Overview
ML/CV musculoskeletal analysis systems for clinical diagnosis from low-quality phone video. 500+ engineered features, 97% accuracy. Built at RoCA Health (Stanford-affiliated).
Sarcopenia Analysis
Sit-to-stand movement analysis for sarcopenia detection using MediaPipe and Random Forest classifiers on the Stanford dataset.
Athletic Injury Rehabilitation
Post-injury functional tests and AI-based analysis for sports medicine. Pose extraction and kinematic analysis tools deployed on AWS (EC2/Elastic Beanstalk).