, and that MAIA analysis will reveal correlations between model focus points and medically significant ADHD-related brain regions. The research also explores the possibility of identifying previously overlooked brain areas that may be significant for ADHD research .

B8-CS Computer Science and Technology
Personalized Workout Trainer: Hybrid AI and Classic Machine Learning Approach
Austin Lin
Jaden Chen
Saayan Rao

The popularity of physical exercises and workouts is surging over the past decades. However, a large portion of this increase is from a rush of younger and inexperienced people. With inadequate workout form, often combined with the drive for unrealistic standards, the effectiveness of exercises is reduced. Furthermore, people are susceptible to injuries without professional guidance. Trainers can provide these valuable and widely needed resources, but they are oftentimes not available or too costly for the general use and regular monitoring and guidance. 

To address these difficult challenges, our team presents an innovative and effective hybrid AI and machine learning driven workout application that utilizes the growing pose estimation technology to provide users with real-time feedback and performance insights. A Resnet-18 convolutional neural network with an accuracy of 96% for distinguishing between 6 different exercises was implemented. Machine learning is then used to process data from pose estimation to provide quantitative analytical calculations including form accuracy, consistency, repetitions and other metrics. As a virtual live trainer, the algorithm pipeline classifies workouts into 3 categories; good, caution, and danger, providing instant feedback. 

B9-CS Computer Science and Technology
Smart Traffic Lights with Reinforcement Learning
Dinesh Babu

The prototype constructed focuses on reducing traffic congestion and CO 2 emissions as a result of the pile-up of vehicles within intersections, specifically 4-way signalized ones. The main focus was intersections within the United States which used outdated signal technology that does not adapt to real- time traffic congestion. An engineering experiment was performed in order to program and train a reinforcement learning (RL) model to control traffic light signal timings dynamically and in real-time by minimizing wait time and CO2 emissions. The simulation of this traffic light control was done through the Simulation of Urban MObility (SUMO) software. The model was equipped with the ability to output a combined efficiency rating out of 100 that is composed of 50% weight in wait time and 50% weighting in CO2 emissions. Upon training the model nine times with a training period of 30 minutes each, the RL model was able to pass the success threshold of the efficiency rating output being 85/100. Overall, this model can be deployed onto a Raspberry Pi for a portable and cost- effective solution to modern-day traffic congestion.

Q10-CS Computer Science and Technology
Charting the Brain: Using Normative Modeling to Predict Adolescent Cognitive Development
Elaine Zhang

Understanding typical brain development and its relationship to cognitive abilities is crucial for developmental neurosciences. Here, I applied normative modeling to a large dataset of adolescent brain imaging (N > 10,000, age 9-15 years) to characterize developmental trajectories of brain structure, function, and white matter microstructure. Using these normative models, I quantified how individuals deviate from typical developmental trajectories and investigated whether these deviations predict cognitive abilities. I found that deviation scores from typical brain development significantly predicted both intelligence and learning ability, outperforming predictions based on raw brain measures. Multiple brain features contributed to prediction, including volumes of memory-related structures, functional connectivity between attention and cognitive control networks, and temporal lobe white matter microstructure. My findings demonstrate that normative modeling can capture meaningful individual differences in brain development and their relationship to cognitive abilities . This approach provides a framework for understanding typical brain development and potentially identifying clinically relevant developmental variations.

Q11-CS Computer Science and Technology
Using Molecular Docking and Machine Learning to Design and Discover Novel Drugs for ER- alpha Targeted Breast Cancer Treatment
Aishwaryalakshmi Saravanan

Breast cancer is the most common type of cancer in women in the United States, with around 245,000 women diagnosed each year. Over 80% of breast cancers are estrogen-receptor positive, meaning that the cells grow in response to estrogen, which is a hormone necessary for sexual and reproductive health. Estrogen carries out its role by binding to receptors such as estrogen receptor alpha. The ESR1 gene, which codes for that receptor to be made, will be expressed more for different people, meaning that more estrogen is able to bind to the receptor, leading to increased proliferation of breast tissue, which could develop into breast cancer. The best method to treat ER-positive breast cancer is through targeted therapy, which involves drugs such as Tamoxifen to specifically slow down the connection of estrogen to its receptor. While there is a search to increase the drug options available, the current drug discovery process costs over $2.8 billion and takes over a decade. In this project, I used machine learning to develop a machine learning model that predicts drug potency for targeting estrogen receptor alpha and treating breast cancer. Using the model, I discovered a  few novel drug candidate molecules with a higher pIC50 (efficacy value) compared to some of the currently used breast cancer medications. I tested my three discovered molecules from last year and modified these molecules using in-silico organic synthesis to add functional groups to specific receptor sites that are likely to improve hydrogen bonding. Using molecular docking and computational modeling to improve their efficacy, I simulated the binding of those molecules to  the ER-alpha protein to improve the binding affinity (efficacy) of the drug molecule. The original molecule that I discovered from the machine learning model had an affinity of -14.2 kcal/mol, but after rounds of modifications based on observations to improve hydrogen bonding, adding a carbonyl group, amine, and carboxyl group to three different receptor sites resulted in the best and improved binding affinity of -15.7 kcal/mol. The molecule also had favorable pharmacokinetic properties after testing in pkCSM. Ultimately, the goal of the project was achieved as I built a high-performing machine learning model to discover drug candidates, which, using computational modeling, I improved the efficacy of, resulting in four novel drug candidates for treating estrogen-receptor positive breast cancer.

Q12-CS Computer Science and Technology
LightningJolt: Hardware-Inspired Acceleration for Zero-Knowledge Virtual Machines
Celine Zhang
Eric Archerman

Keywords: Zero Knowledge Virtual Machines (zkVMs), Cryptographic Protocols, Computer Architecture, Ethereum, Instruction Set Architecture, Blockchain Optimization, Jolt, Offline Memory Checking, Rank-1 Constraint Systems

A major bottleneck in cryptocurrencies like Ethereum is their ability to verify large , complex computations. Zero Knowledge Virtual Machines (zkVMs) are a new cryptographic protocol that allow replacement of these complex computations with small, fast-to-verify proofs showing correct execution. In particular, zkVMs work by compiling programs into a list of verifiable assembly instructions ( also known as an Instruction Set Architecture or ISA) that operates on an emulated memory unit. Memory actions (reads and writes) are then proven using a multiset permutation check that breaks down into expensive polynomial commitment evaluations. To accelerate this proof generation, we develop LightningJolt on top of Jolt, the most performant zkVM to date. Our protocol mirrors physical CPU designs in a verifiable setting by integrating frequently accessed memory structures (registers) directly into the virtualized processor with a novel combination of Rank-1 Constraint System (R1CS) constraints and finite field arithmetic manipulations. The register state at each step is then passed directly as input to the backend prover for a time-optimal breakdown into computation commitments. Mathematical proofs show LightningJolt’s completeness, soundness, and zero-knowledge. We implement our changes in Rust and benchmark performance across standard smart contract equivalent example programs, realizing a 1.8-2.3x speedup in prover time.

Our hardware-inspired acceleration of Jolt represents a new approach to zkVM optimization, and we believe this unique perspective will continue to drive faster, more efficient blockchains.

Q8-CS Computer Science and Technology
Using Machine Learning to Identify Plant Diseases
Avi Amin

New plants are grown every day. If one plant contracts a disease, however, it could also spread to ones nearby. This problem is a dreadful experience for farmers and amateur plant owners. A promising solution to this crisis is computer vision. The hypothesis in this experiment was that if images of apple leaves containing diseases were fed into a convolutional neural network, then the neural network would yield at least 90 percent accuracy because it would learn the images’ key features and the different characteristics of each disease well . By efficiently realizing if a plant contains a disease, steps can be taken to prevent it from spreading further. Using Python and the machine learning software called Tensorflow, a convolutional neural network was designed to identify plant diseases. Throughout this project, 1,100 images containing diseases were split equally into 4 distinct classes: apple scab, black rot, cedar apple rust, and healthy. A neural network model was trained with 195 images per disease over 400 epochs and validated using 40 images per disease. The leftover 40 images in each class were used as unbiased data to test the model after training. This convolutional neural network was expected to have a low loss, not overfit, and high accuracy. It displayed computer vision’s power and correctly identified the classes 96.25 percent of the time on unbiased data. Furthermore, 4 of the 6 misidentified images came only from the apple scab class. 

Q9-CS Computer Science and Technology
Evaluation of Unsupervised Network Intrusion Detection Algorithms without a Gold Standard
Kevin Bai

Supervised machine learning algorithms have been widely used for network intrusion detection; however, they require labeled training data, which can be challenging to obtain accurately. It is worthwhile to develop a cost- effective network intrusion detection solution based on unsupervised machine learning algorithms that do not rely on labeled data. Assessing the performance of unsupervised learning algorithms without labeled data presents a significant challenge. In this paper, we employ the Expectation-Maximization (EM) approach to evaluate the performance of a set of unsupervised learning algorithms. To the best of our knowledge, this study is the first to utilize the EM approach for assessing unsupervised learning algorithms in the context of network intrusion detection. Compared to traditional evaluation methods based on labeled data, the EM approach provides reliable performance measurements, particularly when labels are imperfect. The combination of unsupervised machine learning algorithms with the EM approach for model performance evaluation offers a practical solution for network intrusion detection.

R10-CS Computer Science and Technology
Classification and Diagnosis of Helicobacter pylori Infections in Histopathological Images Using a Deep Learning Convolutional Neural Network: Phase One
Wilfredo Villanueva Erazo

The majority of the human population is host to a bacteria called Helicobacter pylori, a gram-negative bacterium species spread through contamination of food and fluids and most prevalent in third-world countries. The bacteria will live in a human’s body for their entire life if undetected and left untreated , during which it can descend into bacterial infection that will weaken one’s gastric system and leave it more susceptible to several gastric diseases including cancer . While there are already methods in place to detect the bacteria in a human’s system , such as specialists analyzing histology images to find traces or signs of the bacteria, these methods rely on human intuition and training which can sometimes be incorrect and lead to misdiagnosis as a result of human error. This paper reports on the results of a novel AI deep learning model programmed on Google Colab using Python and utilizing a histology image dataset of 33,165 images to train it in its ability to classify histologies on whether or not they contained H. pylori. Following four iterations, or major edits on the code, the model was able to classify given images with a validation accuracy of 89.7%. Based on these and other results, it was concluded that the model was able to accurately detect the bacteria in histologies. Along with this, this paper summarizes the data recorded from each iteration and explores essential concepts and methods to improve on the model for potential applied use in its continuation.

R11-CS Computer Science and Technology
Malaria Detection using Machine Learning
Gaetano Foster
Ritchy Samedy

Malaria is a disease that impacts millions of people every year. We created a machine learning model to detect malaria and solve this issue effectively. With the use of CNNS (convolutional neural networks) we were able to make this project possible. his is better than the traditional RDTs because of the accuracy and scalability . Having access to this tool will lead to early detection of malaria, which will positively impact economically struggling nations.

R12-CS Computer Science and Technology
Refining AI Models for Brain Tumor Classification: A Data-Driven Approach Using MRI Scans
Mrigank Dhingra

Brain tumors pose a significant diagnostic challenge due to their complexity and the limited availability of specialized neuroradiologists. Machine learning models offer a promising solution to augment clinical workflows and reduce diagnostic variability. This study evaluates the effectiveness of machine learning models in classifying brain tumor types and addressing challenges such as variability and time-intensive manual interpretation. Three key hypotheses were explored: (1) model performance varies numerically but lacks statistically significant differences , (2) classification accuracy differs by tumor type, and (3) specific image features disproportionately influence model predictions. Using the publicly available “Brain Tumor MRI Dataset” from Kaggle, comprising 7,023 MRI slices categorized into glioma, meningioma, pituitary, and non-tumorous cases, five classification models were assessed: Decision Tree, k-Nearest Neighbors (k-NN), DenseNet, ResNet50, and a custom Convolutional Neural Network (CNN). Preprocessing included resizing, grayscale conversion, and data augmentation. Model performance was evaluated using accuracy, F1 scores, and Area Under the Curve (AUC). DenseNet achieved the highest accuracy (93.61%), followed by k-NN (91.25%) and ResNet50 (91.12%). Key decision-making features included tumor boundaries and high-intensity MRI regions. Statistical analysis revealed no significant differences among top -performing models (p > 0.05). This research highlights machine learning models’ potential to improve diagnostic accuracy and reduce variability, emphasizing interpretability and robustness for clinical integration.

R1-CS Computer Science and Technology
Comparing Algorithims for Counting Friendship Relationships in Dynamic Graphs
Sloane Brzezinski

 

R2-CS Computer Science and Technology
ImmersiveBuild: Revolutionizing Construction Using VR Applications
Richard Li

With school renovations, upgrades to classroom layouts can help students learn more efficiently . However, the traditional requests for information from building occupants, such as students, staff, and faculty members, are based on 2D drawings or 3D models. Occupants may not understand blueprints or express their true feelings. They need to experience the environment and provide feedback about the renovation plans. This research combined Virtual Reality (VR) software and all-directional treadmill hardware to build an ImmersiveBuild platform for participants to sense the environment and evaluate the renovation plans. Our experiments show that the walk-through virtual environment influenced participants’ decisions , and participants’ physiological data records, such as the heartbeat rate, indicated their involvement and immersion. The ImmersiveBuild has a potential application for future architecture and construction industries in communicating with building occupants to collect feedback on renovation designs.

R3-CS Computer Science and Technology
A Machine Learning Approach to Improve the Optimization of the Traveling Salesman Problem and an Approach to Optimize Town Curbside Trash-Collection Truck Routes
Katherine Ni

The Traveling Salesman Problem (TSP) and its variants are important optimization problems with numerous practical applications in planning, logistics, and manufacturing, etc. However, solving them is computationally challenging as the required time can grow exponentially with the problem size. Inspired by road trip planning, where the sequence to visit cities is first determined, followed by route planning within each city, this project proposes a novel approach that integrates optimization and machine learning techniques to solve the TSP. The method first applies clustering techniques to group locations, then hierarchically decomposes the original problem into two types of smaller sub problems: the first determines the optimal sequence to visit clusters, while the other optimizes the routes within each cluster. The number of sub-problems is proved to increase only linearly with problem size. The new approach establishes a general framework that can integrate existing methods for the TSP to improve performance. Numerical testing results demonstrate significant performance improvement with the new approach. The second phase of the project focuses on an approach to optimize trash-collection truck routes in my hometown. Unlike the standard TSP, where each location is visited exactly once, road patterns may require multiple visits to form a feasible route. Additionally, U- turns are prohibited as they are difficult on roads. A novel model is proposed by strategically assigning locations to road intersections to significantly reduce problem size. Pseudo locations and associated constraints are introduced to form feasible routes and prevent U-turns. Testing results on a real town map demonstrate that the optimal truck route can be generated within one minute.

Key Words: Optimization, Traveling Salesman Problem, Machine Learning, Cluster, Trash-Collection Truck Route Optimization

R4-CS Computer Science and Technology
Applying Generative Artificial Intelligence to Social Skills Coaching for Neurodivergent Teens
Farhan Khan

Currently, many coaching and therapy fields suffer from a significant shortage of trained facilitators , which greatly limits the scope of such impactful coaching. One of these fields is social skills coaching for neurodivergent teens. Fortunately, Generative Artificial Intelligence, or GenAI, is improving in its ability to converse with humans. Consequently, numerous companies have capitalized on this ability. Thus, to help resolve the shortage of trained facilitators, this research focused on applying GenAI to fill part of the role of a trained facilitator. Specifically, this research explored how GenAI could deliver accurate responses to questions based on a set of teaching materials. The teaching materials are intended to guide the facilitator on how to deliver the social skills coaching. A dataset was created by parsing through the set of teaching materials: questions, context passages (from which to answer the questions), and answers were extracted and formatted into data points. Google’s Pegasus transformer was used to paraphrase each question 40 times, expanding the size of the dataset from ~250 to ~10,000 data points and producing a more diverse and robust dataset. Using this dataset, Google’s T5 transformer was fine-tuned to answer questions from the context passages, similar to how a teacher would answer questions from their students. A plethora of parameter configurations were tested to determine which configuration would be most optimal for accuracy. Overall, the model achieved a peak Evaluation F1 score of 97.52% and a peak EM score of 91.86%.

R5-CS Computer Science and Technology
Skeleton-based Assessment of Dyskinesia in Parkinson’s Disease Using Spatial Temporal Graph Convolutional Networks
Zuming Zhang

Parkinsonion gait, data preprocessing, graph convolutional network, pose estimation

Parkinson’s disease (PD) poses significant challenges to movement, with freezing of gait (FoG) being one of the most debilitating symptoms affecting patients ‘ mobility. In this study, we propose a multistage framework for PD symptom detection: subject segmentation in video, skeletal data extraction, noise filtering, and ST-GCN (Spatial Temporal Graph Convolutional Networks)/ST-GCN++ based FoG detection. We extracted the subjects in the video using an image segmentation model and generated their skeletal data using a pose extraction model to construct a high-quality dataset for FoG detection. The framework also takes privacy issues into account by analyzing only the lower half of the body in specific scenarios. Experimental validation demonstrates the pipeline’s effectiveness as a reliable , privacy aware solution from data preprocessing to symptom analysis. In addition, the results were collected and analyzed from  different dimensions and the proposed methodology showed promising results. This study lays the foundation for both improved diagnostic tools and practical real-world applications in healthcare.

R6-CS Computer Science and Technology
Using Contrastive Activation Addition to Combat Societal Biases in Language Models
Niranjan Nair

Large Language Models (LLMs) have recently seen increasing adoption in fields such as law, medicine, and recruiting, where decisions should be kept as unbiased as possible. Previous work has shown that these models’ responses reflect various societal biases based on race and gender. Contrastive Activation Addition (CAA) is a technique that has shown promise in changing the behavior of language models, and previous work has found it to be more effective than fine tuning, the traditional approach to altering model behavior, in various circumstances. CAA generates a steering vector that can be added to the activations of a layer during the feed-forward process, using less data than traditional fine tuning approaches. This project used CAA to reduce the effect of societal biases on the outputs of Llama -3, an LLM by Meta AI. It also observed neurons whose activations are correlated with societal biases, and that neuron activations tended to correlate to various biases at once. Biases are measured with a numerical benchmark before and after CAA is applied , and two-sample t-tests are used to see if CAA had a significant effect on bias benchmark scores . It was observed that CAA has a statistically significant reduction on bias benchmark scores for racial and gender-based biases. This work provides a valuable methodology for future researchers who are looking to investigate internal representations of biases in language models and for AI companies that aim to reduce the societal biases present in the responses of their premier models .

R7-CS Computer Science and Technology
CustomDancer: Customized Dance Recommendation by Text-Dance Retrieval
Xintong Wu

Customized dance plays a significant role in daily life as a unique form of cultural expression, emotional release, and social connection. It fosters creativity, physical health, and mental well-being, offering individuals a personal way to communicate, celebrate, and bond in alignment with their identity and experiences. As artificial intelligence continues to advance in multimodal data processing, Text-Dance Retrieval has emerged as a promising research topic with significant financial value and future potential, addressing the needs of various practical applications such as dance education, dance performance, robotic dancers. Specifically, In dance education, educators can input a specialized textual description into the model, which then generates recommendations to assist students in efficiently identifying relevant references , thereby enhancing their comprehension of specific movements and the emotional expressions conveyed through dance. In dance performance, Choreography is the foundation of dance performances. While finding the perfect dance to match specific scenes and emotions is time-consuming for humans, the retrieval system can quickly identify a dance that aligns with user prompts, even when it’s not explicitly tagged or mentioned. In robotic dancers, users can provide a customized text description to retrieve SMPL-based dances from the model, which can then drive the robot to perform 3D motion. Meanwhile, the research faces challenges, requiring not only alignment at macro level, such as mood and style, but also synchronization at micro level, such as fluidity or rhythm. Thus, retrieving customized dance based on user descriptions presents both research significance and challenge.

R8-CS Computer Science and Technology
Pulse-Fi: A Low Cost Robust Machine Learning System for Accurate Cardiopulmonary Monitoring Using Channel State
Information
Pranay Kocheta

Non-intrusive monitoring of vital signs has become increasingly important in a variety of healthcare settings. In this paper, we present Pulse-Fi, a novel low-cost system that uses Wi-Fi Channel State Information (CSI) and machine learning to accurately monitor heart and breathing rate. Pulse-Fi operates using commodity low-cost devices, making it more accessible and cost-effective. It uses a signal processing pipeline to process CSI data that is fed into a custom low compute Long Short-Term Memory (LSTM) neural network model. We evaluated Pulse-Fi using two datasets: one that we collected locally using ESP32 devices named ESP-CSI dataset and another containing recordings of 118 participants using the Raspberry Pi 4B called EHealth, making it the most comprehensive data set of its kind. Our results show that Pulse-Fi can effectively estimate heart and breathing rate from CSI signals with comparable or better accuracy than hardware with multiple antenna systems that can be expensive.

R9-CS Computer Science and Technology
Aegis Explorer: AI-Powered Real-Time Content Filtering for Child Safety
Chenhang Zhang

Abstract—With the growing risks for children online, this project introduces a Chrome extension that leverages AI driven text and image classification to filter harmful content in real time. The system employs BERT for text classification and YOLOv8 for image analysis, dynamically blocking inappropriate material while allowing safe content to pass through. Through experimental evaluation, we identified classification weaknesses, prompting the removal of the monetary and social categories due to persistent misclassification. Iterative dataset refinement and model retraining led to significant performance improvements. Performance testing revealed that GPU acceleration is essential for real time filtering, as CPU-based deployment resulted in substantial delays. Future work will focus on further dataset refinement, model optimization, and multimodal AI integration to enhance efficiency and accuracy . The results demonstrate the viability of AI-powered real-time filtering, offering a customizable and adaptive approach to online safety . This study lays the groundwork for future advancements in automated content moderation, contributing to a safer digital environment for children.

Keywords—online safety, child internet use, AI content filtering, parental controls, harmful content detection, web filtering, nonprofit technology.

S10-CS Computer Science and Technology
Comparative Analysis of Quantum Approaches for Solving the Electronic Structure Problem
Shreyan Mazumdar

In this project, we investigate the electronic structure problem—an NP-hard problem in computational chemistry. The electronic structure problem involves determining the distribution and energy of electrons in a molecular system, which is essential for understanding atomic-level material behavior. This study has wide-ranging practical applications in fields like material design, photovoltaics, nanotechnology, and drug discovery. Given the high complexity of electronic structure calculations, classical methods have limitations; however, recent advances in quantum computing offer promising solutions for this core problem in quantum simulation. Our primary goal is to accurately and efficiently estimate the ground state energy of a molecular system given its electronic Hamiltonian. The Variational Quantum Eigensolver (VQE) is currently the most promising algorithm for this task. Its low circuit depth, variational guarantees, and hybrid quantum-classical approach suggest it can provide quantum advantages while remaining feasible on Noisy Intermediate-Scale Quantum (NISQ) devices. In contrast to VQE, the Quantum Annealing Eigensolver (QAE) has been developed specifically for D-Wave’s adiabatic quantum computers. While QAE bypasses some of VQE’s challenges with ansatz construction and nonconvex optimization , it faces issues with QUBO scaling and accuracy. Another noteworthy approach, the Eigenvalue Estimation (EE) algorithm, leverages Quantum Phase Estimation and Quantum Amplitude Estimation to achieve a quadratic speedup over the best-known classical algorithms, making it an intriguing but less explored option. In this work, we implement VQE with a variety of optimizers, including a sequential triple-hybrid approach where quantum annealing optimizes VQE’s ansatz parameters. Additionally, we examine QAE, EE, and the classical Hartree-Fock method as a benchmark. We evaluate each algorithm’s speed and accuracy in estimating the ground state energy for a diverse set of molecules. Our analysis of the results reveals insights into how each quantum algorithm performs based on specific molecular properties and algorithmic assumptions. Finally, we present a practical guide for industry users, such as those in pharmaceuticals or chemical manufacturing, summarizing our findings to help them select the most suitable algorithm for determining a molecule’s electronic structure based on its properties.

