rPPG-Toolbox: Deep Remote PPG Toolbox (NeurIPS 2023)

Self-hosted platform to unify wearable health data through one AI-ready API.

[ICLR'24] Enhancing Healthcare Predictions with Personalized Knowledge Graphs

A Python-based computer vision and AI system for skin disease recognition and diagnosis. Led end-to-end project pipeline, including data gathering, preprocessing, and training models. Utilized Keras, TensorFlow, OpenCV, and other libraries for image processing and CNN models, showcasing expertise in deep learning and machine learning techniques.

[ICLR'25] Reasoning-Enhanced Healthcare Predictions with Knowledge Graph Community Retrieval

Official implementation of "UniMedVL: Unifying Medical Multimodal Understanding and Generation through Observation-Knowledge-Analysis" - A unified medical vision-language model that integrates multimodal understanding and generation capabilities.

MetaPhys: Few-Shot Adaptation for Non-Contact Physiological Measurement (ACM CHIL-2021)

Repository for "AI/MD in Medicine" club in clubhouse

MediScan: AI-powered bone fracture detection system achieving 99.8% accuracy through deep learning. Features real-time X-ray analysis, transparent Grad-CAM visualizations, and clinical integration tools. Built with Python/FastAPI backend and responsive HTML/CSS frontend, making advanced medical diagnostics more accessible to healthcare providers.

Healthcare AI Model Evaluator (HAIME) empowers healthcare organizations to independently evaluate and customize AI solutions, addressing challenges of transparency, clinical relevance, and real-world impact. By putting control in the hands of clinical professionals, it enables confident, context-specific adoption of AI in healthcare.

FactorizePhys: Matrix Factorization for Multidimensional Attention in Remote Physiological Sensing [NeurIPS 2024]

An innovative application designed to help pharmacists and pharmacy students quickly research FDA-approved drugs by retrieving relevant information from drug labels and adverse event datasets, and providing AI-generated summaries to streamline the learning process

iBVP Dataset: RGB-Thermal rPPG Dataset with High Resolution Signal Quality Labels [Electronics 2024]

KG-assisted Medical Dialogue Generation

Resources for the paper titled "Evaluating Latent Space Robustness and Uncertainty of EEG-ML Models under Realistic Distribution Shifts". Accepted at NeurIPS 2022.

A new age EMR application using conversational AI at its best.

AI-powered Clinical Decision Support System (CDSS) for chest X-ray analysis, detecting pneumonia (89.9% accuracy) using CNN on Kaggle's Chest X-ray Dataset. Full-stack app with Next.js, Django, TensorFlow. Features drag-and-drop uploads, vitals input, and diagnostic reports.

The AI-Powered Healthcare Intelligence Network is an AI-driven system offering disease prediction, drug recommendations, heart disease risk assessment, and an AI medical chatbot. Using ML, NLP, and LLMs, it provides accurate diagnoses, insights, and recommendations, enhancing healthcare accessibility, efficiency, and decision-making .

Deep learning model for pneumonia detection in chest X-rays using TensorFlow and Streamlit