#1
#1 out of 1562.0 est. views
health1d ago
One night of sleep data can reveal risk of over 100 diseases: study
- Stanford's SleepFM AI can assess risk for over 100 health conditions using overnight sleep data.
- The model trained on 580,000 hours of sleep from 65,000 patients between 1999 and 2024.
- SleepFM predicted several diseases with high accuracy, including Parkinson’s, Alzheimer’s, and breast cancer.
- The model also correctly predicted death in 84 percent of cases in the study.
- Researchers plan to add wearable data to improve SleepFM’s predictions.
- Study participants were from sleep clinics, potentially not representative of the general public.
- SleepFM reads multiple physiological signals as a form of ‘language’ for AI learning.
- Researchers plan to integrate wearables data to further improve predictions.
- The research was published in the journal Nature.
Vote 0