S11-CS Computer Science and Technology
A Sustainable System Considering Trash Problem in Eleuthera Based on Linear Regression and Modified CTP
Jason Ko

In regions with large areas relative to waste treatment facilities like the island of Eleuthera, the concept of collective
waste gathering has emerged as a novel solution. While the island expected that the institutionalization of shared collection services would solve waste management issues and improve resident convenience, limitations were discovered during the commercialization phase. The key issues are that collection request frequencies are too low or collection costs are inefficient, particularly in areas far from treatment facilities. To address this, we propose a more efficient system in this paper. CTP (Collective Trash Platform) solves the inefficient collection route problem in widespread areas . By allowing users to move their waste to strategic collection points within reasonable distances, it ensures a 70% probability of route intersection within 1km, significantly reducing transportation costs and environmental impact. To ensure a higher probability of matching by adopting the Skyline query, the system recommends optimal collection options based on multiple factors including distance to treatment facilities, waste volume, and collection timing. The Group Skyline method is adopted to calculate optimal collection groups for areas sharing similar geographical and demographic characteristics. Finally, we adopted a why-not query to motivate collection drivers and suggest appropriate guidelines for improving performance in extensive service areas. Furthermore, I adopted Linear Regression in the system so that decision-makers can evaluate collection point placements and identify areas for model optimization, including factors such as waste volume patterns, geographical distribution of collection points, and transportation efficiency metrics. In the results section, adopting CTP shows improved matching rates without increasing time complexity compared to the original algorithm, particularly benefiting regions with dispersed populations. The paper concludes with suggestions regarding real-time queries and privacy protection aspects. Future research directions include integrating dynamic route optimization through real-time adjustments and advanced machine learning models beyond Linear Regression. Privacy- preserving data collection methods and IoT integration with smart bins and monitoring systems need exploration. Studies on scalability across different geographical contexts, cost-benefit analyses, and user behavior patterns will enhance system effectiveness . Cross-platform integration possibilities with existing waste management systems and smart city infrastructure present additional avenues for investigation.

S12-CS Computer Science and Technology
Inequality in the Age of AI: An Agent-Based, Machine Learning Approach to Explore AI and Economic Inequity
Iris Yu

Artificial intelligence’s (AI) dramatic rise in popularity and ubiquity in recent years has raised the question of AI’s impact on economic inequality. While AI-induced automation will displace a significant amount of workers, it also creates new opportunities. However, the exact relationship between AI and inequality still has no consensus ; some scholars believe it will exacerbate inequality while others argue it is the perfect tool to mitigate inequality. This study explores algorithmic bias, automation, and reskilling in AI and their double-edged impact on inequality, employing an agent-based modeling (ABM) approach combined with a machine learning approach. Three main models—automation, algorithmic bias, and reskilling—were tested and modified in a thorough modeling process. These models capture complex processes such as job displacement, job transitions, and workforce adaptation. Model validity was assessed using established economic patterns, including the Pareto Curve and Wage Curve, and inequality was measured using the Gini Coefficient and Palma Ratio . It was shown that while automation increases the Gini coefficient and unemployment rate , it decreases the Palma ratio. Additionally, reskilling is highly correlated with reduced inequality and a reduced the Gini coefficient . These key results point toward policy recommendations for governments and companies which can ensure a more equitable future for all. A promising avenue for future research would be to explore the aggregate effects of algorithmic bias , automation, and reskilling in one model.

S1-CS Computer Science and Technology
An Empirical Evaluation of Convergence to Correlated Equilibria: Introducing Multi-Stage Multiplicative-Weights Update
Ashley Yu
Michael Han

No-regret learning algorithms are an important component of advances in solving large-scale games. These algorithms
are commonly used to solve games such as Diplomacy, an AI benchmark with a large action space where agents compete
to dominate a map of Europe. We introduce Multi-Stage Multiplicative-Weights Update (MS-MWU), which shows an
improvement upon existing external-regret minimizing algorithms such as MWU across all our experiments. We also perform an empirical evaluation of classic no-regret algorithms such as Multiplicative-Weights Update (MWU) and Optimistic Multiplicative-Weights Update (OMWU). Furthermore, we test swap regret minimization algorithms such as the no swap-regret algorithm of Blum Ä& Mansour (2007) and the TreeSwap algorithm of Dagan et al (2024). We play these algorithms against each other and randomized adversaries on hundreds of subgames of Diplomacy along with Kuhn Poker and random games. Across all these games, our experiments show that MS-MWU converges significantly faster than MWU/OMWU. We experimentally show that swap regret and external regret remain very similar at all iterations . In other words, external regret minimization algorithms such as MWU outperform swap regret minimization algorithms such as BM in terms of rate of convergence and time complexity, even for very large time horizons.

S2-CS Computer Science and Technology
HM-Detect: Murmur Detection and Classification Methodology Using A Novel C2-LSTM Architecture for Multi-Modal Signals
Ram Sivaraman

Heart murmurs are abnormal sound signals generated by turbulent blood flow and are closely associated with specific heart disorders. Current methods to detect and qualify murmurs do not recognize that an important part of diagnosing a patient depends upon murmur location in the heart valves and timing between fundamental heart sounds. This research presents HM-Detect, a novel methodology to provide accurate heart murmur characteristics to cardiologists, specifically in telemedicine. In particular, can we use signal processing and machine learning techniques to detect and classify heart murmur characteristics (e.g. timing and location) from heart sound signals? A novel “multi-modal” long short-term memory (C2-LSTM) neural network-based architecture is proposed to combat signals with multiple frequency modes. This is a generalized approach for decomposing signals into its separate frequency modes. HM-Detect is a murmur detection and classification methodology using the C2-LSTM architecture. The methodology analyzes heart sound features, namely the filterbank energies and spectral sub-band centroids. A combination of under sampling technique on the heart sound data and the evaluation of the statistical moments of these features are used as input processing methods before feeding into a machine learning architecture. The machine learning architecture used is based on long short-term memory (LSTM) neural networks that incorporate the time-varying aspects of heart sounds. The HM-Detect methodology was validated on signals from the CirCor DigiScope dataset, which is a clinically verified dataset. Three different architectures with varying number of cells and layers were compared. The proposed methodology can achieve a performance accuracy of around 90%, with the best of the three architectures having an F1 score of 0.91 and a test accuracy of 87%.

S3-CS Computer Science and Technology
Authorship Verification of the Caesarian Corpus Using Siamese BERT
Brandon Li

The question of Caesarian authorship of De Bello Alexandrino, De Bello Africo, and De Bello Hispaniensi has puzzled
classical scholars for millennia. Although these three texts have traditionally been attributed to Caesar, stylistic differences between these texts and the rest of Caesar’s corpus have resulted in doubts as to their true author . Prior computational studies on the authorship of these texts have involved traditional stylometric feature-based analysis and methods from the field of distributional semantics. This paper builds upon these previous results by using a state-of-the- art Siamese BERT model to ascertain the authorship of these texts. Additionally, this study seeks to establish the versatility of the Siamese BERT architecture in conducting authorship analysis across multiple languages . Following training on the open-source dataset used by Vainio et al. (2019), the model achieved a 95.5% accuracy on the validation dataset. Finally, following authorship verification of the unknown text using the model, results indicated that De Bello Alexandrino and De Bello Africo could potentially have been written by Caesar, and that De Bello Hispaniensi definitively was not written by Caesar. These findings provide greater insight into a millennia-old mystery of authorship and contribute to historical knowledge about these texts.

S4-CS Computer Science and Technology
Nuclei Detection With Histopathological Images of Skin Tissue Through Machine Learning
Afya Shaikh

Early and accurate cancer detection is vital for effective treatment and improved patient outcomes . In particular, skin
cancers can be highly aggressive and spread quickly, making swift diagnoses crucial. Traditional histopathological methods for identifying cancerous cells can be time-consuming and prone to errors, adversely affecting patient care. This research aims to enhance the efficiency and accuracy of cancer detection through a machine -learning model that uses histopathological images. Skin cancers often involve abnormal cell nuclei that appear larger and darker due to excessive DNA. Detecting these changes early is challenging, as cancer cells may develop deep within tissues and standard diagnostic methods can be slow and error-prone. I created a machine-learning model to automate the detection of cancerous cells in histopathological images. The dataset includes 30 digitized Hematoxylin and Eosin (H&E)-stained frozen sections from ten different human organs, sourced from The Cancer Genome Atlas (TCGA). These sections cover a variety of tissue types and staining conditions, providing a comprehensive basis for training and validating the algorithm. The dataset includes over 8,000 annotated nuclei and various segmentation masks to support the development of the model. The machine learning model demonstrated significant improvements in detecting cancerous nuclei compared to traditional methods. By leveraging advanced segmentation techniques, the model increased the accuracy and speed of identifying cancerous cells in histopathological images. Automating nuclei detection with machine learning reduces the time required for analysis and minimizes the risk of errors in identifying these cancerous cells. This can lead to more precise surgical planning and better patient outcomes by facilitating early detection and accurate diagnosis. Applying machine learning to nuclei detection in histopathological images represents a promising advancement in cancer diagnostics. This model’s improved efficiency and accuracy can significantly enhance the detection and treatment planning for skin cancers , ultimately benefiting patient care and outcomes.

S5-CS Computer Science and Technology
Improving Hearing Loss Using Tactile Sound
Mokkshita Arun

A new way to solve the problem of hearing loss using engineering is the solution to help improve hearing loss for those
who have it using the concept of tactile sound. Hearing loss is a huge issue, as 1.5 billion people worldwide experience this, and it is only becoming worse over the years. Scientific concepts, engineering, and modifications with artificial intelligence can help generate a new, useful, and affordable solution for those who face hearing loss . Some solutions are currently available for hearing loss, but not all the technologies are easy to access or affordable . This project creates a solution for this, however, getting there may take a couple of years. The overall project is to create a wristband, a device that can be worn by people who experience hearing loss, so they can understand the volume of a room or the speech of a person using tactile sound. Furthermore, this wristband will convert audio inputs to vibrations that will allow people who wear this to understand the volume and duration of their auditory surroundings or a conversation they may have with a person. This prototype focuses on adding components such as the Arduino Sound Detection Sensor which can take in live audio inputs and somewhat effectively vibrate the intensity of volume and other characteristics . This prototype is the basis for creating a beneficial solution for people who experience hearing loss.

S9-CS Computer Science and Technology
Phones are Causing Loneliness. Can They Solve It, Too?
Fiona Steeves

When asked the leading predictor of death in the United States, your mind might jump to things like smoking, exercise, or diet. You may not know that social isolation is as good of or even a stronger predictor of mortality than traditional clinical risk factors like smoking. One way to combat social isolation, especially in people with social anxiety, is Virtual Reality, or VR. VR fosters deeper social connections than social media or phone calls alone, reducing social anxiety almost as much as meeting up with people in person, but with a fraction of the infection risk, time, cost, or social pressure. However, traditional VR methods are expensive, requiring hundreds of dollars and sometimes a high-end computer. But there’s another way to experience VR at home through Google Cardboard. By using your phone as the screen for the VR headset, the cost is reduced from hundreds of dollars to under ten. By combining social VR experiences with these cheap and accessible mobile VR experiences into ‘mobile social VR’, we could treat the loneliness epidemic! This project dives into the benefits of social VR, how these benefits could be transferred successfully to mobile VR for the first time, and the best way to treat loneliness in VR. 

Keywords: Virtual Reality, Loneliness, Social Isolation, Mobile VR, Social VR, Anxiety, Depression, Worry, Mental Health

F6-EES Earth and Environmental Science
Automatic Irrigation Using Soil and Weather Data
Noah Proctor

In the United States, many large-scale farms use some form of external irrigation, whether through drip irrigation or
overhead sprinklers. However, these methods can be highly inefficient and result in significant water waste . Because of this, various automated irrigation systems have been developed to minimize freshwater consumption in agriculture. Additionally, farms on hillsides can pose additional challenges, as water will flow downhill, causing the soil at the bottom to become moister compared to the soil at the top, further wasting water. While many systems incorporate some components, such as weather or soil data, they often fail to integrate both data types. This project introduces a new irrigation system that integrates weather and soil data while accounting for topographical variations by utilizing multiple solenoid control valves and capacitive soil moisture sensors. The multiple soil moisture sensors and solenoid control valves will allow us to smartly irrigate the different levels independently. This approach offers a more efficient and sustainable water management solution that shows a nearly 50% reduction in water used compared to traditional irrigation systems. This system could be transferred to large-scale agricultural operations and significantly reduce their water footprint.

F7-EES Earth and Environmental Science
Exploring Seaweed as a Natural Solution to Coastal Erosion
Caitlin Stimpson
Lasya Muthyala

This study examines the role of rockweed (Ascophyllum nodosum) in reducing coastal erosion, addressing
the growing need for sustainable shoreline protection. As climate change accelerates sea level rise and intensifies storms, natural strategies for mitigating erosion have become increasingly essential. This research combines environmental science, coastal engineering, and ecology to assess how the presence of rockweed affects sediment retention and erosion rates . A custom-built wave tank was designed to simulate coastal conditions, enabling controlled trials that compared erosion levels between sediment covered with rockweed and bare sediment. Sediment displacement was measured across multiple trials, with results indicating a significant reduction in erosion when rockweed was present. Statistical analysis, including paired t-tests, confirmed the effectiveness of rockweed in minimizing sediment loss . The erosion-reducing properties of rockweed are likely due to its natural biochemical composition and physical structures. Key compounds, such as alginates, fucoidans, and mucilage, contribute to sand stabilization by binding particles, retaining moisture, and improving adhesion. Additionally, rockweed’s holdfast structures trap sand particles and slow water movement , further mitigating erosion. This study underscores the potential of marine vegetation in erosion control, providing valuable insights for sustainable coastal management. The findings support the integration of natural barriers into erosion mitigation strategies, which contribute to climate resilience. Future research should explore the long- term effectiveness of rockweed in various environments, its interactions with other coastal species, and the potential for large-scale applications in shoreline protection.

W10-EES Earth and Environmental Science
Investigation of the Impact of Soil Microorganisms on the Growth of Vigna Radiata
Valerie Lindh

Previous research has explored the role of plant growth promoting microorganisms (PGPM) in the development of plants. This study quantifies the impact of PGPM in the role of the growth of Vigna Radiata through the comparison of the growth of V. Radiata samples in the presence and absence of PGPM. Individual sample growth was determined through the measurement of various metrics of plant growth with the mean value of each metric within each sample group being calculated and compared to that of the control group. The difference between means was expressed as a percentage of the mean of the control group in each metric. Statistical analysis determined that the presence of PGPM is attributed to 25% of cotyledon growth, 38% of root nodule growth, 19% of root depth development, and 20% of root length development, while the control group displayed superiority in the metric of height by only 1%. These findings imply that PGPM plays a notable role in the development of V. Radiata across various metrics and areas of plant development, indicating the importance of PGPM in industrial V. Radiata production.

W11-EES Earth and Environmental Science
Harnessing Food Waste-Derived Hydrogels for Targeted Herbicide Delivery: A Novel, Sustainable Approach to Controlling
Fallopia Japonica
Marko Mano

The state of Massachusetts continues to face a raging conflict with the rapidly growing Fallopia Japonica (common name: Japanese knotweed). Originally deriving its nativity from the Asian continent, Fallopia Japonica is an invasive plant species known for its rapid doubling time, as well as, the ability to deprive native vegetation of crucial sunlight, while simultaneously releasing chemicals that suppress the growth and germination of other native species. However, urban landscapes are tight-knit, and carcinogenic aerosols of commercial weed killers are carried by winds. Families treating invasive species in their gardens often place themselves and their community at severe health risk of lethal respiratory malignancies and lethal lymphomas by exposing their loved ones to cancer-inducing synthetics, such as glyphosate and acrylamide derivatives. The engineering goal of the project was to synthesize an optimal hydrogel mechanism that could 1) swell and release a herbicide underneath the soil without human interaction, 2) effectively disrupt the production of chlorophyll in photosystem II (PSII), as well as, 3) deplete chlorophyll levels to a SPAD index < 25.00 in the shortest duration of time from the initial herbicide treatment. The project utilized three different protocols to synthesize three types of highly hydrophilic hydrogels composed of biopolymers: chitin, cellulose, and alginate. Commercial herbicide was injected at a 500 mM concentration into each of the 45 hydrogel samples consisting of their unique biopolymer (chitin, cellulose, or alginate). There were three hydrogel samples synthesized for each type of hydrogel (nine total hydrogel samples of chitin, cellulose, and alginate) during each data set, and there were five total sets of data collected with their own Fallopia Japonica root treatment sites (45 total hydrogel samples and treatments). The depreciation of chlorophyll levels was accurately recorded for each hydrogel sample (45 total samples) over programmed intervals for four weeks with a soil plant analysis development (SPAD) meter, where a SPAD index below 25.00 indicated impairment of the leaf’s production of key D-1 quinone-binding proteins, and thus indicates the treatment time at which the hydrogel sufficiently inhibited the cellular function of the Fallopia Japonica species. The SPAD meter recorded data points over exact intervals for four weeks to monitor the rate of release of the herbicide treatment at the roots of the plants, while simultaneously recording the depletion of chlorophyll levels over the 30 days after the initial herbicide treatment. At the conclusion of the herbicide treatment, a core sample of the roots was collected with a soil auger, and plant growth was collected with an auxanometer. The use of digital imaging was used to quantify root systems, where high-definition imagery was quantitatively analyzed with RootLM and RootReader 2D, which quantify root growth responses from whole root systems or specific roots of interest. The root growth (length in centimeters) given by the auxanometer was input into GraphPad Prism to generate figures and perform statistical analysis on the data. The experimental trials determined that the Fallopia Japonica treated with hydrogels compromising of the biopolymer alginate as the cross-linking monomer possessed an average SPAD index below 25.00 (23.88) in the shortest amount of time (23.4 ± 0.2 days), the smallest average root growth (5.55 ± 0.05 cm) and the highest average root response decline of 0.25 ± 0.01, in comparison to the hydrogels comprised of the cellulose and alginate biopolymers. The experiment determined that alginate is the optimal biopolymer for the future synthesis of a hydrogel for targeted herbicide treatment at the root of invasive species.

Keywords: targeted treatment, biopolymer, SPAD index, hydrogel, Fallopia Japonica, herbicide, drug delivery, chitin, alginate, cellulose

W12-EES Earth and Environmental Science
Impact of Calcifying Macroalgae on the Mitigation of Ocean Acidification
Anna Li
Elmeria Cheung
Michelle Chen

 

W8-EES Earth and Environmental Science
Which Water Purification Method is the Most Effective?
William Sheehan

Purified drinking water is required for human survival as consuming contaminated water can cause severe illnesses.
Purification removes these contaminants. The purpose of this experiment is to determine the most effective purification
method an individual can use with limited resources. This experiment determines the effectiveness of the five most common purification methods; UV disinfection, distillation, reverse osmosis, microfiltration, and boiling. The hypothesis of the experiment is: if multiple water filtration methods are used on the same sample of lake water, then the method of reverse osmosis will match the purified drinking water standards most closely because it is the process used in large water treatment plants.

Water was gathered from Lake Pearl, Wrentham, Massachusetts. The water was split into six samples–five experimental and one control. Each sample was purified using a different method, and the samples were tested for the concentration levels of many chemical contaminants using test strips. Biological containment data was gathered using Petri dishes to culture bacteria in each sample, and the number of bacteria colonies was subsequently counted. The data was compared with the WHO drinking water standards to determine the most effective method . The experiment does not support the hypothesis and instead supports distillation being overall the most effective purification method. This method removed more biological and chemical contaminants than the other methods by far. Reverse osmosis ranked second as it was as effective at removing larger contaminants such as nitrate but failed to remove the smaller ones as effectively as distillation.

W9-EES Earth and Environmental Science
Plastic with a Purpose: The Science of Eco-Friendly Plastic Materials
Habeeba Fouda
Hana Abdelnaeem
Maya Ammer

Plastic pollution is a major environmental problem because traditional plastics take hundreds of years to break down.
This project explores eco-friendly plastic alternatives that will be closest to normal plastic using different polymers for each ones with the same plasticizer, glycerin. The plastics were tested for durability, texture, and water resistance to determine which formula creates the best alternative to conventional plastic. Our results help show how natural ingredients can be used to develop more sustainable plastics for everyday use.

X10-EES Earth and Environmental Science
Eco-Prosthetics
Roman Raducan

Habitat destruction is a significant factor in species decline, with bees being some of the most drastically affected species. When a beehive is damaged, the colony faces an incredible challenge with rebuilding or migrating to a new hive. This often results in severe population loss or total hive collapse. Eco-prosthetics aim to develop a temporary structural aid to preserve damaged hives, improving the chances of the colony to survive past the initial damage. The prototype is constructed using printer paper, honey, and beeswax to form a faux honeycomb structure. It is designed to be both heat-resistant (up to ~97–98°C) and structurally strong enough to endure the forces typically sustained by natural honeycombs. This will be tested by a heat test, where the comb will be put in a 100 degree oven for a period of time (1 hr roughly.) Then the “weight test” will be applying vertical pressure to the comb, and it must withstand comparably with regular beeswax or better. This faux comb represents a step in an artificial method of temporarily relieving stress on already strained ecosystems. This project represents the first step in many that could bring new ways to restore our damaged environment.

X11-EES Earth and Environmental Science
Effect of Post-Fire Smoke on Chlorophyll Concentration of Pines
Qinghe Zhao

 

X12-EES Earth and Environmental Science
Soil Fertility in East Boston
Elisen Bonilla
Joseph Puopolo

We will be examining and testing the quality of soil in different areas of East Boston.

X1-EES Earth and Environmental Science
How Does The Depth and The Cover-crop Treatment Affect Soil Carbon and Microbial Community
Xuanyu Chang

With the new focus on sustainability, the global increase in regenerative farming practice marks a new era of agriculture. Cover crops improve soil health by preventing erosion, enhancing nutrient retention, and increasing organic matter, which leads to better crop yields and sustainability. The purpose of this experiment was to determine if the presence of winter cover crops affects (1) soil pH and microbial activity and, (2) the properties of soil across different depths. Several observations were made for soil physical (soil color), chemical (soil pH), and biological (measured as cumulative soil respiration) characteristics. Three replicated soil cores representing the cover crop treatment (3-way mix) and two replicated soil cores representing no cover crop or control treatment were used. All cores were collected to a depth of 60 cm and data presented here is obtained from the top organic soil layer (0-30cm) and the deepest mineral soil layer (> 40 cm depth). Variations were observed in all measured properties when compared between the shallower (organic) and the deeper (mineral) depths. The soil replicates from a shallower depth had higher pH in all soil cores, irrespective of cover crop treatment. Soil respiration activity was highly varied from core-to-core, with a trend of lower activity under cover crop. However, these results are inconclusive due to lack of statistical tests caused by a small sample size. Future studies with increased replication would aid in confirming the observed trends reported here.

X2-EES Earth and Environmental Science
Analysis of Polymeric Microplastic Alternatives on the Preservation of Aquatic Macrophyte Photosynthetic Pathways
Chelsea Adams
Mia Caparrotta

Elodea is a genus of aquatic plants commonly found in freshwater environments, known for their role in producing oxygen through photosynthesis. These submerged plants, often referred to as “water weeds,” are vital to aquatic ecosystems as they produce oxygen, serve as a habitat and food source for various organisms, stabilize sediments, and contribute to nutrient cycling. However, the increasing presence of microplastics in aquatic environments poses a significant threat to these plants and the ecosystems they support. Microplastics, which are persistent pollutants less than 5 mm in size, can disrupt photosynthesis by blocking light penetration, inhibiting oxygen exchange, and releasing toxic substances into the water. Previous research has shown that microplastics can reduce oxygen production in aquatic plants by interfering with their photosynthetic processes, but studies examining alternative polymers are limited. This project investigates the short-term effects of microplastic pollution and its alternatives on the photosynthesis rates of Elodea . Three experimental tanks were set up to simulate varying levels of microplastic contamination: a control group with no microplastics, a low-concentration group, and a high-concentration group. In addition, alternative polymers derived from renewable resources were tested to assess their potential as less harmful substitutes. Photosynthesis rates were measured using a dissolved oxygen probe to assess dissolved oxygen levels, while plant growth was monitored visually. The results aim to provide insights into how both microplastics and their alternatives impact oxygen production in Elodea and similar aquatic plants, contributing to the understanding of their broader ecological implications. Through this study, we seek to predict the future risks of microplastic contamination in aquatic ecosystems and highlight the importance of developing more sustainable plastic alternatives.

