I am a high school junior from Michigan.
My research interests lie at the intersection of Computer Vision, Machine Learning, Human Computer Interaction with a focus on Assistive Technologies.
I enjoy studying and designing large language models, 3d models, image processing algorithms, sensor based wearables that leverage real-time analysis of spatiotemporal data to predict diseases and support mental and physical well-being of people with disabilities.
Prototype - Self-Standing White Cane for the blind
Designed specifically for a visually impaired co-designer, the self-standing cane allows for the person to use both hands to do everyday activities.
Research - ML based Embryo Implantation Forecast
In the realm of In-vitro fertilization (IVF), the timing of embryo transfer and successful implantation in the uterus is pivotal for success. Uterus Peristalsis (UP), marked by rhythmic uterine smooth muscle contractions, holds particular significance during the early lumenal phase, as supported by extensive literature. Our project aimed to assess uterine readiness for implantation in mice through a machine learning model trained on scanned mouse uteri across different pregnancy stages. The model achieved a 60% accuracy for test images and an overall 64% accuracy for random sample images, with potential for enhanced accuracy through hyperparameter optimization.
Research - Role of Vasoconstriction and Dysregulated Ion Channels in the Protection and Growth of Glioblastoma Multiforme Tumors
Why is Glioblastoma Multiforme (GBM) resistant to chemotherapy and body’s own immunity? Using GEO, STRINGDB and custom R based statistical analysis of differentially expressed gene data of 300 GBM and normal patients, I identified a signaling pathway that may describe the reason for such a resistance.
Accepted for publication in the Journal of Emerging Investigators.
A Novel Transformer-Based CVD Diagnostic Tool
State of the art methods to interpret long range ECG sequences to detect Cardio Vascular Diseases (CVDs) have less than 90% accuracy. Using the MEGA transformer model, I show that accuracy can be improved to 92%.
🥇 Outstanding Research Award - Michigan Science and Engineering Fair
Microfluidics based Braille display
An innovative, low cost, low power soft pneumatic actuator
🥇 Awarded first prize at the Regional Invention Convention
Poster - Impact of air pollution and hospitalizations in Metro Detroit region
Regression analysis of the impact of air pollution on the number of hospital visits and the duration of hospital
🥇 State Award - Health Occupation Students of America
Prototype - Machine Learning Based Concussion detection
Prototype Invention - An innovative IOT and machine learning based approach to detecting sub-concussive head injuries
🥇 State Award, Michigan Science and Engineering Fair, Raytheon National Invention Convention Finalist
