Projects

Brain MRI Medical Image Analysis

A cascaded 2D/3D deep learning pipeline for valve artifact segmentation, catheter segmentation, and length estimation in brain MRI.

Documentation:

• Final ReportPDF

BME Design Team Project — i-Crutch

Development of a novel device to improve therapeutic outcomes of nystagmus under the advisory of Dr. Kemar Green, D.O., Dr. Michael Repka, M.D., Dr. Edward Kuwera, M.D., and Prof. Michelle Zwernemann.

Awarded People's Choice Overall Design Award of JHU Design Day 2025.

Provisional patent filed in April 2026.

Documentation:

• PosterPDF
• Final PaperPDF

BME Design Team Project — Precision Pediatric Trauma Care

Development of a field decision support tool that leverages physiology ultrasound and AI under the advisory of Dr. Emmett Whitaker, M.D., Dr. Isam Nasr, M.D., Dr. Elizabeth Logsdon, Dr. Constanza Miranda, and Prof. Michelle Zwernemann.

Apps for Good Project: Unified Grocery App

With my partner Joseph Yu, we created the Unified Grocery App to help solve the problem of food waste in America in addition to the fact that many Americans do not know how to cook. The main feature of the Unified Grocery App is its ability to recommend recipes to users based on the ingredients they have at home.

Documentation:

• PosterPDF
• Project ProposalPDF
• Software Design DocumentPDF
• Testing PlanPDF

ML Model for March Madness

One of the most unpredictable events in the modern era is the NCAA Division I Men's Basketball Tournament, commonly known as March Madness. With 68 teams competing and 67 games played, there are 2^67 possible bracket combinations. That is a number comparable to the number of grains of sand on the Earth. Suhruth Vuppala and I worked to create a machine learning algorithm that predicts each match with a high probability, with the ultimate goal of getting every match prediction right.

Documentation:

• Sourced PreprocessingPDF
• Modified ModelPDF
• Developed PreprocessingPDF

HiMCM

The High School Mathematical Contest in Modeling (HiMCM) is an international contest designed to provide high school and middle school level students with the opportunity to work as team members to engage and improve their modeling, problem solving, and writing skills. My group (Suhruth Vuppala, Kweku Akese, Jaylin Li, and I) had 72 hours to submit our solution. Although it was hard and tiring, it was really satisfying at the end to submit our solution. The general method of our solution was to regress data points in various ways to get different graphs to predict future data.

Documentation:

• HiMCMPDF

Development of a Modular Below-Elbow Prosthesis with Bidirectional Signaling for Children

My project designed a modular prosthetic arm for children via 3D printing. The goal of my project was to reduce the cost of prostheses for children by creating a modular prototype which reduces the number of prostheses need during childhood. In addition to modularity, other focusing points of the project were: functionality, durability, comfort, and cost-effectiveness. The final product proved to be successful in testing.

Documentation:

• ThesisPDF
• PosterPDF
• Background InfographicPDF
• Grant ProposalPDF
• Graphical AbstractPDF
• LogbookPDF
• Project NotesPDF
• ReferencesPDF

Sit to Stand Assistive Device

For this group project, I served as the Chief Manufacturing Officer (CMO) and worked with Chief Executive Officer (CEO) Vanessa Fobid, Chief Information Officer (CIO) Nihitha Reddy, and Chief Technical Officer (CTO) Mihika Chalasani to develop a STS (sit-to-stand) assist device to reduce the burden on the limbs of individuals struggling to do so. The task of getting in and out of the seated position is known to be a struggle for many, especially the elderly, the injured, and the disabled communities. A combination of springs, hydraulic/pneumatic actuators, and 3D-printed parts were utilized in the development process.

Documentation:

• PosterPDF
• Design StudyPDF
• HandoutPDF