X3-EES Earth and Environmental Science
Space Food: Growing Lactuca sativa in a Hybrid Hydroponic System with Lunar Regolith
Lily Swilling

Sustaining long-term space missions requires a reliable method of growing fresh produce. Current astronaut diets rely on freeze-dried and thermostabilized foods, which lack essential nutrients found in fresh vegetables. Hydroponic systems and regolith-based cultivation are two potential solutions; however, each presents challenges, including water retention issues in regolith and susceptibility to mold in hydroponics. This study explored a hybrid regolith-hydroponic system to evaluate its effectiveness in supporting Lactuca sativa (lettuce) growth. Lunar regolith was tested with five different hydroponic substrates—regolith , perlite, coco coir, peat moss, expanded clay stones, and rockwool—to determine which medium provided the most successful plant growth. Lettuce was cultivated in a Deep Water Culture (DWC) hydroponic system, with plants measured over a six-week period. Results were quantified by substrate efficiency, seed success rate, plant height, root length, and total harvest weight.

Results indicated that rockwool provided the most efficient growth, likely due to its moisture retention properties preventing oversaturation. Perlite, while hypothesized to enhance aeration, yielded moderate success, while clay pebbles demonstrated the lowest efficiency. Pure regolith exhibited high failure rates due to poor moisture regulation and particle instability. The study suggests that water-retentive substrates like rockwool, peat moss, and coco coir are ideal for space-based hydroponics. Future experimentation could be improved through a longer growth period and more data, as well as exploring sustainable in-space growth. 

X4-EES Earth and Environmental Science
The Effects of Environmental Concentrations of Caffeine on the Neurobehavioral and Neurobiological Development of Zebrafish
Embryos
Anika Jacob

Caffeine has been classified as a pollutant within freshwater systems due to its abundance in the environment. Caffeine enters the water primarily through wastewater excretion. Zebrafish are freshwater organisms that have potential exposure to lethal caffeine concentrations , as their natural habitats are found in polluted bodies of water. It was hypothesized that exposure to caffeine would negatively affect the neurobehavior and neurodevelopment of zebrafish embryos, as they progressed from embryo to larva. Embryos were divided into four groups containing no caffeine, 35μg/L, 40 μg/L, and 64 μg/Lml of caffeine. The concentrations for the groups were based on caffeine concentrations detected in water samples from Asia , Europe, and the Americas. Embryos underwent two behavioral studies, a touch tap test that measured changes in heartbeat, and a light-dark test, that measured reaction time. Neurobiological differences were determined through body length differences that were measured on Day 6 of growth using DanioScope software. A transgenic line of zebrafish embryos (5kbneurod:mito-mEos) was tagged to fluoresce in the brain, and morphological changes in brain structure were measured. There was a statistically significant difference in behavior for the four groups, it was observed that the resting heart rate was higher for the caffeinated groups than for the control group. There was a statistically significant difference in the morphology of the zebrafish embryos , and initial measurements of brain structure indicate differences in brain size between the four groups . These results support the hypothesis that caffeine exposure affects the neurobehavior and neurodevelopment of zebrafish embryos . 

X5-EES Earth and Environmental Science
Impact of Traffic Density and Precipitation on Conductivity in the Ipswich River (With Analysis Into Potential Effects on Local Organisms)
Helena Brain
Quinlan Kelly
Vanessa Steinmeyer

To investigate the environmental impact of applying road salt in the winter months on aquatic ecosystems which will carry the brunt of this pollutant due to run-off, we measured conductivity levels in three sites of differing traffic
densities. Using a conductivity probe allowed us to monitor the concentration of dissolved salts along the Ipswich River in our hometown of North Reading, Massachusetts, which lays entirely within the Ipswich River watershed. We tested throughout the winter to investigate a possible correlation between winter weather events, traffic density, and
conductivity readings. Data were taken daily from an upstream site (as a control, being far from major roads and near
a wetland which would filter out the water), in a high traffic density area, and a medium traffic density area. Each day, the measurement was taken at roughly the same time, accounting for possible conductivity variation that occurs over the day. One major assumption of our model is that the majority of the conductivity readings are coming from the salt applied to roads, and not from natural sources within these environments.

The purpose of this testing was to determine how proximity to high- traffic density roads as well as precipitation events might correlate to the conductivity levels of the Ipswich River. After the testing period, analysis into two variables, traffic density and precipitation, affect on conductivity readings revealed two main findings. The first finding indicates that there is no significant statistical relationship between mean conductivity and precipitation in the last 24 hours. The second finding shows that there is a very strong relationship between mean conductivity and sites of different traffic densities (high, medium, and low). These findings are significant because they indicate that salt application has the potential to significantly increase the conductivity of the river when close to a heavy traffic area , possibly endangering organisms that require a certain range of conductivity to survive and reproduce properly.

In a temperate wetland ecosystem such as the Ipswich River watershed, it is exceedingly important for these conductivity levels to remain within average limits as especially environmentally sensitive species such as a variety of amphibians (frogs and salamanders) and fish spend a significant portion, if not all, of their lives in contact with the water. Because of our findings into the negative association of conductivity readings close to higher traffic density sites, it may be beneficial for towns bordering the Ipswich River to consider decreasing the amount and frequency of road-salting during the winter, possibly considering other methods of increasing road traction to ensure winter driving safety. This is especially important to consider in areas where the road is very close to the river. Alternatively, these communities could consider planting natural buffers such as local trees, shrubs, and other vegetation to decrease the effect that applying rock salt to roads has on the balance of this river ecosystem.

Key words: Conductivity, Road salt, Run-off, Local species, Traffic density, Precipitation

X6-EES Earth and Environmental Science
Analyzing the Effects of Locational and Weather Changes on Air Quality
Aakash Stewart

Last year, when wildfire smoke clogged up my school, I asked the question: how do environmental factors affect air
quality in my town? After doing some research I created a hypothesis that both location and changes in weather affect the daily fluctuations of air quality, but the weather influences the quality of air more than location within the town. To test my hypothesis I started by compiling online data for the air quality, measured in AQI, temperature, humidity, wind speed, and precipitation data from the last three months. I then took the raw data values and converted them into a usable form by checking for outliers and errant sensors which I accounted for. With the data set ready for analysis, I started by checking for locational influence using a correlation chart which indicated a consistent pattern that as the distance between sensors decreased, the correlation between the AQI readings increased. This indicated a relationship between location and air quality. Then I looked at the weather conditions by graphing each condition separately against average AQI . From this I determined that temperature had no significant effect on air quality , humidity had a positive correlation with AQI, and both precipitation and wind speed had negative correlation with AQI. Finally, I concluded that weather conditions had a larger impact compared to locational influences, since the correlation chart indicated that no matter the distance between sensors, they shared a very high correlation, indicating that other factors, like weather, have a higher impact on air quality.

X7-EES Earth and Environmental Science
Using Filters to Limit Greenhouse Factory Emissions
Christopher Lemus

Greenhouse gas emissions from factories, particularly carbon dioxide (CO₂), contribute significantly to global warming and environmental damage. This experiment tested different air filters to determine their effectiveness in reducing CO₂ emissions from a simulated factory setting using burning candles as the emission source. Three types of filters; activated carbon, HEPA, and standard pre-filters, were evaluated based on their ability to capture greenhouse gases and particulates. A gas detector was used to measure CO₂ levels before and after filtration. The results showed that the activated carbon filter was the most effective at reducing CO₂ and other harmful gases, while the standard pre-filter had little impact. These findings suggest that improving factory air filtration systems using filters, similar to the black PureBurg Air Filter, could help reduce greenhouse gas emissions and slow climate change.

X8-EES Earth and Environmental Science
Best Soil for Sedum Growth on Green Roofs
Shevaun Brown

This project was done to help tackle increasing problems in our environment like urban heat effect , stormwater runoff,
and home insulation. The goal was to determine the best seed starter for sedums with the best ratio of drainage, lightweight composition, insulative properties and water retention. To determine this goal 4 tests were done to determine the weight of soil, insulation, cooling and drainage properties of 4 standard types of soil, using varying ratios of coco coir, sand, potting soil and perlite. These tests helped to determine that a soil mixture with roughly 3 parts Coco Coir to 3 parts Perlite to 2 parts Sand and 2 parts Coco Coir was the best soil mixture as it was the second best in each category. Soil mixture has met the design criteria for all aspects however it lacks in insulating which could be improved.

X9-EES Earth and Environmental Science
Determining the Crustal Structure of Central Antarctica Using Seismic Receiver Functions and Ambient Noise
Cross-Correlation
Richard Chen

The structure of the crust in Antarctica is still poorly understood due to its inaccessibility and high ice sheet coverage
(98%), despite its significance in global climate dynamics. To address this, I present initial results on the ice sheet
thickness, bed topography, and crustal thickness in the South Pole region determined from seismic data collected by two
linear arrays of seismic nodes laid between December 2023 and January 2024. A seismic profile along the 380 km-long
SW-NE leg was found by layering seismic receiver functions extracted from an earthquake event on December 20th, 2023 in Peru. The deconvolved P-wave and converted S-wave arrival times reveal the 2D cross-sectional ice sheet thickness to be between 1-4 km, aligning with bed topography predictions from the Bedmap2 model of the Antarctic ice sheet. The crustal root towards the southern Transantarctic Mountains is deeper than previously thought , suggesting they were formed through plate subduction or recent magma underplating. Ambient noise cross-correlation along the 170-km long SE leg reveals that the period-dependent phase and group velocities of surface waves vary between 3.0-3.5 km/s and 1.1-1.9 km/s respectively. RMA frequency normalization and robust stacking were found to best preserve the phase dispersion for mapping the shear-velocity structure of Antarctica. Overall, seismology is shown to be an effective tool in determining the geophysical landscape of central Antarctica. These results provide important insights on the structure of the Antarctic crust and ice sheet, helping to reveal their geological past and current evolution.

Y1-EES Earth and Environmental Science
Effects of Road Salting on Muddy River Salinity
Isabella Winey

In the past two decades, there have been dramatic increases in road salt usage across the US, especially as urban areas continue to grow and road use increases. However, there has not been sufficient research on how salting affects urban waterways considering the rate at which deicing agents are increasing in use. This project investigates an infamously polluted urban waterway, the Muddy River, which holds a key role in the urban ecosystem of Boston. Testing determined salinity over time after salting events throughout the winter, compared to a baseline level and outside data. Chlorine test kits were obtained from the Izaak Walton League of America’s Salt Watch program, and funded by the Muddy Water Initiative. Water samples were collected from eight locations at time intervals after snows , which correlated with deicing events. The testing sites were chosen based on proximity to major roadways with greater salt use or as controls at the beginning and end of the Muddy River. The data collected displays an increase in salinity after a salting event, followed by a dissipation that varied in duration depending on the location. At least to the extent of time that was sampled for, the water’s salinity did not return to the baseline level, indicating a long term issue that can be studied in more depth in the future.

Y2-EES Earth and Environmental Science
Novel Lichen Analysis at Lower Neponset River Reveals Previously Overlooked Heavy Metal Contamination
Yuxuan Zhang

Heavy metals are toxic pollutants that persist in ecosystems and bioaccumulate, causing chronic health issues such as
neurological damage and cardiovascular diseases. Many of these metals are byproducts of industrial activities. Since the
1630s, industries along the Lower Neponset River in Eastern Massachusetts have inevitably released heavy metals into the environment. Despite cleanup programs initiated since 2023, heavy metal contamination from centuries of industrial activity remains largely overlooked. As a solution, we leveraged Flavoparmelia caperata (L.) Hale, an epiphytic lichen species and cost-effective bioindicator, to assess heavy metal pollution at the site. We collected 88 natural lichen thalli and 25 soil samples from the Superfund site, along with 3 lichen controls from MA suburban areas as local baseline . Using Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES), we measured concentrations of 19 trace metals in lichen and soil samples. Lichen tissue exhibited significantly elevated levels of Al , Ba, Cr, Cu, Fe, Mg, Mn, Ni, Pb, V, Zn (p<0.01), with particularly severe Pb contamination. Enrichment factor analysis reveal that F. caperata is a more effective bioaccumulator for monitoring air quality and modern environmental contamination than soil . Spearman rank correlation analysis further suggest distinct distribution patterns within site and different pollution sources for lichen and soil contamination. We conclude that lichens serve as a modern air pollution indicator , while soils reflect long-term, localized contamination. Given that current remediation efforts focus on polychlorinated biphenyls (PCBs), we recommend a revision of strategies to address heavy metals and its associated health risks.

Y3-EES Earth and Environmental Science
Eastie Farm and its Carbon Footprint / Eastie Farm y su Huella de Carbono
Elijah Tavares
Mariana Monsalve Medina

Fresh fruits and vegetables are suppliers of important vitamins and minerals that the body requires. They are essential
components for having a healthy diet, but they are not always available or affordable. Food deserts are areas where it is
difficult to find and buy fresh food and are typically found where low -income people live. East Boston is considered a food desert, due to its few food distributors–the only one being Shaw’s. To increase access to fresh food, Eastie Farm offers a “CSA” which is a weekly subscription to fresh, local produce that supports farmers, their health, and the Massachusetts community. The “CSA” and growing food locally provides an additional benefit. Purchasing food from local sources, like Eastie Farm, would reduce the ecological impact, specifically the large carbon emissions of industrial agriculture and large scale food distribution.

C10-EB Engineering: Biomedical
Impact of Telehealth Utilization on Management of Chronic Hypertension and Health Outcomes: A Comparative Study
Patrick Xu

Background: The COVID-19 pandemic significantly disrupted in-person healthcare, accelerating the adoption of telehealth. While telehealth has become an integral component of healthcare delivery, its impact on chronic disease management, particularly hypertension, remains underexplored completely.

Methods: This retrospective cohort study analyzed hypertensive patients from the Montefiore Health System between
January 2020 and August 2024. Patients were stratified into two groups: those with only in-person outpatient visits and those utilizing both outpatient and telehealth visits. A 1:1 propensity-score matching controlled for demographic and comorbidity differences. Key outcomes included blood pressure (BP) control, healthcare utilization, and mortality.

Results: Patients who engaged in both telehealth and outpatient visits had significantly better BP control, with lower mean systolic BP (128.34 Å} 13.25 mmHg vs. 134.15 Å} 19.07 mmHg, p=0.002) and lower rates of uncontrolled hypertension (SBP >140 mmHg: 16.8% vs. 36.6%, p<0.001). Antihypertensive prescription rates were higher in the telehealth group (30.9% vs. 22.7%, p<0.001). Although ED visits were more frequent in the telehealth group (17.6% vs. 13.7%, p=0.002), all-time mortality was lower in the telehealth group (5.8% vs. 7.7%, p=0.04). Sensitivity analyses confirmed similar trends in later years.

Conclusion:
Telehealth utilization was associated with improved BP control and lower mortality among hypertensive patients . These
findings support telehealth as a viable strategy for chronic disease management, though its impact on healthcare utilization warrants further investigation.

C11-EB Engineering: Biomedical
Automatic Pill Dispenser Powered By Image Recognition To Ensure Medication Consumption
Sophie Barriault

Many people in their day to day lives rely on different types of medication , and yet according to the national library of
medicine, 50% of prescribed patients don’t take their medication. Why is this? Sometimes, it can just simply be because
it’s hard for individuals to remember to take it. Other times however, people simply choose not to take their medication,
despite the encouragement from their caretakers and doctors. This can be for a variety of reasons, from not understanding the importance their medication serves to having concerns relating to the side effects of their medication to having struggles such as depression that prevent them from taking care of themselves. Whatever the case may be, taking these medications are extremely important for the wellbeing of these individuals, and when they don’t take their needed treatments, their caretakers, family, and doctors become worried for their health, but often times can’t do much about it as they’re not able to be by the sides of these individuals 24/7 to ensure that they’re staying healthy. The caretakers of these individuals may also be extremely busy in their lives and are not always able to come by and update any necessary information pertaining to their loved ones medication intake needs, especially if they often travel abroad. My project directly combats this issue by creating an automatic medication dispenser. This dispenser not only reminds people to take their medication, but also utilizes an Optical Character Recognition model and a Named Entity Recognition model to automatically upload necessary information about the dispensed medication through a photo of the prescription label. This dispenser also uses an image recognition model to ensure that the patient actually takes their medication.

Key words: Medical apparatus, bioengineering, biomedical engineering, AI, machine learning, medicine, engineering, computer science, image recognition, software engineering

C1-EB Engineering: Biomedical
Empowering Low-Resource Settings with RetinAI: An AI System with Wearable Headset and Retinal Imaging for Eye Tumor Home-Screening
Ethan Yan

Early diagnosis of retinoblastoma (RB) in the current clinical practice remains challenging due to the lack of access to
timely and accurate eye examinations which is essential as late stages of RB often lead to enucleation and blindness. To
address this issue, this research develops RetinAI, the first low-cost wearable headset and retinal camera with artificial
intelligence systems for early detection of RB. RetinAI consists of two components: an extraocular detection device coupled with a YOLOv11 deep learning model to detect the early sign of RB, leukocoria, and an internal retinal image system powered by ResNet-50 or YOLOv11 deep learning models to detect retinal tumors. The hardware includes a 3D printed headset and lens/camera connector, Pi Camera 3, infrared/white-LED light, LCD screen, 20D lens, and Raspberry Pi 5. For leukocoria detection, computer vision color analysis of leukocoria and normal pupils showed they had different Value , Hue, and Saturation clusters. Then YOLOv11 model was developed and showed a high performance of 98% mAP, and can detect leukocoria as small as 1 mm in diameter using the RetinAI headset. For RB retinal tumor detection, ResNet50 transfer learning model and YOLOv11 were developed and achieved 97% accuracy and 96% mAP respectively. Clinically, using RetinAI on normal and retinoblastoma patients demonstrated that RetinAI successfully captured clear retinal images and detected the normal retina and retinoblastoma tumor. RetinAI is the first low-cost system that can not only detect eye tumors from pupil images but also capture clear intraocular retina images without any pupil dilation medication and detect tumors using high-performance deep learning systems. RetinAI can significantly improve RB early detection and can be used at home or in small clinics without eye specialists, especially in underserved communities.

C3-EB Engineering: Biomedical
Sweat Lactate Detection For Early Diagnosis Of Decompensated Heart Failure
Srivibhu Piratla

This year’s findings underscore some of the most alarming trends seen in heart failure in recent years . While HF rates
have steadily increased over the last decade, the 2024 report shows that the problem is growing even more severe,
particularly in younger populations, racial and ethnic minority groups, and those with multiple health conditions. This project identifies decompensated heart failure, in a cost effective, non invasive, and user friendly way by using machine learning powered data collection.

C4-EB Engineering: Biomedical
NeuroVox: Voice-Activated Spinal Injury Muscle Regeneration
Komal Jasuja

Background: Spinal cord injury (SCI) is often accompanied by severe mobility limitation with many patients becoming wheelchair bound. Depending on the level of the injury, survivors may struggle with movement in the legs and arms. Physical therapy in conjunction with functional electrical stimulation (FES) is used for rehabilitation in denervation injuries. Attendance in therapy sessions for wheelchair-bound patients is challenging and compliance is low.

Mobility limitation leads to disuse atrophy, which results in severe adverse physical, metabolic, and psychological disease states. In addition to visible detrimental effects, being dependent on others for exercise/tasks
takes an emotional toll – frustration, helplessness, and depression, collectively leading to a decline in overall quality of
life. Accordingly, the aim of this project was to develop a low-cost, portable system, which empowers SCI patients in
taking charge of their own muscle rehabilitation.

Hypothesis: I posit that the device will enable a paraplegic patient to independently strengthen their muscles through personalized stimulation. Particularly, the voice command feature will allow for easy accessibility to regaining a positive sense of control of their treatment, restoring self-esteem, thereby improving general well-being.

Overall Design Requirement: To enable SCI patients in taking charge of their own muscle rehabilitation through: a) Voice-activated programming of muscle stimulation routines; b) delivery of stimulation sequences to selective muscles and c) providing real-time assessment of the dynamics of exercised muscle during the rehabilitative interventions.

Execution: The integrated framework, NeuroVox, utilizes an ESP32 microcontroller, voice-recognition module, inertial measurement unit (IMU), configurable SSR-10DD relay, and programmable TENS 3000 units. The device uses adjustable stimulation settings, real-time electromyography sensor feedback, and data logging to track patient progress. Python scripts were used to implement libraries for speech recognition in the graphical user interface. The ESP32 microcontroller consists of Arduino IDE that is programmed to trigger TENS units selectively and to send notifications to the loved -ones or the caregiver, updating them of the successful completion of the exercise routines.=

Conclusions: I describe the development and successful implementation of a scalable therapeutic framework that enables voice-controlled activation of muscle stimulation routine by the SCI patient’s paralyzed muscle groups and integration of additional sensors . Using a voice-activated FES system will reduce the need for caregiver assistance, increase exercise compliance, and promote a greater sense of social independence in individuals with SCI.

C5-EB Engineering: Biomedical
NeuroLens: Multimodal Deep Learning Application for Effective Multi-Class Classification and Semantic Segmentation of Invasive Brain Tumors
Abhiraam Venigalla

According to the American Brain Tumor Association, around 90,000 people get diagnosed with brain tumors every year. Brain tumors are masses of abnormal cell growth in the brain caused by DNA mutations or other genetic factors. There are two types of tumors: non-cancerous and malignant. This study focuses on categorizing and locating malignant brain tumors like gliomas, pituitary tumors, and meningiomas. In the clinical world, quick and accurate diagnoses are essential, and manual classification can be prone to errors and very time-consuming. Current deep-learning models tend to use large amounts of computational power, yet yield average results. For this reason, NeuroLens was created. NeuroLens provides multimodal features that no other app has. It can accurately detect, classify, and segment brain tumors from MRI scans within seconds, compared to larger models that take more time. An AI chatbot called NeuroLens Assistant is also implemented for further diagnosis and prognosis of results given by NeuroLens. NeuroLens first uses special image preprocessing techniques like ADF (Anisotropic Diffusion Filtering) to increase the sharpness of images and class weights that guide the model’s attention to minority classes . Additionally, it uses techniques like Gaussian Blur and Histogram Equalization in parallel. A custom binary mask is then employed for further refining of important features, increasing the analytical preciseness and timely diagnosis of invasive tumors. The blurriness of the image was then measured using a constant called the Laplacian variance. Images with an ideal blur score of around 250-300 are essential when fed to the classification model. After image pre-processing, scaffolds of models were created and trained on their respective datasets. These scaffolds were then built upon either by adding more layers or regularization techniques like batch normalization and dropout layers to prevent overfitting and preserve the learning of the model until they presented high-performance metrics. Additionally, another problem that is faced is the ambiguity of MRI scans. It may be unclear to the surgeon where the tumor is for excision based solely on the scan. That is where a highly sophisticated segmentation model can be most useful. This app uses an MRI scan and mask dataset from multiple sources on the Kaggle database. The segmentation model in this app was a custom, robust UNet with an encoder, bottleneck, and decoder pathway. This segmentation model was then trained, validated, and iterated upon extensively. NeuroLens presented with extraordinary performance metrics. The train accuracy for the binary detection model was 99%, while the test accuracy was 96%. The train, test, and validation accuracies for the classification model were 99.2%, 96.42%, and 95.77% respectively. NeuroLens’s segmentation model achieved an IoU (intersection over union) ranging from 0.71-0.91, close to industry-leading segmentation models with only a fraction of their parameters. These results underscore the potential of NeuroLens as a ground-breaking tool to deliver high-precision detection, classification, and segmentation of brain tumors in a clinically relevant environment, demonstrating its vitality in real-world medical applications.

