Stanford's TRACER: training-free multimodal AI tracks patient behavior from bedside video + EHR
StanfordAILab · x · 2026-10-17
Stanford AI Lab (SAIL) unveiled TRACER (Trajectory Reasoning Across Clinical Event Records), a training-free multimodal system enabling LLMs and VLMs to reason over bedside video and electronic health records to track how patient behavior changes across hospital stays. The team includes Li Fei-Fei, Kevin Schulman, and Arnold Milstein.
- Motivation: patient behavior and mobility are key clinical signals, but observations are incomplete and temporally scattered
- Builds on years of ambient-intelligence work (hand hygiene detection, ICU mobility)
- Revealed behavioral patterns associated with delirium
More from Apps
- Developer has Claude cut a game trailer from real gameplay, music included — walkingriver · 2026-10-17
- Early user verdict: Dots beats Muse as a trustworthy email agent — prd_008 · 2026-10-17
- YouTube Music adds AI labels separating fully, partly, and non-gen-AI tracks — lmoroney · 2026-10-17
- AI influencer workflow: Codex + CapCut Web takes one character from image to video campaign — alifcoder · 2026-10-17
- Debate over using an AI agent to watch friends' Instagram Stories — omooretweets · 2026-10-17
- Pika partners with OpenAI to embed its full video toolset inside ChatGPT — HeyToha · 2026-10-17