Stanford's AI can spot 130 diseases from a single night's sleep

Technology Jan 07, 2026

Stanford researchers have built an AI, SleepFM, that predicts your risk for 130 diseases—just by analyzing one night of sleep data.
The model was trained on over half a million hours of sleep studies from people aged 2 to 96.

What makes SleepFM special?

SleepFM looks at brain waves, heart rate, breathing, and even leg and eye movements to forecast health risks years in advance.
It's especially accurate for tough-to-detect issues like Parkinson's (0.89 accuracy), several cancers, dementia, and heart disease.

Why it matters

This could mean getting early warnings about serious health problems—using tech that fits right into a normal sleep study.
It hints at a future where catching health risks early might be possible with just a single night's sleep data.