C6-EB Engineering: Biomedical
Designing an Economical Multi-Purpose Care Blanket for Neonates Part III
Emily Bulan
Luke Bulan

Designing an Economical Multi-Purpose Care Blanket for Neonates Part III Luke E. Bulan, Emily A. Bulan, Westfield High School, Westfield, MA This project is a result of research conducted on premature babies and their health risks, especially those involving temperature regulation and the condition Neonatal Abstinence Syndrome , or NAS. These conditions called for a blanket to be made that would aim to incorporate many different aspects of neonatal care simultaneously while a baby is swaddled, at full function and at a low cost. A vibration component, a heating pad, and blue LED lights were sewn into a swaddle blanket. The irradiance of the lights would be measured once using a bilimeter to ensure the irradiance was sufficient for high intensity phototherapy . The heating pad would be plugged into a temperature controller, which would be set to keep the temperature in the blanket within a range of 35-38ÅãC. Trials would be run tracking the progressive temperature of the blanket, and any deviations outside the desired range. The strength of the vibration component would be measured using an accelerometer to ensure the vibration could be felt all throughout the baby’s body. In a 4-hour trial, the blanket strayed above the maximum temperature bound several times, in intervals of about 20 minutes. Despite the deviations, the controller kept the relative temperature very consistent. The swaddling and phototherapy components were both functional, along with the added vibration component being able to be felt throughout the entire model. Even though the temperature controlling function was not flawless, its consistency allowed it to be deemed a success, as with reasonable refinement and more sophisticated material, a fully functional controller could be achieved. As this model is further developed and optimized, the result could be a great societal benefit. A safe, cost-effective, comfortable swaddle blanket that contains fully-functioning treatments for a variety of neonatal conditions could make basic neonatal care accessible in the home, relieving an immense load from NICUs around the world, as space could be cleared up not only from fewer patients, but also smaller equipment. The more accessible technology would also be able to spread throughout the world, even to places where current expensive  are technology is not accessible currently.

C7-EB Engineering: Biomedical
Smartbell: Chip-enabled and AI-driven Dumbbells for Optimized Upper-Body Strength Training
Tim Wang

This project hopes to create a pair of weighted handheld devices for upper body strength training, especially in the elderly community or other individuals with lower limb immobility. The devices will incorporate sensors, microcontrollers, and real-time feedback mechanisms to enhance workout engagement and efficiency . The device will be equipped with sensors to track its orientation and receive user feedback, transmitting data wirelessly to a computer system. A camera will monitor the user’s exercise, identifying and tracking the device in real-time. A computer program will display ‘targets’ on a monitor, guiding the user through a series of pre-designed upper body exercises by directing the device to specific positions. Each successful movement toward a target will reward the user with points and trigger musical notes that contribute to a song. By incorporating live monitoring and game-like elements, the device aims to increase the motivation and effectiveness of upper-body strength training.

C8-EB Engineering: Biomedical
Wearable Cardiovascular Disease Prediction Device Powered by a Convolutional Neural Network Based Transfer Learning Algorithm
Alex Ning

The goal of this project is to use different algorithms to detect cardiovascular diseases through the interpretation of
ECGs. A python library, Tensorflow, will be used to create the machine learning model, and the ECG data used for testing will be from online libraries. Besides using new algorithms, another goal is to use a raspberry pi, an inexpensive computer that is much smaller than typical hardware used for ECG readings. This change will potentially make diagnosing CVDs more cost effective and equally reliable.

C9-EB Engineering: Biomedical
Xylem-Inspired Irrigation: Mimicking Nature for Efficient Watering Systems
Dhanusha Ramadas

Xylem is a crucial vascular tissue that provides structural support and transports water from the plant’s roots to its
leaves. Xylem, composed of tracheids and vessel members, transports water in stable increments, regulating water flow. It transports water to plant leaves by adhesion (water molecules’ ability to attach to plant tissues), cohesion (water molecules’ ability to attach to each other), and transpiration pull (the pull of water upward due to transpiration). Additionally, it transports water to the area of the plant that needs water the most by the transpiration demand (more water is needed where there is more evaporation). Inspired by the fascinating work of xylem, this project aims to create an automated irrigation system with similar structures to xylem. This irrigation system only waters plants when necessary by reviewing soil moisture sensor readings. Additionally, it does so stably without flooding the soil by incorporating the biological structures of plant xylem. The idea of watering plants in this manner would not only increase a plant’s health , by preventing either over or under watering, but also make caring for plants much easier for individuals, either beginners or experts in growing plants.

D1-EB Engineering: Biomedical
SkinRAI: A Novel Approach to Leveraging AI and ML for Skin Cancer Detection by Mitigating Racial Bias
Venkat Surya Sai Kotur

Skin cancer is one of the most prevalent forms of cancer worldwide, yet its detection remains disproportionately skewed due to significant racial and ethnic disparities in dermatological datasets. SkinRAI (Skin Recognition Artificial Intelligence) presents a novel approach to mitigating racial bias in AI-driven skin cancer detection by leveraging synthetic data augmentation, deep learning architectures, and fairness-aware machine learning models. Existing AI-based skin cancer classifiers often exhibit substantial performance gaps, especially in underdiagnosing malignant lesions in individuals with darker skin tones. This bias stems from the overwhelming representation of lighter-skinned individuals in training datasets. SkinRAI addresses this critical issue by developing an innovative synthetic dataset generation framework called Pasta, which is integrated with a high-performance convolutional neural network (CNN).

SkinRAI utilizes Pasta, a unique synthetic lesion transplantation technique developed by the author, which transfers malignant skin lesion features from lighter skin tones onto darker ones. Using advanced techniques like StyleGAN2, Poisson blending, and inpainting, Pasta enhances the diversity of training datasets, ensuring better representation of varied skin tones. The SkinRAI model is built on the powerful EfficientNetB4 architecture and is trained using the synthetic dataset generated by Pasta. This methodology optimizes feature extraction across diverse skin types, significantly improving diagnostic accuracy for both lighter and darker skin tones.

To ensure fairness and reduce bias, a fairness-aware algorithm is incorporated into the SkinRAI model, actively mitigating disparities and enhancing the model’s diagnostic equity across different demographic groups.

The model was trained using Nvidia GPUs on AWS EC2 instances, ensuring scalability and optimized computation. A comparative analysis with a model based on the SimpleCNN architecture, trained on the well-known HAM10000 dataset, demonstrates SkinRAI’s superior performance. With an impressive 92% accuracy, SkinRAI excels in detecting
melanoma and other malignant skin conditions, particularly in underrepresented populations, while significantly reducing false-negative rates.

For real-world applicability, SkinRAI is deployed as both a web and mobile application, facilitating real-time, AI-assisted skin cancer detection for patients, healthcare providers, and community clinics. The mobile app offers intuitive, on-device analysis, making it especially valuable in underserved regions with limited access to dermatological care. The web application is designed for seamless clinical integration, providing diagnostic support to medical professionals.

Preliminary results indicate that SkinRAI outperforms existing AI-based dermatological models in both fairness and accuracy, setting a new standard for equity-driven AI-powered healthcare solutions. By combining synthetic data augmentation, bias- mitigating AI models, and scalable cloud-based deployment, this research underscores the pressing need for inclusive medical AI and paves the way for ensuring equitable, high-accuracy disease detection for diverse populations worldwide.

D2-EB Engineering: Biomedical
UV treatment of Vitiligo based on Deep Learning and 6-DOF Robotics Arm
Xuhong Ding

Vitiligo, the most common depigmenting skin disorder, often results in significant psychological harm for patients. Existing treatments like UV phototherapy face limitations in cost, accessibility, and precision. This study attempts to address such challenges through an innovative approach based on deep learning and robotics. The proposed system combines a 6-degree-of-freedom (6-DOF) robotic arm, 308nm UV phototherapy, and machine vision to automate detection and targeted treatment.

The machine vision module employs a Hikrobotics depth camera and a UNet-based algorithm for high-accuracy (94.8%) segmentation of vitiligo patches, leveraging depth and RGB data to delineate lesion boundaries. The treatment module utilizes 308nm UV light, optimizing melanocyte stimulation and minimizing side effects on healthy skin . A 6-DOF robotic arm, calibrated via hand-eye coordination and inverse kinematics, ensures accurate path planning and UV delivery. System integration involved coordinate transformation, trajectory interpolation, and hardware assembly, supported by a computing unit for real-time control.

Experimental evaluation on 186 vitiligo images demonstrated robust performance, with 92.5% precision in lesion detection and efficient treatment path planning . Safety protocols maintained UV exposure within medical guidelines. The major challenge was detecting small or irregular patches, highlighting future directions for algorithm refinement and device portability.

This system represents a significant advancement in vitiligo care, reducing reliance on specialized clinics and enabling an affordable, accessible, precise, and user-friendly treatment. Future work will focus on enhancing detection algorithms, optimizing portability, and exploring synergies with drug delivery systems to improve therapeutic outcomes.

Keywords: Vitiligo Treatment, 308nm UV Phototherapy, Deep Learning (UNet Algorithm), 6-DOF Robotic Arm, Depth Camera, Image Segmentation, Trajectory Planning

D3-EB Engineering: Biomedical
Utilizing Physics Informed Neural Networks to Effectively Predict the Development of Atherosclerosis in Patient -Specific
Models
Eric Nie

Atherosclerosis, a leading cause of cardiovascular diseases such as stroke and myocardial infarction, is strongly
influenced by hemodynamic factors like wall shear stress (WSS). Traditional methods, such as computational fluid
dynamics (CFD) and experimental techniques, struggle to accurately capture near-wall flow and WSS due to the complex and turbulent nature of blood flow. This study explores the application of Physics-Informed Neural Networks (PINNs) to model blood flow in carotid artery bifurcations, integrating the Navier Stokes equations into the loss function to combine data-driven learning with physical constraints. By systematically investigating sensor density and location, I found that higher sensor densities improve accuracy, with optimal placement at 0.9 R providing the best predictions for both sub boundary and near-wall flow. Application to a patient-specific carotid artery geometry demonstrated PINNs’ ability to adapt to complex scenarios, achieving an average velocity difference of 0.0037 m/s compared to CFD simulations. These results highlight PINNs’ potential to bridge the gap between idealized models and clinical applications, offering a computationally efficient and accurate tool for personalized hemodynamic assessments and advancing cardiovascular disease diagnosis and treatment.

Keywords: Atherosclerosis, Physics Informed Neural Networks (PINNs), Computational Fluid Dynamics (CFD), patient specific

D4-EB Engineering: Biomedical
A Personalized Fatigue Metric Using M-Space Evaluation of Movements In a Fatigue Model of Movement Degeneration
Inversely Correlated with the IJS Skating Skills Component Score
Nicholas Ying

Background: Currently, there is a need for a method to accurately measure and evaluate movements for clinical
management and in clinical trials. This study uses fatigue in ice dance movements as a model system for movement
degeneration. This could provide a more objective and quantifiable way to assess and monitor neurodegenerative movement disorders.

Methods: Two skaters performed sequential patterns of the Dutch Waltz and self -reported fatigue. Individual dance steps, measured using Noitom Perception Neuron inertial measurement unit (IMU) sensors, were mapped to m-space, as represented by the first three components from principal component analysis (PCA). Video review of IJS Skating Skills component score (SSCS) was rated by a figure skating expert.

Results: Using Hotelling’ s T-squared statistic, after correction for multiple comparisons, step groupings were significantly different, and individual steps were significantly different from each other (p<0.001), with the exception of duplicated steps. Using Student’s T, SSCS for early/late sets for each step were significantly different (p<0.001). Using Pearson’s R, SSCS’s inversely correlated with fatigue score as defined by step-specific m-space and individualized step specific m-space (R<-0.7, p<0.05) but not generalized m-space (-0.70.05).

Conclusion: In this proof-of-concept pilot study, we demonstrate that: 1) when movements are mapped to dimensionally reduced “movement space” (m-space), distinct steps cluster together, 2) movement quality decreases with both measures of fatigue, 3) the location of a given movement in step-specific m-space can be used to define a “fatigue score” that correlates with self-reported fatigue and inversely correlates with movement quality.

Key Words: Injury Prevention, Disease Progression and Management, Wearable Sensors, Motion Capture, Sports Medicine, Quantified Self, Decision Support

D5-EB Engineering: Biomedical
Reading Emotions Using Arduino and Brainwaves
Mannat Markan

This project aims to develop and test a DIY EEG Brain wave reader and generate art using those emotions . Converting
brain waves into art is significant because it connects neuroscience and creative expression. This allowed us to visualize
emotions. It is designed to help disabled individuals communicate and express their emotions/feelings. The project will
explore how the EGG brain wave reader interprets feeling to art. The ultimate goal is to enhance communication methods for people with disabilities. The project must accurately read emotions (70%), must light up one light at a time, must display Delta, theta, low beta, high beta, low gamma, and Mid gamma on-screen online.

D6-EB Engineering: Biomedical
Cuffless Blood Pressure Measurement Monitor via Pulse Transit Time Analysis and Arduino
Louis Chiu
Shuting Zhu

This project addresses the accessibility gap in hypertension monitoring by developing a low-cost, cuffless blood pressure device. With 48.1% of U.S. adults affected by high blood pressure, many cannot afford standard monitors ($50-$100) or find traditional cuff-based systems uncomfortable. We designed and built an Arduino-based alternative utilizing Pulse Transit Time (PTT) methodology at a $25-$40 price point. Our prototype integrates an Arduino Uno R3 microcontroller with pulse sensors in a 3D-printed housing, eliminating the need for uncomfortable cuff inflation while maintaining clinical accuracy . Testing against commercial monitors demonstrated relatively accurate measurement precision with an average of 7.64% error for SBP and 8.88% error for DBP. This innovation demonstrates how engineering solutions can address healthcare accessibility challenges while potentially improving hypertension management for underserved populations.

F1-EB Engineering: Biomedical
A Pipeline Prediction Tool of Cerebral Hypoxia Using Digital Histology Staining , U-Net Segmentation, and Multimodal Deep Learning Frameworks
Avaneesh Mohan

Clinicians use a multitude of data and their own expertise to decide treatment options of patients and tumors. A large factor in determining treatments is the incidence of hypoxia, which is the lack of oxygen transport due to irregular vascularization or inefficient vascularization of tumor tissues . The lack of oxygen in tumors creates a microenvironment for tumors to grow more resistant to treatments such as radiotherapy, chemotherapy, and facilitating tumor growth through increased invasion, immune evasion, and increased angiogenesis and tumor vascularization. A tool to predict cerebral hypoxia would be invaluable information to clinicians to help decide on treatment type and intensity. Using pre-trained U-net models and fine-tuning them to apply to histological images, the model will segment blood-vessels, tumor-tissue, necrotic regions and perivascular regions. Using the python library Sci-Kit image, we extracted features such as vessel diameter, spatial distribution of vessels, shape metrics of vessels, vessel density, etc. Finally, using convoluted neural networks the model is trained in addition with genomic data on over 600 cases.

T10-EB Engineering: Biomedical
Implications of Eye Rotation in Optimizing Prism Glasses for Clarity Alignment
Tony Badawi

Prism glasses are one of the many treatments for patients with strabismus, which essentially is a condition where one
eye is off-center. Although it can lead to prism adaptation when worn long-term, prism glasses are a way to align the
misaligned eyes of a strabismus patient and unify double vision. This paper discusses the findings of a few experiments that suggest the existence of a point between full alignment and no alignment, where the crossed eye in a specific strabismus patient exhibits vision the clearest. The more powerful a prism is, the blurrier it tends to be, so this point implies that a weaker prism could plausibly generate a clearer image than a typical corrective lens, assuming the brain would attempt to fuse the images after the prism partially aligns the double vision. If this phenomenon can be generalized for other strabismus patients, it could possibly mitigate the effects of prism adaptation in current long -term users, while potentially opening the doors for treatment to strabismus patients looking for glasses with better clarity.

Keywords: strabismus, double vision, alignment, prisms, prism adaptation, crossed eye, point of clarity

T11-EB Engineering: Biomedical
Anticipating Thyroid Cancer Recurrence via an Advanced Integrative Computational Machine Learning Framework
Ishika Kumar

Thyroid cancer is a cancer located in the thyroid gland. It is one of the fastest growing cancers worldwide, and has a
significant risk or recurrence. Can one prognosticate thyroid cancer recurrence and its most impactful factors by creating an artificially intelligent model? By developing a machine learning program ( a certain facet of AI) one will create an artificially intelligent model that detects any chance of remission in thyroid cancer patients. In order to create such a thing, an understanding of complex mathematics, machine learning programming, random forest algorithms, and the dynamic nature of diseases (specifically, thyroid) is vital. Testing of the model includes accuracy calculations and examining its performance using an AOC Curve. This project necessitates technical excellence and has the potential to create a profound impact on the world as it concerns the indispensable fields of medical sciences and artificial intelligence.

T12-EB Engineering: Biomedical
A Path to Improved Fetal Cardiovascular Health Outcomes Using Machine Learning
Christina Quin

Reduction of child mortality is a significant focus of the UN’s Sustainable Development Goals (SDGs). The UN seeks to
eliminate preventable neonatal and under-5 deaths by 2030 [1]. Maternal mortality, which accounted for 295,000 deaths during and following pregnancy and childbirth in 2017, is closely linked to child mortality. Most of these deaths 94 percent occurred in low-resource settings and were preventable [2]. Cardiotocograms (CTG) offer a simple and cost-effective method to assess fetal health, which is crucial in preventing child and maternal mortality. CTG equipment works by sending ultrasound pulses and analyzing the responses to monitor fetal heart rate (FHR), fetal movements, uterine contractions, and other parameters in the dataset. Using CTG data, machine learning models—Multinomial Logistic Regression, Random Forest, Gradient Boosting and Fully Connected Feedforward Neural Network (FCNN) —were employed to classify fetal health into three categories: Normal, Suspect, and Pathological. Strategic feature preprocessing and class-balancing techniques were implemented to enhance model performance. Results indicate that FCNN achieved the highest overall accuracy, while Gradient Boosting provided superior recall for pathological cases. Additionally, FCNN demonstrated strong recall for the suspect category, ensuring accurate identification of at-risk cases. These findings suggest that FCNN is an effective model for fetal health classification, with Gradient Boosting serving as a strong alternative for prioritizing recall of clinical scenarios. By minimizing observer dependency, the study aims to mitigate unnecessary medical interventions while providing a consistent, accurate, and cost-effective approach to neonatal health assessment [3].

Keywords — Artificial Intelligence, Fetal Health, Cardiotocogram (CTG), Machine Learning, Neural Networks.

T1-EB Engineering: Biomedical
Detection of Bacterial Contamination in Water
Joseph Cordero
Sully McCarthy

This project aims to design an affordable testing system that creates a clear marker of bacterial contamination in water
samples. To achieve such a design, experiments were conducted to develop the elements of the kit. A combination of gram staining, a method of dying bacteria, and buoyancy agents the testing system will create a colorful positive line at the surface of a given water sample that indicates bacterial contamination. To complete the necessary experiments with minimal risk of exposure, we used Lactobacillus acidophilus bacteria as a stand-in for more harmful bacterias. Ultimately, our test kit is designed to be cheap and therefore available for humanitarian aid around the globe, and potentially preventing infection and death due to unsafe drinking water.

T2-EB Engineering: Biomedical
Screening Ligand-Protein Interactions for Exosome-Based CRISPR Delivery Across Multiple Cancer Types
Aru Sharma

CRISPR-Cas9 gene editing offers revolutionary potential for cancer treatment , yet existing delivery methods face critical limitations, including off-target mutations, immune activation, and inefficient cellular uptake. This project proposes a novel, targeted approach: identifying optimal ligands to design ligand-functionalized synthetic exosomes to selectively deliver CRISPR-Cas9 to cancer cells. Through molecular docking simulations, I identified the optimal ligand-receptor interaction from ten ligands and seven cancer-specific receptors, ensuring precise tumor targeting. The optimal ligand was also docked with a few healthy cell receptors to understand the chances of off -targets. The selected ligand is conjugated to exosome surfaces, enhancing their affinity for cancer cells while minimizing systemic exposure . These exosomes encapsulate CRISPR-Cas9 as an mRNA payload, promoting efficient gene editing within malignant cells. The design is supported by thorough and proper research as I incorporated functionalization and loading techniques into my design. An additional part of my design could be adding pH-sensitive linkers to ensure the therapeutic molecules are specifically being delivered to the target cells. Computational validation helps validate the accuracy of targeting and delivery. By addressing the limitations of current CRISPR delivery systems, this research aims to improve the safety and efficacy of gene therapy , offering a transformative step toward precision oncology.

T3-EB Engineering: Biomedical
Development of a Microfluidic Device for Improving Cryoprotected Cell Therapies
Seiji Ting

Chimeric Antigen Receptor (CAR) T cells are an exciting new cellular therapeutic platform applicable to a wide range of different cancers. Although they can be extremely effective, CAR-T cells must be cryopreserved in dimethyl sulfoxide (DMSO), which is toxic to humans and causes negative side effects . Removing the DMSO from the cells before administering them to a patient would alleviate many of the side effects of the treatment; however, diluting the cells too fast could cause them to lyse due to highly hypotonic conditions. Centrifugation is one method of removing DMSO, but this is impractical in the clinic. Microfluidic devices offer a method to dilute and separate DMSO from cells gradually in a platform that can be easily deployed in a clinical setting.  By comparing different channel designs and developing prototypes, I was able to develop a working model of a cell therapy concentrator that was optimized for cell viability and speed. I hypothesized that smaller pipe diameter and smaller helix diameter would enhance mixing using calculations with the Reynolds and Dean numbers. I tested this hypothesis by printing iterations of a single loop and analyzing channel intensity profiles, and found that smaller pipe diameter and larger helix diameter tended to mix better. The fluid exchanger utilized 6 loops with a 1 mm pipe diameter and 15 mm helix diameter to continually dilute and reconcentrate the cell solution. This yielded more live cells than a standard dilution, demonstrating that microfluidic devices are effective at removing DMSO from a cell solution.

T4-EB Engineering: Biomedical
Simutaneously Targeting Photons to Mitochondria and Lysosome via Nanophotosensitizers
Eagle Pei

Photodynamic therapy (PDT) utilizes photosensitizers (PSs) such as porphyrins to generate reactive oxygen species (ROS) that induce cancer cell death upon light activation. Porphyrins are particularly effective due to their strong absorptivity, high quantum yield, photostability, and low dark toxicity. However, the efficiency of PDT can be limited by factors such as PS aggregation, reduced ROS generation, and organelle-specific localization. To enhance PDT efficacy, we designed a dual-targeting nano-photosensitizer capable of co-delivering 5-aminolevulinic acid (5-ALA) and Chlorin e6 (Ce6). Upon cellular uptake, 5-ALA preferentially localizes to mitochondria and converts into protoporphyrin IX , whereas Ce6 remains in lysosomes. Our results demonstrate that this dual-organelle targeting significantly amplifies ROS production, leading to enhanced cytotoxicity compared to single-targeting strategies. Nanoparticles were synthesized using a one-step sonication method and characterized by dynamic light scattering, electron microscopy, and spectroscopic analyses, confirming appropriate size, charge, and photophysical properties for efficient cellular uptake and lysosomal escape . Confocal microscopy verified precise co-localization of Ce6 in lysosomes and 5-ALA-derived protoporphyrin IX in mitochondria. Cytotoxicity assays confirmed that dual-targeting nanoparticles significantly improved PDT effectiveness , killing approximately twice as many cancer cells compared to single-organelle-targeted nanoparticles. This study highlights the potential of dual-organelle-targeted nano-photosensitizers to increase PDT effectiveness and offers promising insights for future clinical translation.

T5-EB Engineering: Biomedical
Innovative and Cost-Effective Solutions to Prevent Cobalt Toxicity Exposure for Miners in DR Congo
Seoyoung Chang

With the rise of climate change, the demand for electric vehicles (EVs) has increased as a way to reduce fossil fuel
consumption. Particularly, in regions like Massachusetts, rising sea levels and increasing temperatures highlight the
urgency of transitioning to sustainable energy solutions. A key component in EV batteries is cobalt, with 70% of global
production concentrated in the Democratic Republic of Congo (DRC). Heavy metal exposure from mining operations
threatens miners, as many don’t have the financial capabilities to be equipped with protective gears, left with facing
hazardous conditions, exploitation, and long-term health consequences. This study explores whether plant-based materials readily available to DRC miners can absorb heavy metals and be incorporated into protective equipment. Moringa, perilla, and sunflower, widely found in the region, are known for their ability to absorb toxic substances and form protective barriers. The results of the research revealed that cobalt disrupts tumor-suppressing mechanisms in skin cells, increasing vulnerability to cellular damage. A plant powder mixture (PPM) composed of moringa seeds and leaves, perilla leaves, and sunflower leaves significantly reduced cobalt-induced cell damage and minimized cobalt penetration when applied to fabric. These findings suggest that integrating PPM into recycled clothing to create protective workwear offers a sustainable and locally accessible solution to mitigate heavy metal exposure. By utilizing indigenous plants, this approach not only enhances workers safety but also promotes environmental sustainability in cobalt mining regions, addressing both health risks and the broader humanitarian issues of cobalt extraction.

