Projects
Agentic AI That Thinks in Healthcare
We build an agentic AI that reasons across biomedical text, genetic information, 3D molecular structures, and medical imaging. By connecting drugs, diseases, and patient biology, it uncovers meaningful relationships and delivers transparent, evidence-backed insights that help clinicians and researchers make sense of complex data faster and with greater confidence.
Super Resolution using Deep Generative Denoising Diffusion on Video Sequence Data
Single-image super-resolution framed as an ill-posed problem; explores diffusion-based generative models that outperform conventional SISR approaches on video sequences.
Retitrack™
Class II device (21 CFR 886.1510). Intended for recording, viewing, measuring, and analyzing fixation and saccadic responses to visual stimuli in healthcare settings.
Deep Active Learning Using Retinal Image Embedding Vectors to Optimize Training Data Selections
Uses embedding vectors from retinal imagery to guide active learning, reducing labeling load while preserving model performance for medical applications.