T6-EB Engineering: Biomedical
Nanoparticle-Based Disruption of Biofilms: A Novel Strategy to Counteract Antibiotic Resistance
Jianna Bixho

The emergence of drug-resistant strains has increased due to the presence of biofilms. Biofilms are large communities of
bacteria that cover themselves in a slimy matrix. These communities can develop in extreme environments and cause major problems in the medical world including the development of antibiotic resistance. Antibiotic resistance is the process in which bacteria mutate to resist antibiotics. Biofilms are especially good at this because of their slimy matrix. Each type of biofilm has a different composition, so there is no universal treatment, and the current treatments are losing effectiveness. Current treatments for biofilms are bacteriophages but they face lots of backlash in the medical world due to their invasive nature. To attempt to find an alternate treatment, this work will focus on creating a silver nanoparticle that can eradicate two types of E. coli biofilms. Specifically, biosynthesizing silver nanoparticles using Aloe vera to reduce the chemical waste emitted from traditional ways of creating nanoparticles. The importance of nanoparticles is that they range from 10-100 nanometers. This allows the nanoparticles to surpass the slimy matrix of the biofilm due to its small size. The results of this experiment indicate that silver nanoparticles can hinder the formation of biofilms. In conclusion, this will aid the process of finding a universal treatment for biofilms and avoiding antibiotic resistance.

Keywords: Biofilm, Antibiotic Resistance, Nanoparticle, Biosynthesis, Aloe Vera, E. coli

T7-EB Engineering: Biomedical
Development of a Novel SHERLOCK Diagnostic Assay for Rapid Detection of Borrelia burgdorferi
Kevin Branda

Lyme disease, caused by Borrelia burgdorferi and transmitted by tick bites, is a common vector-borne illness in the United States. Despite over 63,000 reported cases in 2022, the true annual incidence is estimated to exceed 500,000 due to underdiagnosis. Early detection is critical, as untreated infection can lead to serious complications. Current diagnostics rely on serologic testing, which lacks sensitivity in early-stage disease, and PCR-based methods, which require expensive laboratory infrastructure and extensive training. This project aimed to develop a rapid, point-of- care molecular diagnostic assay for Lyme disease using the SHERLOCK (Specific High-sensitivity Enzymatic Reporter unLOCKing) platform, which utilizes CRISPR-Cas13a in combination with isothermal amplification. The assay targets the ospA gene of B. burgdorferi and is designed to work with a minimally invasive skin fluid collection device ( Tasso) for direct pathogen detection from erythema migrans lesions.

Two primer sets (KB1 and KB2) and a CRISPR guide RNA were designed and evaluated for this purpose. The KB1 primer set demonstrated higher sensitivity compared to KB2, achieving a limit of detection of approximately one target copy per microliter. In silico analysis confirmed high specificity for Lyme-related Borrelia species. Future work will focus on clinical validation and adapting the assay to a lateral flow format. This CRISPR-based assay represents a promising tool for improving early Lyme disease diagnosis at the point of care.

T8-EB Engineering: Biomedical
Orthotic Hand Brace to Improve Range of Motion and Reduce Pain in Low-Functioning Children with Cerebral Palsy.
Sophia Caramanica

Children use their hands to explore and learn about the world. This can be a challenge for children with spastic cerebral
palsy who develop hand contractures, a muscular deformation resulting in stiff muscles and reduced range of motion (ROM). Current treatments include orthotics and electrical stimulation, but the orthotics are either bulky or static, and electrical stimulation is still relatively unexplored, but promising. A hand exoskeleton/robotic glove was designed to provide electrical stimulation to relax the muscles in the hand while using the robotic components to stretch out the contractures and increase range of motion over time. The device was tested over a period of a few weeks. The finger joints were measured before the glove was on, after it was on, and after it was activated. When it was activated, the electrical stimulation unit was turned on and each motor was turned 720Åã. Additionally, an EMG was used to read muscle activation before and after testing. After 4 weeks of testing with one subject, the baseline of 2 out of 4 metacarpal joints increased, 4 out of 4 proximal interphalangeal joints increased, and 2 out of 4 distal interphalangeal joints increased. While these results are statistically insignificant, they indicate a possible upward trend. Future testing will be done over a longer interval with a larger subject group. Children who have cerebral palsy live with chronic pain and are limited by their reduced ROM, but this device could increase their ROM over time, improving their overall quality of life.

Keywords: cerebral palsy, robotic orthotic, electrical stimulation, contractures, range of motion

T9-EB Engineering: Biomedical
A Translational Spectrophotometric Apparatus for Noninvasive Glucose Monitoring in Type II Diabetics
Anoushka Nair
Dylan Striek

Over 462 million individuals suffer from Type II Diabetes worldwide. However, current glucose monitors are painful, costly, and unsustainable. This project establishes a novel, non-invasive monitor that uses the spectrophotometric detection of glucose biomarkers (potassium, sodium, and phosphate) to estimate blood sugar through the finger at wavelengths of 400-800 nm. This uniquely deviates from other spectroscopic monitors that measure glucose directly, requiring inaccessible, expensive equipment emitting wavelengths exceeding 1000 nm.

Our previous research encompassed 250+ experiments involving the induction of varying levels of diabetes in an invertebrate model, extraction of hemolytic fluid (blood equivalent), and the development of a novel prediction model effectively correlating potassium , sodium, and phosphate to glucose (R^2 > 0.9 for all three electrolytes).

This translational model was integrated into a novel prototype (the GlycoCharge). It was created using three distinct LEDs to measure the absorbances of each of the electrolytes (766.5nm, 595nm, and 470nm), a potential fourth LED to measure path length (650nm) as used in the Beer-Lambert Law, an Arduino microcontroller, a photodiode to measure transmittance, and an OLED display.

To verify the prototype glucose output, 100 trials were conducted by comparing readings from a medically verified glucometer to the GlycoCharge device, using previously acquired invertebrate samples. Residual graphs and error charts yielded p > 0.001, indicating that the error associated with the GlycoCharge had no influence from outside factors and that there was a significant increase in accuracy when detecting lower glucose levels. Comparatively, the majority of current glucometers fail to measure the lower extremes of glucose concentrations properly. The overall error for the device is 4.57%, on par with current glucometer market standards.

U1-EB Engineering: Biomedical
Automated Prediction of Autism Spectrum Disorder Severity using Retinal Photograph based on Feature Aggregation with Retinal Blood Vessel Segmentation
Juwon Lee

Early diagnosis of autism spectrum disorder (ASD) is essential for effective intervention, yet current diagnostic methods rely on subjective behavioral assessments and costly imaging techniques, leading to delays in treatment. This study proposes an AI-assisted screening tool using fundus imaging and retinal vessel segmentation to develop an objective, affordable, and efficient method for ASD classification. The hypothesis is that integrating vessel segmentation with convolutional neural networks (CNNs) improves diagnostic accuracy by emphasizing relevant geometric vascular features.

The experimental design involved training a CNN-based model with two input streams: raw fundus images and corresponding vessel segmentation maps. The extracted features were merged using element-wise multiplication and processed through subsequent layers to classify ASD severity (normal, mild, or severe). A publicly available dataset of 57,000 individuals aged 7 to 21 was used for training and evaluation.

Three evaluations were conducted: (1) performance assessment using accuracy, precision, recall, and F1-score, (2) classification accuracy using a confusion matrix, and (3) the impact of incorporating vessel segmentation. The proposed model achieved 84.2% accuracy, outperforming prior methods. The confusion matrix indicated high classification performance, though distinguishing between mild and severe ASD remained a challenge. Additionally, vessel segmentation improved accuracy by 1–2%, demonstrating the potential of retinal biomarkers for ASD screening.

This study demonstrated the feasibility of retinal imaging for ASD classification, paving the way for a cost-effective screening tool. Future research should explore model refinement with younger age groups and real-world clinical validation to enhance early detection and accessibility.

Keywords: autism spectrum disorder, fundus imaging, retinal vessel segmentation, artificial intelligence in healthcare, AI-assisted diagnosis.

U2-EB Engineering: Biomedical
Designing a Novel Bamboo-Inspired Lattice Structure for 3D-printed Total Knee Arthroplasty Implants: Finite Element
Method and Machine Learning-based Optimization Approach
Yaejoon Jung

Total knee arthroplasty (TKA) is a widely performed surgical procedure aimed at reducing pain and restoring mobility in patients with severe osteoarthritis. Despite past advancements in the field, TKAs still face challenges related to implant personalization, mechanical performance, and longevity. The combination of 3D printing technology and biomimetic lattice structure can offer promising solutions, but due to the vast design space of the lattice geometry, the design process and its optimization has been traditionally slow. In this study, a novel bamboo-inspired lattice structure was designed for metal 3D-printed Ti-6Al-4V knee implants, with a focus on optimizing its energy absorption and porosity. Utilizing finite element method (FEM) simulations combined with Gaussian process and Bayesian optimization, the strategy successfully identified an optimal structure within the defined design space within 50 simulations. A final structure with a 70.69% improvement in energy absorption and a low packing density of 0.48 could be achieved. Moreover, the optimized pore sizes were found to be in the ideal range for promoting effective osseointegration . This study highlights the potential of machine-learning based optimization in accelerating the lattice design process and significant implications for improving implant performance and personalization in orthopedic applications.

U3-EB Engineering: Biomedical
Which Non-Absorbable Suture is the Strongest?
Manal Mahfudh

Suturing is the primary method to close major wounds on the human body. This method of closing wounds is commonly
used throughout the medical industry, along with staples, bandages, and the like. For this project, I have scientifically tested which non-absorbable suture type is the strongest. I hypothesized that if I measured the holding strength of polypropylene monofilament, then it would hold the most weight before snapping because this material is most commonly used in the medical field out of the four other non-absorbable sutures I will be testing. I measured and recorded the Newtons of force each material can sustain before breaking against a force meter. The materials I tested on included polyester braided, silk braided, nylon monofilament, and polypropylene monofilament sutures. In terms of results, polyester braided was found to be the strongest suture recorded at 29N. I noted that braided sutures tended to be stronger compared to monofilament-made sutures. Polypropylene monofilament was not the strongest suture recorded, therefore my hypothesis was rejected. I hope that doctors and other professionals in this field will be able to consider my data findings and possibly be useful in their future endeavors regarding suturing types.

U4-EB Engineering: Biomedical
Early Diagnosis of Parkinson’s Disease by Nano Detection of Biomarker Alpha Synuclein
Claire Liu

In recent years, Parkinson’s Disease (PD) has become a more global issue, impacting over 10 million people worldwide,
and is expected to rise continually due to an increase in the aging population. As slowing down the disease progression is its only treatment, diagnosing the disease as early as possible is critical for an effective intervention . Thus, finding a
biomarker that accurately and consistently reflects the early stages of PD is necessary. In recent studies, researchers have analyzed many biomarkers and identified several that allow for an accurate and early diagnosis of the disease. In particular, alpha-synuclein(α-syn) and its oligomers (α-syn olig) showed potential for a widespread screening and an early noninvasive diagnosis. This study aims to detect salivary α-syn and α-syn olig at low concentrations, utilizing surface-enhanced Raman scattering (SERS) spectroscopy. By enabling the analysis of biomarker concentration in different fluids combined with silver nanoparticles (Ag np), this method would develop a diagnostic procedure for the early stages of the disease. This study employed various experimental techniques to analyze the biomarker under different circumstances by demonstrating the functioning of silver nanoparticles on α-syn, finding a concentration-signal intensity curve, and simulating in artificial salivary samples. The outcomes of this study may enhance future research in Parkinson’s Disease and other biomarker -based diagnostics, improving the quality of life of many people.

Keywords— Alpha-Synuclein, SERS, Parkinson’s Disease, Silver Nanoparticles, early diagnosis

U5-EB Engineering: Biomedical
Lifestyle Modifications in Patients with Alzheimer’s Disease
Hwiseo Choi

Major neurocognitive disorder, which is also known as dementia, is a neurodegenerative disorder that is characterized by a significant decline in one or more than one of the cognitive domains; learning/memory, social cognition, complex attention, executive function, language, and perceptual-motor function. Alzheimer’s disease (AD) is the most common type of major neurocognitive disorder and affects about 1 in 9 people (10.8%) age 65 or older. Despite advances in medical research, no cure has been found yet, making AD one of the leading causes of morbidity and mortality in elderly patients. There are hypothetical protective factors for AD; exercise, diet control, and cessation of smoking. However, the previous studies to explore the protective effect of these factors on cognition in patients with AD were inconclusive due to the conflicting results and small sample size. As such, this systematic review investigates the effects of lifestyle interventions– exercise, diet, and smoking cessation–on cognitive function in AD patients by utilizing meta -analysis based on the previous clinical trials. A quantitative analysis revealed that there is no evidence of improved cognitive functions by aerobic exercise measured by MMSE and ADAS-Cog. Further, dietary interventions, specifically the Mediterranean diet and Omega-3 supplementation, suggested cognitive benefits. In addition, the data on smoking cessation and cognition highlighted the need for supportive interventions in smoking patients with AD. Overall, this analysis suggests that the progression of ADA may not be reversible by lifestyle modifications alone and these interventions should be administered as an add-on therapy on top of the current pharmacological interventions.

U6-EB Engineering: Biomedical
AI-Powered Skin Disease Detection App: Convenient Disease Detection through Image-Based Analysis
Alex Kuai

Skin and oral health conditions often go undiagnosed due to limited access to timely medical consultations and
diagnostic tools. This project introduces a novel mobile application that leverages artificial intelligence to analyze
user-uploaded photographs and provide medical recommendations for various conditions, including melanoma, eczema, acne, bug bites, cavities, oral cancer, dry sockets, and more. The goal of this app is to give recommendations for everyday people on whether they should seek medical attention and proper diagnoses. This app is by no means an official medical diagnosis and should not under any circumstance be used as one. Utilizing state-of-the-art image recognition and deep learning methodologies, the app ensures high accuracy in detecting both common and rare conditions. By offering general risk assessments and suggesting whether professional medical intervention is necessary, the app empowers individuals to take proactive steps toward their health. This solution aims to bridge gaps in accessibility, providing a convenient, cost-effective, and user-friendly platform for preliminary medical insights while reducing healthcare disparities.

D10-EE Engineering: Electrical and Mechanical
Mitigating Vibrations in High-Speed Helicopters: Model to Investigate an Electric Power Transmission System for a Vibratory Control Moment Gyroscope
Jacob Lee

To counter intensive vibrations produced by Sikorsky’s X2 helicopter’s unique coaxial rotor setup, engineers at Sikorsky
designed the Vibratory Control Moment Gyroscope (VCMG) which uses pinion-bevel gears to spin high-speed discs and a rotating cage, generating a counteracting moment to cancel out the vibrational forces applied. However, the constant grinding of these gears due to the discs need for a high, sustained RPM leads to rapid wear and failure. My research introduces, improves on, and tests a novel electric power transmission system to replace the VCMG’s mechanical pinion -bevel gears. Instead of needing a gear assembly to spin the discs, a single motor would turn a generator that spins the rotating cage of the VCMG while also producing an AC current to be rectified to power the gyroscopic discs simultaneously . Through rigorous testing, I will determine the viability of this electric solution, and if proven capable, retrofit a mechanical VCMG with the electric solution. If successful, this innovation would greatly alleviate the stress, enhance reliability, and lengthen the VCMG’s service life without sacrificing its effectiveness, solving a major hurdle for the X2 family.

D11-EE Engineering: Electrical and Mechanical
Optimizing an After-Market Prototype that Operates a Window Remotely
Brandon Guyott
Justin Romanelli

This project seeks to optimize a method for those with muscular inhibitions, especially the elderly, to be able to open
windows without the help of others. The ability to open a window is vital for increasing ventilation to reduce the spread of airborne illness and regulate temperature. The shafts of two stepper motors mounted on the top of the window are coupled to lead screws that feed through corresponding nuts bolted to the window sash. The window moves up or down depending on the direction of the motors. Python code was written for a Raspberry Pi, such that the window opens when a button is pressed on either the IR Remote or the Touchscreen. A safety system was implemented, consisting of three HC-SR04 Distance Sensors, to ensure that harm could not be dealt from operation of the window. Forty trials were set up, twenty for controlling the window with the IR remote and twenty for the Touchscreen. For each trial the following measurements were recorded: noise of operation, time to open the window, one operation consistency test (if the window opens and closes fully without interruption) and three safety consistency tests (if the window stopped moving when an object was detected) for each of the three objects: stuffed animal, brick, and plastic hand. The consistency of operation (97.5%) and the consistency of the safety system (100%) both met their goals of 95% and 100%. The time to open the window exceeded its goal of being quicker than 15-20 seconds with an opening time of about 7.62 seconds, on average. The noise of operation exceeded its goal of being lower than 60 db with a mean noise level of 54.9 db. However, the current design failed its qualitative goal of sleekness and modularity to reflect a potentially marketable product. Overall, the project is considered successful in meeting its goals, though there are many improvements that can be made. One improvement that can be made to the design is to better accommodate manual operation, such as with a hand crank that uses bevel gears to turn the lead screws and move the window. Temperature and humidity sensors can be implemented to close the window in undesirable ambient conditions. A printed circuit board can be designed as a more modular and sleek substitute for the current breadboard system.

D7-EE Engineering: Electrical and Mechanical
From Crude to Extrude: A Novel Approach to Combining Pultrusion and Extrusion Technologies into a Single Filament Production System to Recycle PET Bottle Waste and 3D Printing Purge Materials into Additive Manufacturing Filament
Benjamin Davis

As 3D printing becomes more and more common, homemade filament becomes easier to make. Traditionally, filament is made with large, industrial machines that can cost 10s of 1000s of dollars. Alternatively, DIY and homemade options can cost less than $1000, but at a much lower production rate and value. The purpose of this project was to develop a homemade extrusion machine that is similar to the commercial ones, at the cost of a DIY machine. In addition, this project aimed to combine two different filament production technologies -extrusion and pultrusion- in a single machine, something which there is no documentation of ever having been done before. This combination of technologies would lead to more efficient production. The project began by creating a machine to manufacture a compression screw, which is the most expensive and most critical component of the extrusion process. Then, an extrusion machine was made. Following this, a water cooling system, air cooling system, spooling system, filament pulling system, and pultrusion system were created. Different plastics (PLA; PET) were extruded at different temperatures to analyze the proper extrusion temperatures of those materials. The water cooling system and pultrusion unit went through multiple iterations until they were able to work together and produce filament. The expected outcome of this machine was that it would produce filament efficiently and work together as a system, with a high power consumption and commercial grade tolerances. The power consumption of the machine was analyzed, an average of ~0.89 kWh, as were the tolerances of produced filament, Å}0.129 mm. Overall, each component of the system worked, but they did not function together as a group. There is a need for further experimentation in the implementation of the pultrusion unit, water cooling system, and auger parameters.

D8-EE Engineering: Electrical and Mechanical
Regenerative Shock Absorber with an Automatic Damping Adjustment System
Xiaoke Yu

Shock absorbers are components in a car that convert excess kinetic energy in the suspension to heat and improve the vehicle’s drive experience. Traditional shock absorbers use hydraulic systems that efficiently dissipate the kinetic energy. However, these hydraulic shock absorbers cannot reuse and recover the energy they dissipate and typically fall short in the variability of their damping force. This paper introduces an innovative Regenerative Shock Absorber with Adjustable Dampening based on two rack-and-pinion mechanisms and a disc-type generator. The rack and pinion mechanisms placed in opposite orientations convert the vertical oscillations from road irregularities to unidirectional rotational motion of the generator’s shaft, which outputs electrical energy stored in a supercapacitor. To adjust the damping force, a sail-winch servo, controlled by an ESP32 board, drives a screw- nut mechanism that moves the generator’s magnets farther or closer away from its coils in response to detected vibration levels. This shock absorber has a recovering efficiency of more than 53 percent within vibration amplitudes of 4.4 to 34.4 mm, and the efficiency goes up to 97 percent under higher vibration amplitudes and frequencies. As the frequency and amplitude of the vibrations increase, the time needed for the shock absorber to charge a supercapacitor decreases exponentially. Moreover, this shock absorber achieves consistent levels of shock absorption under various vibration frequencies and exerts greater damping forces as the vibration amplitude increases.

D9-EE Engineering: Electrical and Mechanical
Amphibious Rescue Vessel Based on Planetary Rotating Paddled Wheels
Haokun Xie

According to the 2021 World Trade Data, maritime transportation accounts for approximately 80% of global cargo
transportation. However, with the increasing demand for trade, accidents at sea are also on the rise. When rescue personnel are dispatched to the scene for recovery, their lives are threatened due to the lack of good rescue ships, and there is an urgent need for amphibious rescue ships with fast speed and strong obstacle crossing capabilities through narrow spaces. Therefore, this project develops an amphibious rescue vessel to provide life support for rescue personnel. The amphibious rescue ship mainly consists of a hull and four inclined spoke paddles. When the wheels rotate, six paddles are rotated to change the driving force, thereby generating a combined force in the direction of the robot’s movement and meeting the speed requirements during rescue. In terms of electronic design, Arduino nano, closed-loop stepper motor with FOC, voltage reduction module, and HC-12 are mainly used as hardware modules. The movement of the structure is achieved by driving the gear module and planetary reduction structure through the stepper motor. Driving tests were conducted in various environments including ground, water, and obstacle crossing, and it was found that the amphibious rescue ship performed excellently, meeting the practical requirements of fast speed and strong obstacle crossing ability.

E10-EE Engineering: Electrical and Mechanical
Oldie but Goldie: Redesigning Constructing Vintage Musical Speaker Systems
Gabriel Walters

This project aims to emulate, redesign, and construct a speaker system similar to a Leslie speaker cabinet, but with a
more affordable and more customizable design and construction .

E11-EE Engineering: Electrical and Mechanical
WATTer Works
Siddharth Gowtham

There are a couple of modern technologies that try and harness the power of flowing water in homes by creating
electricity from it. There is one problem though, the systems that do this limit the water pressure of the pipe that the water is exiting from, often making the water inefficient and unusable for the owner . After finding out how turbines, generators and water pressure works, it was identified that the water pressure regulations for sewer pipes is much lower than for a standard water intake pipe. After this research, a CAD model of a turbine design that could be tested was created. Eight other ones with 3 different blade counts and 3 different blade curvatures combined to make the final 9. These models were 3D printed with PLA+ filament. Then they were tested by connecting them to a PVC pipe and pouring water down it. The electricity produced was measured with a multimeter in mV. Tests were run on the data collected to find that all the data was significant and to help identify some trends in the data. After this, a design was chosen as the most important with possible suggestions and improvements made for future iterations.

E12-EE Engineering: Electrical and Mechanical
Developing A Low-Cost Accessible Musical Instrument Device for Individuals with Physical Disabilities Utilizing Capacitive Touch and Multimodal Input
Arnav Shah
Ruhi Shah

Improvements in computational resources and miniaturization of sensors are enabling the development of digital musical instruments that use non-conventional interaction interfaces. This shift has opened opportunities in the creation of accessible instruments to enable music expression toward persons with disabilities. This work in particular focus on utilizing open source microcontrollers, nontraditional conformable materials such as electrically conductive paint and multimodal input dedicated to people who are physically disabled, for whom the only means for musical expression are the voice and a small range of operable instruments. An adaptable technology stack created with a microcontroller and electrically conductive paint discussed for the design of accessible digital musical instruments targeted at performers with motor impairments. Multimodal interaction channels are analyzed in terms of potential and limitations for musical interaction. Through a series of engineering prototypes and illustrative example uses, we show that our research could enable new opportunities for music expression for disabled individuals through capacitive touch technology and multimodal input.

E1-EE Engineering: Electrical and Mechanical
Detection of Black Ice in Autonomous Vehicles Using Inertial Measurement Based Binary Classification Neural Network
Shengzhen Wang

Black ice poses significant challenges to driving, specifically autonomous driving, due to its difficulty to detect and its impacts on the vehicle safety. Present methods for detecting black ice, although accurate, are still vulnerable to external environmental influences and cannot function in certain environments. Therefore, the research looks into novel methods of all environment black ice detection, using inertial measurement data collected with a scale model of vehicles to train neural networks for binary classification of road conditions. The resulting method from two separate neural network structures are 98.8% and 99.5% accurate respectively, and deployment of the neural network onto Tensor Processing Units (TPU) is proved to be feasible with the average inference time being 0.75 milliseconds and the standard deviation being 0.13 milliseconds. A Two Proportions Z-Test also proves the method’s improvement in accuracy to be statistically significant. 

E2-EE Engineering: Electrical and Mechanical
DAL: Drone-Assisted Acoustic Localization System for Human Detection and Rescue in Disaster-Stricken Environments
Vladyslav Mishyn

Ukrainian rescuers face significant challenges in rubble conditions, including restricted movement, entrapment risks, and the growing threat of secondary rocket strikes targeting first responders. To address these dangers, a universal search system was proposed to detect, locate, and transmit the positions of victims, enabling rapid rescues without manual searching. The system integrates an onboard computer and a microphone array, keeping costs under $1000 three times cheaper than thermal imaging drones. It employs an SVM model for scream detection, a neural network enhanced GCC algorithm for localization, and Wi-Fi for data transmission. The current goal is to develop a prototype capable of pinpointing a screaming victim’s location within a three-meter radius, despite interference from drone propeller noise. Named DAL, from the Ukrainian word “даль” (dal’), meaning “perspective” or “long way,” the project envisions further advancements toward establishing a company under Ukrainian government patronage.

E3-EE Engineering: Electrical and Mechanical
Creating a Soft Biomimetic Robotic Snake With an Anisotropically Frictional Skin Design for Rubble Search and Recovery
Nichelle Thinagar

128 earthquakes over the magnitude 6.0, the threshold after which infrastructure damage is sustained, were recorded in 2023. Active conflict zones worldwide see buildings heavily targeted, and thousands of other disasters occur every year resulting in the deaths of a total of 60,000 – 65,000 people last year alone (Vergin). Technology capable of assisting those trapped under rubble without disturbing its stability in the aftermath of these events is of the utmost importance. In the first 6 hours following such an event, the chances of finding an injured survivor is approximately 60%, but this drops to less than 10% by the 48th hour (WHO). Thus, technology capable of assisting those trapped under rubble without disturbing its stability in the immediate aftermath of these events is of the utmost importance. However, it commonly takes up to a week for search and rescue (SAR) crews to arrive at a disaster site, by which time the number of survivors left to be rescued has already dropped significantly (Mike Cannon, Personal Communication). Moreover, rubble structures must be thoroughly analyzed before rescuers enter to ensure their safety (Gareth Miller, Personal Communication; Mike Cannon, Personal Communication). Though there are several devices used to assist in SAR operations, the majority of these are sensors and cannot directly assist survivors (Gareth Miller, Personal Communication; Mike Cannon, Personal Communication). One device that can enter rubble, FLIR’s PackBot 510, has difficulties navigating particularly small gaps in rubble . It was also designed more specifically for bomb disposal, reconnaissance, and HazMat handling operations, and costs $178,125. 11 in its most basic version (FLIR; Hazmat Resource). One notable alternative for PackBot is robotic snakes. These devices, though, are just as expensive and cannot carry payload, thus not truly assisting survivors. Neither robotic snakes nor PackBots are commonly used in SAR operations due to their limitations and high cost. A bioinspired soft robotic snake is ideal for a device to navigate these environments due to their biological counterparts’ maneuverability in tight spaces . Though such devices have been created in the past, their rigid components prevent them from fully unlocking the capabilities of organic snakes, and semi-soft snake-inspired robots are unable to allow for the improved locomotion of many rigid devices. Furthermore, many of these devices do not follow the modes of gait seen in natural snakes, which are impacted in no small part by their scales, and thus lose the advantages of this adaptation. This research aims to develop a biomimetic soft robotic snake robot that locomotes in a way more like organic snakes do. This research involved the extensive testing and development process of an artificial snake skin with the friction qualities of true snake scales, which aims to facilitate better locomotion in a soft robotic snake. A prototype of the final design of the robot was also created . The eventual design will cost $3,600, a cost much more feasible for the local search and rescue operations that take place in the first 6 hours following a disaster.

Keywords: Biomimetic Robotics, Limbless Locomotion, Biomimicry, Robotic Snake, Anisotropic Friction, Search and Rescue, Natural Disasters, War Zones, Survivor Assistance, Rubble

E4-EE Engineering: Electrical and Mechanical
Fire and Smoke Detection Safety System Leveraging Computer Vision and Thermal Imaging
Ashley Defilippo

Many homes around the world use their stove for cooking. Many of these homes do not have a way to detect potential risks associated with an unattended stove or a fire. Fire risk can come due to 1) what is being cooked on the stove, and 2) due to the unattended stove. Many homes may not have working, or adequate, fire alarm systems and are at risk. In addition, homes in developing countries often do no have stringent building codes that would enforce such requirements. Our project aims to build a low-cost, battery operated device, that can 1) visually assess if the stove is on, 2) detect a non-normal flame on a stove (i.e. a flame that looks like it might start a fire), and 3) assess if the stove is unattended for too long (e.g. attendance time is related to how often a person, or person’s hand for example, interacts with the stove). The device should be easy to operate, require only standard batteries and be easy to install on an existing hood or range stove. It should also be a possible alternate for low income homes all around the world such as in third world countries that do not have fire alarms.

E5-EE Engineering: Electrical and Mechanical
A Novel Electric Propulsion System for Watercraft: Harnessing Magnetohydrodynamic Thrust for Solid-State Water Transportation
Tayyab Afzal
Thomas Le

In our experiment, we optimized the existing technology behind the magnetohydrodynamic (MHD) drive to create a
powerful and high-velocity propulsion system. By using and manipulating principles of electromagnetism and fluid dynamics, we aim to develop a three-stage model capable of efficient and sustainable maritime transportation in the future : the research process combined primary literature review, mathematical predictive modeling, and physical prototyping. A linear model was first created, optimized, and then tested to refine performance. After maximum optimization was achieved, more complex stages were designed. The process of testing and improving allowed for the identification and remedying of challenges surrounding thrust generation, energy efficiency, and material derogation. By optimizing electrode alignment, magnetic field strength, and electrical input, we maximized the Lorentz force while minimizing electrical leaks. Material selection and cost are critical considerations, ensuring the MHD drive remains operational under varying conditions, while still remaining in the budget. To achieve the desired velocity and consistent reliability , computational predictions and real-world testing enabled a comprehensive understanding of the drive’s capabilities . The goal was to demonstrate the feasibility of an advanced MHD drive as an environmentally friendly and high-performance propulsion system for maritime applications.

E6-EE Engineering: Electrical and Mechanical
Extraction Hive
Ethan Okeke

The project will consist of the constant heating and cooling of air in order to separate ambient gasses.

E7-EE Engineering: Electrical and Mechanical
Smart Module For Canes
Vinesh Kanagavel Chithra

The current solution that addresses the mobility issue for blind people indoors and outdoors is canes– regular
and smart. Smart cane provides a promising opportunity that enables safe travel for blind people because it combines the features of a regular cane, having a feel on their surroundings as well as the features of a smart cane; detect how far obstacles and in which direction obstacles are and advanced navigation assistance. This is such a valuable solution because this combination secures blind people from threats as they receive fast inputs from built-in devices to act accordingly. But, current models can cost anywhere from $850 to $1050 which is not accessible for all. The purpose of this project is to make this helpful technology accessible to everyone. In this process though, none of the important abilities of a regular smart cane must be sacrificed. In addition, this device shall be user-friendly by having an extended battery life, being easy to attach, and providing the information required like the proximity of obstacles in an efficient way. This problem should be addressed because there are people out there with no access to such technology because of its price. This project follows the field of assistive technology. Assistive technology can be defined as products or types of equipment that enhance the normal living conditions of a person with a disability. This will be the goal of the project– make mobility easier for blind people.

E8-EE Engineering: Electrical and Mechanical
Two Way Communication Link in the Sub-GHz Band
Carson Mayer
Wyatt Lee

Sub-GHz offers reliable wireless communication with low power consumption . This research contributes to existing
knowledge by demonstrating how Sub-GHz technology can extend communication range and reduce power usage
simultaneously. This research is crucial for search and rescue missions in areas with limited infrastructure, as it enhances the effectiveness of communication over longer distances with low power consumption . The development process involved custom programming of microcontrollers, bespoke wiring of components, and testing using a SDR (software defined radio) to visualize the signal transmission between modules. We anticipated achieving a communication range of at least 100 meters between devices, while maintaining the ability to transmit messages. Our findings indicated that the communication range was limited to approximately 2 meters, owing to the transmit power being only 0.1mW. We concluded that a transmit power of 0.1mW is insufficient for achieving significant communication distances between two devices equipped with standard omnidirectional antennas. The next steps include integrating an amplifier to achieve the desired communication range and designing a unified PCB (printed circuit board) to streamline the circuitry and enhance convenience.

E9-EE Engineering: Electrical and Mechanical
Research And Demonstrate The Effect of Light Intensity(Lux) And Temperature on The Output of A Solar Panel And Explain Software Defined Micro Grid Which Is Resilient And Self Healing against Failure and Temperature /Lux changes
Siddha Karjee

The goal of this project is to study and demonstrate the effect of light intensity (measurement of how much light is falling on a surface, essentially the brightness of a specific area, and is measured in units called “lux”) and temperature on the output of a solar panel and explain software defined micro grid which is resilient and self healing against failure and temperature/lux changes. For demo purpose Arduino platform and accessories will be used .

U10-EE Engineering: Electrical and Mechanical
Making It Accessible: Scientific Hand Models
Gabriella Scottoni
Matthew Rothus

The information on lower income schools are based off of the statistics on non funded schools across America according
to 2013. Our project pushes a concept of accessibility for lower income schools by demonstrating how you can use the
materials available to you in order to create an easy to understand hand model for the classroom. This research will pave a new way of learning for schools unable to fund or be provided with the hands-on activities that are seen to better help kids of all ages process information from the classroom according to classroom surveys. Our methods during this process consisted of calculating the cost of materials, finding the most cost effective and scientifically accurate resources to create an easy to understand hand model as well as calculating the retail cost of providing these objects to classrooms without funding from the school. We also planned our future materials to better improve our design and maximize the price for these low income schools. We expected to conclude our cause was effective and could help a large percentage of educators to go forward with lessons that best suit their students. By using cheap to build models that demonstrated the mobility tests of a hand to prove that the sum of resources does not affect the success of the model . We found that our model was made of materials that were less than our goal of 35$ and ended up being 11$ retail instead. We also found that the hand passed almost all mobility tests except flexion and extension. Angles tests were also conducted and the hand model reached all the following degrees of 90, 60, 45, 30 and 0. In conclusion we gathered that our scientific hand model was able to surpass our expectations on cost effectiveness and did decently on mobility . We were able to create a concept that allows educators to use what is available to them and use it in a creative way in order to create a better learning experience for students which will grow their interest in science or other fields of interest. When continuing with our concept we hope to experiment with more ways to construct a scientific hand with differing materials and how it affects its durability . We also hope to make a version that is more sturdy with a widened range of mobility. Lastly we would like to experiment how higher our budget would compete with the cost effective option and if there would even be a notable difference .

U11-EE Engineering: Electrical and Mechanical
Piezoelectric Energy
Matthew Sawyer

A piezoelectric floor pad that converts footsteps and other energy usually exerted onto the ground into usable energy.

U12-EE Engineering: Electrical and Mechanical
A Modular Ion Wind Cooling System Based on High-Voltage Pulse Generator and Sensors
Alex Hu

This ionic wind project will focus on developing a next-generation cooling device with ionic wind technology. The main
component of this device is a high-voltage emitter and collector electrode system that shoots ions from the emitter electrode to the collector electrode, pushing air molecules alongside. This innovative approach is expected to provide several advantages over traditional fan-based cooling systems. By avoiding mechanical components, the cooling system can be flexible in form, thus giving it more freedom to be applied in compact and irregularly shaped devices. Another feature is that its non-mechanical traits will allow it to operate at approximately 50-60 dB, which is quieter than typical environmental noise levels and causes minimal disruption in settings such as offices or homes without sacrificing its cooling performance . It still achieves comparable wind speeds to conventional fan coolers while consuming less energy as the direct electron movements between electrodes have the potential to push the air molecules more efficiently . Furthermore, the electrodes in the device can be placed flexibly to allow the airflow to enter and exit devices more directly without distortion in the airway, like a typical mechanical fan cooler would. Combining all the features mentioned above, the overall efficiency of the final product is expected to be higher than traditional fan-based coolers. This high efficiency consumes less energy to cool devices and also allows the device to work more efficiently , further reducing the amount of energy wasted through heat. Therefore, considering the vast number of devices that need to be cooled down, it can also significantly reduce carbon emissions. In my plan, the primary application of this technology will target the computer hardware industry to address the increasing need for efficient and silent cooling solutions in high -performance systems. Additionally, the flexible potential of this device gives it broader application opportunities, such as climate control in confined spaces and advanced thermal management in industrial and medical equipment. I will combine theoretical analysis and experimentation to achieve these goals during the product’s design. Prototypes will be designed, constructed, and tested to evaluate performance metrics, including airflow rate, thermal dissipation efficiency, noise levels, and power consumption. The anticipated results will demonstrate that ionic wind cooling devices can be a viable alternative to traditional fan coolers that offer superior adaptability and energy efficiency. Overall, in this project, I want to improve the existing cooling systems, reduce environmental impact, and enhance the functionality of electronic devices by integrating ionic wind technologies.

U7-EE Engineering: Electrical and Mechanical
An Automated Machine Vision Training Machine for Real-Time Sorting Applications
Emma Capaldi
Nina Capaldi

Object recognition in real-time enables transformative applications across various industries. Real-time solutions are crucial in areas such as quality control, security, and inspection. Applications like sorting waste or recycling streams with numerous product types each existing in various states, require recognizing thousands or even tens of thousands of object classes. These systems rely on large datasets of manually labeled images for each class, which is expensive and
time-consuming. To address these challenges, we developed a method for automatically identifying and labeling objects for training recognition systems, along with a machine to implement it. This low-cost device uses a camera to capture images in both UV and visible light spectrums, allowing marked objects to be recognized. The objects are moved along a conveyor, imaged from multiple angles, and shuffled for repeated imaging. The machine then generates training labels, including bounding boxes or object masks, by combining data from both light spectrums. Initial testing shows promising results, and we are working to scale it for more object classes. The impact of labeling errors on recognition accuracy is also demonstrated by introducing controlled variations in the label boxes. In conclusion, our automated labeling system speeds up the process, reduces human error, scales efficiently to large datasets, and opens new possibilities for object recognition applications.

U8-EE Engineering: Electrical and Mechanical
SkyProbe: A Thrust Vectoring UAV for Enhanced Inspection Performance Under Windy Conditions
Jingxuan Zhou

My research introduces the novel thrust vectoring UAV, which has altered the fundamental control logic of a UAV,
specialized for contact-based inspections. This special UAV with its eight arms and four propellers, represents a
revolutionary step in the advancement and innovation of UAVs. Unlike a traditional drone, the thrust vectoring UAV is installed with eight quadrant-shaped arms and four high-torque servos, each connected to two arms. The thrust- vectoring UAV tilts its propellers to generate thrust moving in different directions , with its fuselage maintained level and horizontal to the ground. The design and control logic of the thrust-vectoring UAV overcame the traditional issue of oscillation and performed instant reaction to turbulence under windy conditions, turning the drone into a perfect carrier of inspection tools for stable inspection and detection. The real-world application of the thrust vectoring UAV is profound and abundant , and the drone is able to accomplish countless tasks that traditional drones fail to perform, significantly increasing efficiency during inspections.

U9-EE Engineering: Electrical and Mechanical
Solar Umbrella
Giovanni Santaniello

What is already known about the topic you’re researching? Solar energy has a lot of versatility and is completely green
energy so it does not harm the environment. It can also be used almost anywhere on Earth. How does your research add to what’s already known? What makes it different or special? My research is used to create a power source without having electricity immediately accessible. It’s special because there are not many devices like this . Why is this research important? Shows how solar energy can be used to power devices so that they can still be used in places where electricity is not available, which saves the environment from pollution and devices from being left without power. What were your methods? My phone had to be at 10%. The Solar Umbrella was placed in my backyard or on Hampton beach and my iPhone 12 was plugged into it. Data collection occurred every 10 minutes, where I clicked on the screen to check the power. The amount of power was collected indirectly, however the amount the iPhone was charged over a 10 minute period was the idea for monitoring the output of the umbrella. What did you expect to find? To find a way to charge any device without outside electricity and have it be clean energy to purely power devices. Also to meet my engineering goals. What did you actually find? Be specific with numbers. That my device can charge 0.5625% battery per minute at maximum and 0.05% battery per minute at absolute minimum. What did you conclude from this? It is concluded that the Solar Umbrella is working almost as well as a charger for the iPhone would, with the charging data medians being within 1 of each other(charger at 6 Solar Umbrella at 5). What’s next? The umbrella will be strengthened to hold more solar panels on top, which will result in more surface area for charging. Higher efficiency solar panels or bendable solar panels could be invested in so that more charging could occur with more surface area or higher effectiveness .

V1-EE Engineering: Electrical and Mechanical
Innovative Robotic Rescue System for Avalanche Situations
Yubo Liu

Time is everything in rescue missions. This is especially true in more extreme situations such as avalanches, where
time is a significant factor of survival. As victim are buried up, the limited oxygen in snow is quickly consumed, causing
asphyxia to kick in rapidly after the burial. However, there is not really a great way to ensure the survival of the victim;
conducting the rescue as an amateur or waiting for a rescue team both are not ideal options time wise. However, the time it takes for the rescue team to locate and arrive also great compresses the window of survival for the rescue. Victims have a survival rate of over 91% if rescued within 10 minutes, but this rate drops sharply to 31% after 10 minutes. To address this problem upon avalanche rescue, the paper explores the current problems and solutions upon avalanches, proposing an innovating automatic solution towards the problem of avalanche rescue. the solution is to design a robot based on a wide track chassis, equipped with a UWB positioning beacon to search for trapped people. After reaching the vicinity of the trapped people, the robot can automatically complete the detection task using the detection rod with a force sensor. Finally, after finding the trapped people, air is pumped into the hollow detection rod to extend the rescue time of the trapped people, thereby increasing the probability of successful rescue.

V2-EE Engineering: Electrical and Mechanical
Comparative Analysis of Aerodynamic Performance and Fuel Efficiency in Conventional and Hybrid Wing -Body Commercial
Aircraft Designs
Arin Nazarian

This study examines the aerodynamic efficiency and fuel consumption of two commercial aircraft designs : the traditional wing-fuselage configuration and the hybrid wing-body (HWB) model. Through wind tunnel testing and digital image correlation, aerodynamic forces and airflow behavior were analyzed for both configurations at varying attack angles. The conventional wing-fuselage design, while established and well-understood, suffers from higher parasitic drag at the wing-body junction. In contrast, the HWB design offers reduced drag through smoother airflow and improved load distribution . This study aims to quantify the fuel efficiency benefits of the HWB design, particularly at cruising speeds, supporting its potential as an environmentally sustainable alternative in aviation.

V3-EE Engineering: Electrical and Mechanical
Self Driving car
Daivik Patel

Using a miniature model of a road and traffic features, I will create a self-driving car that drives on a circular track, and try to show how we can add existing features to roads to prepare for future self-driving cars.

V4-EE Engineering: Electrical and Mechanical
AeroClimatic Explorer: An Anti-Collision ROS Drone with SLAM Navigation for Air Quality Monitoring
Yanlin Zheng

This project introduces “AeroClimatic Explorer”, a novel high-performance quadrotor UAV designed for surveillance and environmental monitoring in challenging settings such as chemical plants and caves. Unlike traditional fixed monitoring stations, which are limited by their geographic locations and struggle to provide accurate data in dynamic environments, the “AeroClimatic Explorer” offers superior mobility. Equipped with a PID control algorithm, a GPS unit, a barometer on the flight control panel and a combination of TOF, IMU, and optical flow sensors, the UAV ensures automatic attitude calibration during flight in outdoor environments, significantly enhancing its stability. Additionally, the “AeroClimatic Explorer” integrates a 2D SLAM (Simultaneous Localization and Mapping) algorithm utilizing the Keli LS Lidar, enabling stable flight and data acquisition in indoor environments. Its anti-collision design further protects against three-dimensional collisions, making it a versatile and robust tool for environmental monitoring. Additionally, a real-time camera with a communication range of 20km is mounted on the UAV, providing a monitor to assist the operator in controlling it. Furthermore, to enhance efficiency and compactness, two Printed Circuit Boards (PCBs) have been designed to ensure the stability of the “AeroClimatic Explorer.” One is an SBUS Bridge PCB that controls data in response to control actions, while the other is a multi-channel power supply RF (radio frequency) PCB, which is essential for data transmission and plays a key role in power supply. To ensure accurate and comprehensive monitoring, the “AeroClimatic Explorer” is equipped with both a four-in-one gas detection unit and a seven-in-one air quality monitoring unit. These sensors deliver real-time data on a range of parameters, including humidity, temperature, PM2.5, PM10, CH₂O, CO₂, TVOC, CO, H₂S, O₂, and CH₄, ensuring rich and precise data.

Keywords: Environmental Monitoring，UAV，Sensor Integration，SLAM，Air quality monitoring

V5-EE Engineering: Electrical and Mechanical
ViSA: Virtual Swashplate Aircraft: Design and Implementation of a Novel Control Surface-Less Fixed-Wing Aircraft
Rodolfo Wang
Yuze Cai

One of the paradigms of fixed-wing flight is to go control-surfaceless. Traditionally, fixed-wing aircraft have used the deflection of control surfaces like ailerons, elevators, and rudders, to achieve control in forward flight. These control surfaces add redundant weight due to their complex hinge/hydraulic/servo-mechanisms, creating parasite drag through the discontinuities they create on the wing of a plane, and increasing manufacturing complexities. There have been numerous efforts to replace conventional control surfaces with more efficient alternatives, with an example being the morphing wing, where the entire wing changes shape to mimic the deflection of a control surface. Morphing wings reduce the parasitic drag during cruise flight by eliminating the need for cut-outs on the wing for control surfaces. Still, they add significant material, weight, and maintenance complexities. Another alternative for control-surfaceless flight is pneumatic blowing, where pressurized jets of air are blown from slits on the wing to achieve control. Such a system also adds weight complexities—the need for compressed air storage makes it impossible to be implemented on smaller-scale unmanned aerial vehicles (UAV).

ViSA—Virtual Swashplate Aircraft, is our novel solution to the problem of control-surfaceless forward flight. Inspired by Paulos et al’s (Paulos & Yim, 2015) work on swashplateless and underactuated UAV motor hinges, we devised an underactuated motor drive and propeller mechanism capable of 2D thrust vector control via motor speed modulation alone. The idea is to create a propeller hinge whose tip path plane changes angles in response to cyclic variations in motor speed and acceleration. We’ve verified the control authority of such a mechanism in sustained outdoor fixed-wing flight, paving the way for future micro and full-sized fixed-wing aircraft to achieve forward fixed-wing flight control without any control surfaces. In addition, since our motor speed-modulation scheme requires the connection of multiple sensors and high-speed peripherals which exceeds the computing power of most hobby-level controllers such as the STM32 and ESP32 families, we also developed a novel, cheap, fully operational parallel computing SMD printed-circuit board consisting of three ESP32 microcontrollers which perform flight-critical tasks such as multi-voltage power management, PID, Kalman filter, motor speed feedback/modulation, and SD logging, in tangent. Last but not least, we developed a special low-cost test stand for testing two-dimensional thrust output from our thrust-vectored propeller as well as a blade kinematics and aerodynamic model for predicting thrust-vectoring behavior based on different flight conditions and motor excitations.

C2-EEN Engineering: Environmental
Multi-Environmental Monitor Drone
Jingrui Zhang

The detection of biodiversity in water resources and water quality is an indispensable part of protecting water resources
and the ecological environment. However, in some complex water bodies, the efficiency of collecting biological samples or detecting complex waters is relatively low. This project aimed to design a multi-environmental robot that can be used for marine environmental and ecological detection. At the same time, ensuring the controllable cost and high efficiency is an important factor in promoting its widespread adoption.

F2-EEN Engineering: Environmental
Protecting Native Forest Ecosystems: Using AI-Powered Drones for Identifying and Mapping Invasive Insect Populations
Alex Bace

Invasive insect species pose a significant threat to native forest ecosystems, disrupting biodiversity. In New England, species such as the Spotted Lanternfly (Lycorma delicatula) and the Emerald Ash Borer (Agrilus planipennis) have decimated hardwood forests. Current methods to monitor invasive species rely on individual reports and manual surveys, which are often limited in scope and efficiency.

This project presents an autonomous drone system designed to detect and map insect populations. The drone captures images of insects on tree bark with an onboard camera. These images are processed using a trained AI model and a color -based image filter to identify invasive insect species. The drone’s flight computer logs GPS coordinates for each detected location , enabling the creation of a spatial distribution map of infestations. By providing more detailed insect data, this system enhances prior detection efforts and aids targeted intervention strategies to stop the spread of invasive insect species.

F3-EEN Engineering: Environmental
Cost Effective Atmospheric Water Harvester
Shyam Srinivasan
Swaroop Srinivasan

Developing a cost effective atmospheric water harvester for harvesting water from the atmosphere.

F4-EEN Engineering: Environmental
Does Obsolete Technology Have A Place In Science?
Ethan Parmentier

Using only sensing equipment, software, and computer technology which is considered “obsolete” I attempted to
construct a weather station to collect temperature, humidity, wind speed and direction, and air pressure within scientifically acceptable levels of accuracy. For the purpose of this project, sensing equipment, software, and computer technology is defined as “obsolete” when it is no longer in production by it’s manufacturer, no longer supported by it’s manufacturer, and is no longer in circulation/use. By comparing the range of accuracy of the data produced by the weather station to the range of accuracy it was advertised to have when it was new, and the range of accuracy of modern equivalents, it was determined that the weather station fell well within both the it’s initial range of operational capacity, and acceptable ranges for use in research. The purpose of this project was to answer the question of whether or not age alone can be used to determine the usefulness of scientific equipment, and to demonstrate how though it is ideal to have state-of-the-art equipment to collect scientific data, those who don’t have access to this equipment may still find alternatives in hardware and software which has been replaced or discarded.

F5-EEN Engineering: Environmental
Incorporating Aquaponics into Fish Farms to Increase Efficiency
Emma Lu
Suma Bhiravarasa

With the demand for urban farming increasing, current solutions are inefficient and costly. The issue is that there are many nutrients in fish excrement but it is difficult for farmers to use it. This project aims to utilize aquaponics to create a low-cost, space- efficient and eco-friendly farming system. By using fish excrement to fertilize the plants, farmers will be able to use the nutrients in it and will be able to save money on fertilizer for plants. 

A fish tank filled with fish poop water was attached to a water pump to pump water through a test group of plants (lettuce and cilantro) while a control group was watered with plain tap water. A tube with the water from the tank was attached in a “S shape” pattern to a corrugated plastic sheet directly on top of the plant.

Compared to the control group, the test group grew significantly faster and was more mature for the same amount of time. By day 20 the lettuce control group had 41 sprouts while the lettuce test group had 46 sprouts in the pot. For the cilantro control group there were 32 sprouts while the cilantro test group had 38 sprouts. 

This study showed that the test group that was watered with the fish excrement water grew faster than the control group. This result is important because it means farmers can use this method in a bigger set up to fertilize their plants and they do not need to pay for anything else except set up. Once they have a system set up, they do not need to pay any other money. Something we could think about for the future is that we could try this on a larger scale and see what other costs could come up.

V10-EEN Engineering: Environmental
Which Biodegradable Hydrogel Conserves the Most Water for Farming?
Jalani Lopez
Kinsey McClelland

The purpose of this experiment was to create and test different types of hydrogels , to determine which type retained the
most water. Climate change is an increasing issue worldwide, and one of its side effects is drought, negatively impacting our agriculture, thus leading to increased world hunger. By creating an agent of water conservation, it can help to protect against droughts. The creation of hydrogels was done by combining different organic powders , citric acid, and water. The resulting hydrogel was then mixed into corresponding pots of soil and a set amount of water was added. These pots were then placed in the testing environment where their weight was measured each day for 15 days. Results showed that Carboxyethyl Cellulose hydrogel performed the best, retaining the greatest amount of water. These findings could be beneficial to the workers in agriculture if they choose to implement hydrogels into their farming system.

V11-EEN Engineering: Environmental
Solar Powered Water Purification Device Utilizing Membrane Distillation Methods
Maria Mishechkina

About 2.2 billion people worldwide do not have access to safely managed drinking water services (WHO, 2019). Many of
the earth’s freshwater sources are full of chemicals, heavy metals, and microorganisms due to improper disposal of various wastes (Khan et al., 2022). Consumption of contaminated water results in waterborne diseases which facilitate about 1 million deaths annually (WHO, 2022). Current methods of water purification only focus on the removal of a specific contaminant and require large amounts of electricity to function, making them unaffordable and ineffective in full decontamination. Membrane distillation is the process of evaporating contaminated water through a porous membrane and condensing purified water. Preliminary methodology includes testing the rate of vapor permeation of various membranes, utilizing spectrophotometry and methylene blue to conduct a proof-of- concept that determines the effectiveness of the membrane distillation process, and prototyping condensing units. Upon prototyping, a final device design was created and built using Computer Aided Design and purchased components. Water samples were collected from Institute Pond and Northborough reservoir to test levels of contamination before and after purification. Focused chemical and microorganism testing occurred as well to determine the device’s effectiveness with higher levels of contamination than pond water . A solar panel and battery were acquired to power the device. Results show that the device achieved complete or mostly complete removal of chemicals, heavy metals, and bacteria. All results after purification fit safe drinking water standards of the EPA, WHO, or WQA. Future work includes increasing the scale of clean water production, further iterations on materials and design, and experimenting on a broader range of contaminants.

V12-EEN Engineering: Environmental
Algae-Pumpkin Bioplastic: A Sustainable Alternative to Conventional Plastics
Emily King

Plastic pollution is one of the most pressing environmental challenges of the modern era. Each year, over 460 million metric tons of plastic are produced globally, with the majority ending up in landfills, oceans, and other ecosystems. Traditional plastics, derived from fossil fuels, take hundreds of years to decompose and contribute significantly to environmental degradation. Despite growing awareness, plastic recycling rates remain critically low, with only 5-6% of plastic waste being recycled annually in the United States. Given the urgent need for sustainable alternatives, this project explores the development of an algae-based bioplastic reinforced with pumpkin seed fibers as a biodegradable and environmentally friendly substitute for conventional plastics.

To evaluate the feasibility of this bioplastic, a mini spoon prototype was created and subjected to a series of mechanical and biodegradability tests. The bioplastic was formulated by combining 1/4 teaspoon of algae powder, 1/4 teaspoon of sorbitol, 1/4 teaspoon of gelatin, and 1/4 cup of glycerol solution with 1/4 cup of water. The mixture was heated until a gel-like consistency formed, and after cooling, 9 grams of pumpkin seed powder were incorporated to enhance strength. The material was then molded into spoon shapes and left to dry overnight.

The durability of the bioplastic spoons was tested against conventional plastic spoons through multiple experiments. In the Stress and Strain Test, a force sensor measured how much pressure each spoon could withstand before bending or breaking. The bioplastic spoon required an average force of 23.39 N, whereas the conventional plastic spoon required 16.64 N, indicating the bioplastic’s superior strength. In the Drop Test, both spoons were dropped from a height of 1 meter (5 feet) onto a hard surface multiple times, and neither showed visible damage. The Scratch Test involved using a sharp object to test surface resistance , where the bioplastic proved more resilient, as the conventional plastic spoon cracked under pressure. Additionally, in the Repeated Bending Test, the bioplastic withstood 50 bends before cracking, whereas the plastic spoon broke at 29 bends.

To assess biodegradability, the bioplastic and conventional plastic samples were buried in a compost bin maintained at 50-60ÅãC (120-140ÅãF) and monitored for decomposition. The bioplastic spoon broke down within one week, displaying mold growth and structural disintegration, while the conventional plastic spoon remained unchanged even after several weeks.

The results of this experiment demonstrate that algae-based bioplastics
reinforced with pumpkin seed fibers can provide a durable yet biodegradable alternative to traditional plastics. This research highlights the potential for bioplastics to reduce plastic waste and promote sustainability in everyday products such as utensils and packaging. Future research will focus on refining the material’s flexibility and water resistance , as well as expanding its applications to other consumer goods. With further advancements, biodegradable bioplastics could play a significant role in mitigating plastic pollution and creating a more sustainable future.

V6-EEN Engineering: Environmental
Harvesting Drinking Water from Fruits and Vegetables
Siddharth Peruri

Many fruits and vegetables are primarily composed of water. I want to test if I can separate drinking water from fruits and vegetables. Tomatoes and Grapes contain 90% water in them. Food processing plants separate water from tomato juice to make paste used in ketchups and sauces. Similarly, wineries process grape juice to make wine by distilling water out after fermentation to increase alcohol content. If this works, I believe this can be part of a solution to the drinking water shortage in some places.

V7-EEN Engineering: Environmental
AeroRain Generator
Ishan Kasam
Rishabh Mathukiya

Vertical Axis Wind Turbines (VAWTs) play a vital role in optimizing wind energy to mitigate climate change. Especially while utilizing piezoelectricity, the energy is enhanced and the overall system has the potential to enhance renewable energy while assisting with global energy demands. As climate situations worsen, there is a growing need to optimize wind turbines, particularly in urban and residential environments where horizontal axis wind turbines face limitations due to noise pollution, maintenance challenges, and spatial constraints. VAWTs are more efficient in these settings due to their omnidirectional nature , ability to harness wind from all angles, and compact design. Integrating piezoelectric transducers into wind turbine design allows for converting mechanical vibrations (from rain or wind) into additional electrical energy, maximizing output from even low-speed winds, which is especially relevant in densely populated areas, homes, or regions with fluctuating wind conditions. Research into VAWTs with piezoelectric elements could also lead to decentralized energy production, reducing dependence on fossil fuels and advancing sustainable development goals. This project is a scaled-down model of Darrieus H- rotor VAWT outfitted with piezoelectric transducers, focusing on harnessing vortex-induced vibrations (VIV) and synergistic turbine arrangements to improve efficiency of renewable energy in urban settings (compact environments). Additionally, by placing multiple VAWTs strategically in an array so that the wake interferences from upstream turbines amplify rotational speeds and vibration, the model showed up to a 15–30% boost in wind and piezoelectric output, supplementing primary wind energy capture. Overall, this work underscores the potential of small-scale hybrid systems combining aerodynamic synergy and multimodal energy harvesting, offering a compact solution for clean energy, especially in compact, urban environments.

V8-EEN Engineering: Environmental
Low-Frequency Energy Harvesting Device Based on a Winch-Driven Structure
Charles He

Marine environments face increasing pollution challenges from floating debris and microplastics, prompting the need for self-sustaining solutions that can both monitor and protect ocean ecosystems. In this study, we introduce a dual-purpose system designed to address this issue: a tilting-induced wave energy converter that powers a real-time trash detection module. By leveraging low-frequency, low-amplitude waves, the device aims to generate electricity while simultaneously enabling on-board image recognition for marine trash identification. The power generation component uses a sliding-weight mechanism and helical pulley system to capture wave-induced tilts (2Åã–5Åã). Under controlled laboratory conditions simulating slow wave frequencies, the system achieved a peak voltage output of 1.6V at a 4Åã tilt, demonstrating moderate efficiency on a small scale. Reinforced structural elements (track redesign, one-way spinning rods) reduced friction and instability, improving energy transfer. In parallel, a K210-based AI module was trained to detect surface trash with promising classification accuracy, showcasing the potential for integrated, self-powered monitoring. Although further testing is required to ensure consistent performance in real-world marine conditions, initial results suggest that scaling up the wave converter and refining the AI model could create a truly autonomous platform for environmental protection—offering both clean energy extraction and proactive pollution detection for sustainable ocean stewardship.

V9-EEN Engineering: Environmental
Benthic Microbial Fuel Cells: A Novel Power Source
Maverick Pil

This project aimed to design a benthic microbial fuel cell (BMFC) capable of generating power over an extended period. It was hypothesized that the cell would produce power near the lower end of the previously researched spectrum (30-50
mW/m.), given its design, which lacked pumping, added substrate, or high-surface-area electrodes. The prototype was
constructed by drilling a hole in the container, installing a check valve with the flow direction facing outward, and sealing the hole with silicone adhesive to prevent leaks. Electrodes were connected to 18 AWG wire using bronze screws and ring lugs. The BMFC was submerged in a simulated environment of seawater and sediment. Current was tested in 10 mA increments until the maximum current output was reached. This maximum current was tested, with the voltage measured over time. The power produced ranged from 17.4 mW to 49.6 mW, with the power produced per square meter ranging from 145 mW/m. to 321mW/m.. The decline in power during the constant discharge and recovery during staggered discharge cycles were consistent with previous research. While the power production was within the expected range, it was closer to the higher end, suggesting the potential for higher energy yields. Factors such as organic matter decay, corrosion of the cathode, and the small experimental setup likely influenced the power produced. This study demonstrates the potential of BMFCs for marine sensor networks, though further optimization is needed to improve power stability and performance in real-world applications.

W1-EEN Engineering: Environmental
Biofilm-Based Plastic Bioremediation in Water Systems
Benson Wong
Martin Wong

Bacterial biofilms are essentially groups of bacterianthat stick together as a layer onto surfaces. Their sticking ability is attributed to their production of extracellular polymers (EPS), which also form matrices that hold the microorganisms. Such biofilm matrices are responsible for the encapsulation of foreign materials – including microplastics. This paper explores the use of the non-pathogenic planktonic bacteria, B. subtilis, in large-scale water filtration, ultimately in hopes of supporting the hypothesis: If B. subtilis biofilms are utilized in a layered water filtration system, then water flowing through systems with biofilm will contain less microplastics than without. Furthermore, this paper also explores the efficacy of known water filtration materials, such as activated carbon and sari cloth, as well as in conjunction with biofilms. An experiment was conducted using a fluorometer in order to calculate the concentration of microplastics through dye staining. Based on the results, not only were the efficacy of known filtration materials confirmed , including activated carbon for microplastic filtration and sari cloths for bacterial filtration, but the effectiveness of biofilms as a microplastic filtration material was also supported when paired with other materials, ultimately supporting the hypothesis.

Key terms: Extracellular polymeric substances (EPS), matrix, B. subtilis, water filtration

W2-EEN Engineering: Environmental
Leveraging Drone-Based Spectral Analysis of Plant Moisture to Combat Wildfire Risks
Junwon Park

Global warming increases dried vegetation worldwide, intensifying wildfires as the annual average burned area in 2020-23 was thrice that of 2010. Identifying areas with concentrated dried vegetation, which ignite and spread wildfires, is essential in preventing wildfires. Plant spectral analysis could help with this task. Fluorescence, emitted from plants’ leaves, varies according to the leaf’s moisture level and signals dryness faster than visible light does. This experiment explored the possibility of using a portable infrared sensor to distinguish between healthy leaves, dried leaves, and fallen leaves using fluorescence. Blue Light Emitting Diodes were used as excitation light, and the infrared sensor measured intensities of six different wavelengths based on fluorescence . Compared to fallen and dried leaves, healthy leaves emitted more red fluorescence, successfully distinguishing the leaves. Finally, the sensor system was placed on a drone to detect a large area of leaves via wireless communication and effectively distinguished healthy , dry, and fallen
leaves. AI (artificial intelligence) was used to measure Dried Point per area, a metric quantifying the likelihood of ignition of leaves and dryness for a given area, with different variables such as leaf types, the number of dried leaves, and leaf arrangements. Using the random forest regressor model, AI successfully predicted Dried Point per area with R-squared value of 0.99953 and the highest feature importance in the U spectrum (760 nm). Once areas with dried vegetation are identified using Dried Point per area, water could be distributed, or dried vegetation could be removed via controlled burning to prevent wildfires. 

W3-EEN Engineering: Environmental
AI-Driven Drone Surface Damage Detection for Wind Turbines
Kathy Lin

Wind turbines play a crucial role in renewable energy production, generating 10% of the energy in the United States. However, surface damage can compromise their efficiency and introduce safety hazards. With over 700,000 wind turbines worldwide, some towering over 900 feet, blade failures present a significant challenge, with more than 3,800 incidents reported each year. This project proposes the implementation of autonomous drone inspections equipped with computer vision and live video feeds to capture high-resolution images of wind turbines. These images are analyzed using advanced deep learning models, including Convolutional Neural Networks (CNN) such as EfficientNet and InceptionNet, to automate the detection of surface damage. The study fine- tunes the EfficientNet and InceptionNet models using over 13,000 drone-captured images of wind turbines to enhance model accuracy. The proposed method achieves an accuracy of 84% and a precision of 90%, surpassing human detection rates. By enabling more frequent and precise inspections, this system extends the operational lifespan of wind turbines while significantly reducing the risk of malfunctions and costly repairs.

W4-EEN Engineering: Environmental
AI Meets Geothermal: Cost and Location Analysis for Energy Efficiency
Evan Channa

This project introduces a novel neural network-based application designed to identify and rank potential geothermal sites using a combination of geological, thermal, and spatial data. The model integrates heat low data, borehole temperature measurements, thermal conductivity values, and geochemical risk factors to predict geothermal viability with over 85% accuracy. By incorporating convolutional layers for spatial heat map analysis and dense layers for tabular data processing, the system provides a comprehensive assessment of geothermal potential. The application is designed to process 100 km. of data in under five minutes, offering a user-friendly interface for data input and visualization. Validated against known geothermal sites in Massachusetts, this tool demonstrates significant improvements over traditional assessment methods, making it a valuable resource for advancing geothermal energy adoption.

W5-EEN Engineering: Environmental
Using a Bio-Reactor to Reduce Nitrogen and Phosphorus Pools for Eutrophication
Aleksander Mohmand-Borkowski
Grant Silver
Oliver Affonso

Increased nutrient pollution of marine environments has been fueling eutrophic waters
all around the world for decades now, resulting in harmful algae blooms, massive fish kills, and threats to all sorts of life. Hence, research such as ours, aimed at reducing nutrient pools in water, may potentially alleviate Eutrophication. This would be achieved using a chlorella vulgaris photobioreactor.

W6-EEN Engineering: Environmental
Making Mobile Homes Amphibious: A Flood Avoidance System
Juhi Kundu

Flooding presents a major threat to mobile homes, which are vulnerable due to structural limitations and lack of flood
protection systems. As climate change increases the frequency and intensity of floods, affordable solutions are urgently
needed to protect mobile home communities. This study investigated whether synchronized-length extendable/retractable tension cables, tethered to the ground, could allow a mobile home sitting atop a floating foundation of inflatables to rise and fall with floodwaters while maintaining lateral stability. The research question centered on how much these cables, which adjust equally in length, could constrain the home from drifting laterally as it floats vertically. The hypothesis proposed that the buoyant force of the water would keep the cables taut, stabilizing the home enough laterally to prevent collisions with neighboring structures and to bring it back to its original position after the flood. A scaled-down prototype was built and tested in a controlled flood environment simulating a 6m (20 ft) flood with various wave intensities. The results showed that lateral displacement was effectively minimized, with the maximum recorded at 126.77 cm (4.16 ft) in real-world scale during the most aggressive wave conditions. Once the water receded, the mobile home returned to within 18.45 cm (7.26 in) of its original position, also in real-world scale. These findings suggest that such a system could offer a practical , cost-effective solution for retrofitting mobile homes without major structural modifications, making it an accessible approach for protecting vulnerable communities from the increasing risks of climate-induced flooding.

W7-EEN Engineering: Environmental
Scientific Machine Learning and Photodynamic Therapy for the Prevention , Anticipatation and Mitagtion of Toxic Algae Blooms
Kaizar Rangwala

Harmful algal blooms (HABs) are a significant global environmental issue, impacting marine ecosystems, economies, and human health by depleting oxygen in water bodies and causing biodiversity loss. The economic cost of HABs is estimated at approximately $82 million annually, along with posing health risks through toxin-contaminated water and seafood. Traditional methods for combating HABs often harm non-target organisms and ecosystems.This project offers a solution combining photodynamic therapy (PDT) with artificial intelligence (AI). PDT, a cancer treatment used to selectively target tumor cells, is adapted here to precisely neutralize harmful algae without collateral damage to non-algal cells. To ensure that the delivery of PDT against a HAB is not excessive or insufficient,  AI locates and assesses the magnitude of blooms using a visual feed from a camera, with an integrated algorithm using this assessment to dynamically adjust PDT treatment to match the bloom’s size . To ensure adequate performance, a prototype was considered successful if the AI model had a 90% accuracy in algal bloom detection from the camera feed and if its estimate of the size of the bloom and appropriate intensity of PDT for that size were within 10% of the actual necessary amount (loss). To measure the quantity of algae alive before and after PDT (hence measuring the treatment’s effectiveness), spectrophotometry was employed, with an 85% reduction in absorbance after treatment considered successful. The final prototype was 97.49% accurate, with a loss of 0.051%, and an 85.07% decrease in absorbance.

F8-MA Mathematics and Computational Modeling
The Blind Timekeeper: The Synchronization of Coupled Oscillators
Shrinivasa Makaram

Our body is made up of intricate dynamics of organs and cells. Among these, the heart and the brain are some of the most essential, with the brain directly controlling the body and the heart pumping to keep it alive. Both of these organs rely on synchronization to function.  A lack of synchronization in the heart pacemaker cells can lead to diseases such as arrhythmia. On the other hand, unwanted synchronization in sections of the brain can lead to diseases like epilepsy. Moreover such synchronization can also be seen in other aspects of nature, for example the synchronized firing of fireflies in South- East Asia and the chirping of crickets all around the world. However, these examples of synchronization are not just limited to nature. Examples of spontaneous synchronization can be seen in traffic patterns and computer networks .

Such natural self synchronization can be modeled as a group of pulse coupled oscillators. In this project a simulation of pulse coupled oscillators with several variables and a spatially local connection was developed. Utilizing models such as 4-Way 2d connection, 8-Way 2d connection, 4-Way Torus connection, 8-Way Torus connection, and a Global connection, I show how the spatial orientation affects the synchronization . I apply analytical techniques, such as calculating the number of oscillators going above the threshold at a given time and using windowing techniques, to perform statistical analysis (mean and standard deviation) on each time period. Globally connected models always synchronize. This observation lines up with what was proven in previous published results. For the spatially oriented simulations, synchronization was not always seen. A change in certain parameters showed greater tendency towards synchronization. 

This work has pushed the boundaries of what is known about spatially connected self-synchronizing oscillators. This has
immense applications in the medical field, other biological applications and in the reduction of traffic congestion in large communication networks. Further work in this field may find cures for epilepsy and cardiac arrhythmia.

F9-MA Mathematics and Computational Modeling
Estimating Periodic Solutions within Pendulum-like Chaotic Dynamical Systems
Zachary Medjamia

This study explores the estimation of periodic solutions within chaotic pendulum-like dynamical systems by analyzing
the Lyapunov characteristic exponent (LCE) of various pendulum models, including double, triple, and sprung pendulums. Numerical simulations will be used to examine small perturbations in initial conditions and assess their impact on trajectory divergence. The equations of motion for each system will be derived using Lagrangian mechanics and solved via fourth-order Runge–Kutta numerical integration. Python will be used for visualization, while C will handle large-scale computations. The Lyapunov exponent will be calculated across the phase space with increasing resolution to identify regions of minimized chaos. This approach aims to detect conditions under which periodic behavior emerges, contributing to a deeper understanding of stability in non-linear and chaotic dynamical systems.

S6-MA Mathematics and Computational Modeling
The Likelihood of Developing Huntington’s Disease Across Human Subpopulations
Sara Leonard

Huntington’s disease is a dominant genetically inherited neurodegenerative disorder affecting the central nervous system; indicated by continuous psychiatric disturbance, loss of coordination with motor functions, and cognitive deficiency. It is caused by a mutation of the Huntingtin (HTT) gene located on chromosome 4. This mutation involves the expansion of repeated CAG trinucleotide sequence at least 40 times, which results in an abnormally long polyglutamine tract on the N terminus of the HTT protein . Huntington’s is fatal roughly 15-20 years after its onset, and currently there is no known cure. In this analysis, public data from the 1000 genomes project was utilized to compare the number of CAG repeats among human subpopulations. Through coding, the low coverage, alignment, binary angular measurement files from each participant in the phase 3 of 1000 genomes project is analyzed for the HTT gene, counting the number of CAG repeats that occur on both sets of chromosome 4.  First the number of CAG repeats on both sets of chromosome 4 for each individual was averaged together, and then the average number of repeats per sub-population was found. Each number of repeats per subpopulation was then compared, to identify the subpopulation carrying the highest number of repeats and thus determine that those from Toscani, Italy have the highest likelihood of developing Huntington’s Disease.

S7-MA Mathematics and Computational Modeling
Circadian Rest-Activity Rhythms and Cognitive Resilience in Prodromal Alzheimer’s Disease
Kevin Wang

Alzheimer’s disease (AD) is the most common form of dementia, accounting for around 60%-80% of dementia cases in the US alone. AD can be divided up into preclinical AD, where cognitive symptoms are largely unnoticeable but pathology has started to accumulate; prodromal AD, also known as mild cognitive impairment (MCI), where cognition is impacted but not significantly; and finally diagnosis of AD, which can further be divided into mild, moderate, and severe stages. Currently, in the US, 5 million older adults above age 65 are living with AD, and another 7 million are living with MCI1. Circadian rhythms represent the near 24-hour internal clocks in the brain that regulate various biological and physiological processes, such as sleep, alertness, and cognitive performance2. Disruptions in circadian rhythms are frequently observed in patients with Alzheimer’s disease (AD) even during its preclinical stage, and especially in its prodromal stage/MCI3. Rest-activities rhythms are an extension of circadian rhythms and can be measured using non-invasive measures such as motor actigraphy. Previous research in cognitively healthy older adults has shown that disruptions in rest-activity rhythms are associated with an increased risk of developing AD3, 5. Given that aging is the greatest risk factor for AD, understanding how circadian disruptions contribute to cognitive decline in elderly populations is crucial. Cognitive resilience refers to an individual’s capacity to maintain their cognitive function against factors that would typically promote cognitive decline, such as AD pathology4. It remains unclear whether better-maintained rest-activity rhythms represent a resilience phenotype that protects against cognitive decline or dementia in people with prodromal AD3. Identifying potential proxies of cognitive resilience is essential for predicting the progression and severity of AD, which may inform lifestyle-based interventions, such as structured sleep schedules and light therapy, to promote healthy aging and cognitive function.

S8-MA Mathematics and Computational Modeling
Energy-Efficient Routing to Pick Up Heavy Objects
Alex Kuriakose
Sanjay Iyer

This paper addresses the challenge of determining the most energy- efficient method for collecting heavy objects distributed across the coordinate plane. This problem is significant due to its real-world applicability, where the task involves not only distance but also the accumulated burden of the collected objects. Our research focuses on a scenario where a carrier collects objects placed at specific coordinates in the first quadrant of a Cartesian plane, starting and ending at the origin. The carrier’s objective is to gather all the objects with minimal energy expenditure, which is a function of the distance traveled and the weight added by each ball. We developed a computer program that calculates the optimal route by analyzing coordinate points provided by the user. This program identifies the most effective path that minimizes the total cost, factoring in both the traveled distance and cumulative weight of the basket. We analyzed the data from a simulation of 5000 random problems and found the following: our algorithm finds a more cost-efficient path for 44.61% of problems and in those situations, our algorithm saves on average 6.59% of the total cost.

Y10-MA Mathematics and Computational Modeling
Social Credit: Leveraging Artificial Intelligence (AI) and Social Media for Inclusive Lending
Ryan Denny

The Data-Augmented Credit Scoring (DACS) methodology introduces an AI-driven approach to creditworthiness
assessment, addressing the exclusion of underbanked populations from traditional credit systems. This proof-of-concept project leverages alternative data sources, including social media activity, to predict financial behavior and broaden participation in global capital markets. The DACS model combines synthetic demographic, psychometric (Likert) survey responses, and AI-generated social media media data to create an alternative credit scoring system. This project developed 119 synthetic personas to simulate real-world financial behavior and creditworthiness. Three independent scoring mechanisms were applied: Demographic Score (92% accuracy), Psychometric Survey Score (85% accuracy), and Social Media Score (87% accuracy). The final DACS score, computed as a weighted sum of these components (30% + 40% + 30%), demonstrated a strong predictive model for credit worthiness when validated against synthetic FICO scores. Overall accuracy in classifying individuals into creditworthiness categories was approximately 88%. In addition to ChatGPT and Gemini, six other Large Language Models (LLMs) were used: Claude, DeepSeek, LLaMA, Qwen, Mistral, and Zephyr. As part of this project, the following statistical measures were used: Mean, Median, Standard Deviation, Frequency Percentages, Confusion Matrix, One-Sample T-Test, P-Value and Kullback-Leibler (KL) Divergence. This research demonstrates that a multi-faceted credit scoring approach incorporating AI-driven behavioral analysis and psychometric indicators can enhance financial inclusion for individuals lacking traditional credit history. While social media behavioral analysis can serve as a viable supplementary credit assessment tool, demographic data remains the strongest predictor followed by behavioral analysis. Future work will focus on expanding the dataset, optimizing weight distribution, and addressing ethical considerations related to AI-driven credit scoring.

Y11-MA Mathematics and Computational Modeling
Novel Computational and Machine Learning Approaches Shed Light on Parkinson’s Disease Heterogeneity and Subtypes
Ariv Vaidya

Key Words: neurodegenerative, heterogeneity, PD subtypes, predictive analysis, extreme phenotypic analysis, machine learning, personalized clinical trials, multi-omics 

Parkinson’s Disease (PD) is a neurodegenerative disorder characterized by immense heterogeneity. Due to the variation in disease severity and disease progression and a lack of understanding on accurate PD modeling, this study aimed at defining PD subtypes based on their disease progression and conducting a predictive analysis of these subtypes using baseline phenotypes. Due to the practical constraint of sample size, extreme phenotypic analysis was conducted to create the subtypes to boost the statistical power and simulate a larger sample size and machine learning (ML) approaches were utilized to conduct the predictive analysis. The study uncovered that subtypes could accurately create distinctions amongst a patient population and the ML model uncovered an accuracy of 78% when predicting the subtype from baseline characteristics. Furthermore, key clinical indicators were further validated such as age, Dyskinesias, and the overall severity index created from the Pairwise analysis as being the top endpoints from the ML model. This creates the possibility for personalized clinical trials where each PD subtype could receive different treatment and multi-omics data analysis using these subtypes could uncover biomarkers of PD disease progression.

Y12-MA Mathematics and Computational Modeling
Causal Discovery and Supervised Learning for Identifying Cancer Risk Factors
Shengtao Ding

This paper aims to identify the essential factors that lead to cancer incidence using the PC (Peter & Clark)
algorithm (Graph-based algorithm), Least squares regression, and Neural networks. The data used for this analysis was
collected by the CDC (Centers for Disease Control and Prevention) over the past decade (2014-2023), and thirteen variables were selected for analysis, including BMI, exercise, asthma, and diabetes. The PC algorithm generates a directed acyclic graph (DAG) to represent the causal relationships, while least squares regression gives the weight of each factor. Neural networks, in conjunction with counterfactual thinking, serve as a tool to describe the relationships and effects and to predict the possible outcomes of the system in hypothetical settings.

The data analysis presents two types of visualizations: A 10-year causal graph that shows the general picture of the data and annual weight graphs that depict the standardized (Z-score) fluctuations in the relationships between the factors. Through the analysis of the changes, the study determines the shifts that have occurred in the relationships between these variables, and this is done while considering the influence of external events such as the COVID-19 pandemic. The results of the study show that there is an opportunity to target specific factors in order to decrease the incidence of cancer, and the findings can provide policymakers with information on health disparities and future trends to guide health policy decisions.

Y4-MA Mathematics and Computational Modeling
Using Math Modeling to Identify and Correct Gerrymandering
Ravena Arun

There are several gerrymandering-preventative legislations focused on maintaining district compactness and prohibiting discriminatory redistricting plans, outlined in Section 2 of the Voting Rights Act (Section 2 of the Voting Rights Act, 2015). However, these laws are ultimately deemed as ineffective because gerrymandering cannot be objectively measured or solved . Current models aimed at outputting optimized non-gerrymandered maps fail to account for all factors. These existing models are accurate but solely consider geographical size, compactness, and preserving old district cores. They leave out one crucial factor that is one of the main causes of the continued usage of the archaic electoral college system: just representation of minorities. Thus, the project goal is to develop a mathematical model to objectively and accurately identify and correct gerrymandering, specifically of minorities. This research could significantly change the way we redistrict and provide objective legal evidence to prosecute  gerrymandering to make every vote count the same and ensure political justice.

Y5-MA Mathematics and Computational Modeling
Using Machine Learning to Predict Chronic Kidney Disease: Analyzing Critical Indicators for Early Diagnosis
Amogh Hiremath

Chronic Kidney Disease is a global health problem that affects millions of people, and often progresses silently before entering more advanced stages. Early detection of Chronic Kidney Disease is especially important for this reason. This project aims to develop predictive models for diagnosing Chronic Kidney Disease, using machine learning algorithms on clinical data. Including 24 features, such as albumin, serum creatinine, hemoglobin, packed cell volume, and red blood cell count, this dataset provided the clinical data to train these machine learning algorithms. By training and evaluating multiple machine learning models, including logistic regression, Random Forest Classifier, and Gradient Boosting Classifier, the experiment aims to identify the most effective algorithms for accurate prediction of Chronic Kidney Disease. Furthermore, a feature importance analysis highlights the main clinical factors that contribute to Chronic Kidney Disease, offering very valuable insight into Chronic Kidney Disease’s underlying mechanisms. In addition, model evaluation metrics such as accuracy, precision, recall, and F1-Score were used to compare and evaluate model performance. The models were compared to optimize predictive accuracy. This project shows how machine learning can help improve the early detection of Chronic Kidney Disease, and the results contribute to addressing the global need for efficient, non invasive diagnostic tools, which highlight the importance of personalized medicine in combating chronic diseases.

Y6-MA Mathematics and Computational Modeling
Manufacturing Capacity: A Math Modeling Analysis of PresMet Corporation
Abigail Lei
Xin Zheng

PresMet Corporation faces capacity challenges when manufacturing metal parts for its main three patron companies Walabe , Fiord, and Costly. Due to spikes in demand, they are unable to produce enough parts, resulting in shortages. By employing mathematical modeling, it was possible to analyze how PresMet’s current production schedule and machine performance align with fluctuations in weekly demand . Gap-filling techniques were applied to estimate missing values in the dataset through averages and regression lines, allowing for a more comprehensive analysis.

PresMet’s current production capacity was calculated for standard and overtime workweeks and compared with actual
demand, revealing that regardless of schedule, PresMet could not produce enough parts. Three potential solutions were
evaluated: the sole acquisition of new honing machines, an overtime schedule, and the modified overtime schedule. While adding new machinery could address capacity issues, the high cost and installation time present major challenges. As such, a combination of new machines and a new production procedure are needed to fulfill demand. Alternatively, the two overtime procedures combined with purchasing new machines offer a cost -effective approach to managing fluctuations, though they may still struggle with extreme spikes.

Overall, this project demonstrates how mathematical modeling can be used to identify and address real-world manufacturing challenges, providing valuable insights for optimizing production processes.

Y7-MA Mathematics and Computational Modeling
Analyzing the Local Approximation of Trigonometric Functions Using Chebyshev and Taylor Polynomials
Jamison Ballou

Trigonometric functions are very useful for modeling natural processes , like the swinging of a pendulum, planetary
motion, and sound waves; however, they are not nearly as easy to work with algebraically as polynomial functions. There are many ways of approximating functions, but the two most versatile are the Taylor and Chebyshev polynomials . The goal of this study is to determine which method of approximation best emulates the original trigonometric function over the range [-1, 1]. This was determined through calculating several metrics of error as well as qualitative observations of the data. Python was used to generate the approximations, and they were logged into a spreadsheet. Then, JavaScript was used to calculate error and create visualizations so further observations could be made. The results reveal that in general, the error between the Chebyshev approximation and original function was lower than the error for the Taylor approximation . Therefore, over the interval examined, the Chebyshev polynomial is more effective at approximating trigonometric functions .

Y8-MA Mathematics and Computational Modeling
Multivariate Analysis and Prediction of the Risk Factors for Concussion in High School Students
Leo Lim

Concussion, also known as mild Traumatic Brain Injury, is prevalent among athletes, with short and long-term consequences impacting neurological health and athletic performance. Current diagnostic methods rely on subjective measurements, leading to high rates of undiagnosed concussions, and advanced methods that are highly inefficient and expensive . This study thus aims to develop a multivariate logistic regression model to enhance early detection and risk assessment of adolescent concussions. Using data from federal repositories, such as FITBIR and Normative Athlete Data, 19 risk factors were identified using a logistic regression model through the programming language R. Then, with the identified factors, a logistic regression model was designed, trained, and tested using a 7:3 train-test ratio. To validate the model, a calibration plot and a Receiver Operating Characteristic (ROC) curve were used. The Area Under the Curve (AUC) of 0.7045 indicates acceptable predictive performance. As the final product, the model was implemented into a ShinyApp-based website, allowing athletes to readily input data and receive concussion probability estimates.

C12-PH Physics
The Effect of Atmospheric Conditions on Plasma
Robert Simpkins

Last year, I had a project in which I tested how uncharged and charged dust affect plasma generation in order to test my
hypothesis that suspended charged dust would allow for easier creation of plasma. Variability between testing days was
observed leading to this year’s experiment. In the current experiment, different atmospheric conditions were tested to
determine their effect on plasma generation. I also continued to refine my dust trials to continue to look at how dust affects plasma creation. This was actually the original purpose of the experiment but due to issues with changing the dust insertion method, as well as time and material constraints the main focus of this experiment shifted to looking at the effects of atmospheric conditions. To test the various atmospheric variabilities plasma was repetitively generated at various pressures on different days and the ignition voltages were recorded . In addition, the humidity, temperature, and atmospheric pressure were measured each day. Then the graphs of the ignition voltages each day were compared to the measured variables to see if there were any correlations. Of the variables tested, none seem to have correlations, however humidity may have caused a jump in ignition voltage on day 4. If this is true (more tests are needed to confirm) then humidity simply requires a dramatic increase to have any effect, and the effect was very small. On top of the tested variables, some other observations were noted such as a slow leak, chemical reactions inside the chamber which created new compounds, and a change in initial voltage required after initial ignition. None of these observations however explain the differences between days so I hypothesize that there is a variable that is unaccounted for by this experiment. Finally, dust trials were attempted like originally planned. Last year there had been issues with the dust insertion method as varying levels made it into the chamber and other than a noted difference in the amount of voltages required for ignition there was little proof that dust was inserted and suspended inside of the vacuum chamber. This year a new way of dust insertion was created where a pvc pipe was magnetically attached to the side of the vacuum chamber’s bell jar and a magnetic door on the bottom was opened after a vacuum was pulled to release the dust. This new method took time to create and used up many materials for each trial so refinement and repetition beyond the initial successes was not feasible. The inserted dust did not appear to be suspended but rather simply fell to the bottom and there was no noticeable change in ignition voltage leading to the belief that the dust insertion was ineffective and no dust was truly suspended .

D12-PH Physics
A Quantifiable Analysis of Launch Angle to Distance in Different Heights of Shot Put Throwers. (And Why It Matters)
Ari Levine
Holly Riddell
Pia Obrador

 

F10-PH Physics
Entanglement Sudden Death in Interacting Quantum Circuits
Samyak Jain

The theoretical aspects of entanglement dynamics in quantum circuits have been at the forefront of quantum information theory research. In particular, random unitary circuit models provide a simple platform for studying entanglement dynamics in generic quantum circuits. Here, we study quantum entanglement dynamics in a random Clifford circuit model , interspersed with single-qubit noise described by quantum channels. We study entanglement dynamics in the presence of noise and competing unitary evolution numerically using Qiskit. The simulation presents numerical evidence that the dissipation time in noisy Clifford circuits is the same as that of noisy circuits . Our work lays the foundation for developing a generalized mathematical proof for the critical time at which a quantum state becomes fully separable when subject to noisy unitary dynamics. If proven, such a theory can be applied to real-world quantum devices and algorithms, assisting with developing fault-tolerant code and hardware.

F11-PH Physics
Ultrasonic Waves for Sustainable Water Purification
Riya Kanury

The purpose of the experiment is to explore the potential impact of ultrasonic waves on water purification. By
investigating a sustainable solution, filtration systems can reduce chemical waste production and improve water quality. The hypothesis states that if contaminated water is treated with ultrasonic waves in an ultrasonic water bath, the absorbance of Methylene Blue will decrease due to the cavitation process breaking down organic pollutants. For the experiment, the contaminated water is separated into three samples. The first sample uses an ultrasonic water bath to produce sound waves, while the second imitates traditional filtration with a water filter. The last sample goes through no filtration or treatment as it is the control. The absorbance of Methylene Blue is measured by using a spectrophotometer on all samples before and after the treatment. Lower absorbances correlate with cleaner water and less turbidity. The data on dye reduction helps evaluate how efficiently the techniques break down organic contaminants. The expected outcome was that ultrasonic waves reduce the amount of Methylene Blue in water. The reduction would happen because the tiny bubbles that cavitation produces collapse with each other, breaking apart the matter and degrading the dye. Through this, the absorbance of Methylene Blue in the water would decrease (measured by the spectrophotometer) since the amount of suspended particles reduces. The experimental results included that the ultrasonic-treated water had a lower average absorbance at 1.10 abs when compared to the control at 1.405 abs. The water filter performed the best with a 0.04 average absorbance. The results show how ultrasonic waves can contribute to a more eco-friendly and efficient filtration system.

F12-PH Physics
How To Control and Optimize Electromagnetic Force Under Non-Superconducting Temperatures?
Zihua Mai

Most existing locomotives with magnetic levitation use liquid helium that are rare
and nonrenewable to cool electromagnets and enhance its force. We attempted to explore the principle of levitating objects through electromagnetic force by investigating electromagnetic levitation systems and optimizing for energy efficiency . For our project, we wanted to explore, how do current, temperature, and wire thicknesses affect the performance of the electromagnets? Previous research concludes thinner wires have higher electrical resistance due to a smaller cross-sectional area increasing collision of electrons against the conductor’s atoms , converting electricity into heat, higher temperature increases the collision rate. We propose that lower temperature, higher electrical current, and thicker wires for the coil result in a stronger magnetic force and also be more energy efficient . 

Our independent variables are temperature, current, and thickness of copper wire. The dependent variable is
the levitation height of the permanent magnet. The controlled variables are the soft iron core, number of turns, and the
levitated permanent magnet. During testing, we change current, water temperature, wire on the electromagnet and observe the height of the rising magnet.

We found a strong positive relationship with current and force. Temperature ranges of 2ÅãC-73ÅãC did not provide a noticeable impact on the electromagnetic strength. The thickness also did not have a strong correlation, the thicker wire (22AWG) had a more stable performance but sometimes thinner wire(30 AWG)produced higher magnetic force when in max current. Thus, continuing to lower the temperature to further reduce electrical resistance is a solution. 

N3-PH Physics
Taking Safety to the Next Level: The Design and Safety Assessment of a Magnetic Levitation Elevator
Yasmine Bridaa

According to the American Elevator Group, traction elevators are the most frequently used mode of transportation in the world. However, despite their popularity, traction elevator rides can be fatal.

Repulsive forces exist between like magnetic poles. Thus, if like magnetic poles are brought close together, magnetic levitation, because of these repulsive forces, can be achieved, providing, for a particular object, stability along the vertical axis. Considering this, can the implementation of magnetic levitation in an elevator system improve the safety of that elevator system?

With the purpose of developing a safer alternative to traction elevators, research was conducted on magnetic levitation and used to design and build a functional concept model of a neodymium-based magnetic levitation elevator, the independent variable of the engineering project, the safety of which was to be evaluated through a set of reliability assessments, the dependent variable of the engineering project, that would assess the velocity, durability, and susceptibility to malfunction of the designed elevator. Prior to the design and assessment of the concept model, it was hypothesized that magnetic levitation elevators could serve as safer alternatives to traction elevators because of their characteristics.

Through evidence provided from the assessments conducted, it was found that the designed magnetic levitation elevator consistently travels at a constant velocity and is not susceptible to elevator malfunction. It was thus concluded, to a high degree of certainty, that magnetic levitation elevators can serve as safer alternatives to the typical traction elevator. 

N4-PH Physics
Ultrafast Pulsed Laser-Induced Degradation and Charge Carrier Dynamics in Multilayer Perovskite Photovoltaics via
Cross-Sectional Multimodal Atomic Force Microscopy
Thantham Jittham

The degradation mechanisms and charge carrier dynamics in single- and multi-layer perovskite photovoltaics under
ultrafast pulsed laser ablation are explored through a comprehensive suite of experimental and computational methods. A custom laser ablation setup, varied in terms of thermal and spatial precision, is utilized to study laser-material interactions. The study contrasts mechanical and photonic ablation techniques to determine optimal approaches for cross-sectional analysis. Structural, potential profile, charge transport dynamics, photoluminescence defects, and chemical impurities are investigated via topographic atomic force microscopy (AFM), Kelvin probe force microscopy (KPFM), conductive AFM (c-AFM), photoluminescence AFM (PL-AFM), field-emission scanning electron microscopy (FESEM), and energy-dispersive X-ray spectroscopy (EDX), respectively. Damages via laser thermal dynamics investigations using infrared imaging and heat spread simulations are presented. Defect propagation and ion transport dynamics are also modeled to elucidate cross-sectional degradation pathways in fabricated devices. Furthermore, laser-processed single-layer and multi-layer perovskite modules are evaluated for performance, with computational device simulations offering predictive insights into module efficiency. Additional chemical and optical analyses, including ultraviolet photoelectron spectroscopy (UPS) and surface UV-Vis transmittance and X-ray diffraction (XRD), provide deeper insights into the graded energy band alignments and material stability of the multi-layer cells. This multimodal approach advances the understanding of perovskite degradation under laser processing, paving the way for improved perovskite photovoltaic module fabrication.

N5-PH Physics
Sensitivity to Discovering Toponium from Simulated LHC Monte Carlo
Nikichi Tsuchida

The purpose of this research is to answer the question, “What is the expected sensitivity to discover ground state
toponium at the Large Hadron Collider?” Our null hypothesis is that toponium does not exist. To determine if our null
hypothesis can be rejected we used simulated Monte Carlo of toponium to model the signal and simulated Monte Carlo of top quark-antiquark pairs as the primary background. In our analysis, we used the selection of one electron and one
anti-muon, with at least one b-tagged jet, and calculated the dilepton mass and angular phi difference for each of the
selected events. After scaling the Monte Carlo to an integrated luminosity of 347 fb-1, the expected integrated luminosity for Run II + Run III, and applying the Monte Carlo weights to the events, we computed the histogram of dilepton mass and angular phi difference. With the histograms, we calculated the chi-squared values and used the chi-squared value to
determine the p-values for both histograms. In the end, the p-values were both <1Å~10-6 for the dilepton mass and angular phi difference distributions respectively, thus rejecting the null hypothesis with a very strong significance. However, there were many limitations in this analysis such as not taking into account systematic errors that could have led to such unrealistically low p-values.

Y9-PH Physics
Examining the Forces that Drive a Crookes Radiometer
Matthew Kellogg
Robert Ronan

The widely accepted theory attributing the mechanism behind the rotation of Crookes radiometer vanes fails to account
for certain aspects of its behavior. This study aimed to identify and confirm the existence of discrepancies between thermal transpiration-theory-predicted and empirically observed radiometer velocity and acceleration trends while simultaneously proposing and examining the accuracy of a new explanation for radiometer rotation based on an atomic electron transition-induced disruption of electrostatic forces between molecules, termed the orbital repulsion theory. Radiometers were tested under a variety of experimental conditions, including heating under low-intensity light focused on just one vane face, rapid heating under high-intensity light, rapid cooling after placement in a low-ambient temperature environment, and heating under limited light frequency ranges using color filters. Velocity and acceleration measurements were gathered by comparing vane positions between subsequent frames extracted from recordings of the experiments. When heated, radiometers exhibited rapid initial acceleration followed by a gradual decline in velocity and an indefinitely stable period of steady-speed rotation, while cooling induced temporary rotation in the opposite direction. Additionally, higher-frequency light was found to correspond with greater rotational velocity. As was hypothesized, the acceleration trends predicted by the orbital repulsion theory more closely resembled the data than those predicted by the thermal transpiration theory, as the latter lacked an explanation for the observed rapid initial acceleration and backward rotation when cooled. These findings suggest the orbital repulsion theory may provide a more comprehensive understanding of radiometer behavior. The theory may also apply to other unexplained phenomena in physics.