Archive · 8 Jan 2026
AI Predicts 130 Diseases From Sleep Data
Scientists have created a smart computer program called SleepFM that can predict the risk of over 100 diseases by studying how people sleep.
The program was trained using sleep data from 65,000 people, totaling nearly 600,000 hours.
It looks at different signals from the body during sleep, like brain activity, heart function, and breathing.
The program is really good at predicting serious conditions like cancer, heart problems, and mental health issues.
It can even predict the risk of diseases like Parkinson's and developmental delays.
The scientists used a special method to train the program, which helps it understand how different signals relate to each other.
This could be a big step forward in using sleep data to improve health.
The AI identifies subtle patterns in sleep data that might indicate future health problems, allowing for earlier interventions and potentially improving quality of life.
AI model SleepFM predicts risk of 130 diseases from sleep data.
Trained on 600,000 hours of sleep data from 65,000 participants.
Strong predictions for cancer, pregnancy complications, and mental disorders.
Uses multiple data streams like brain activity, heart function, and breathing.
Achieves high C-index scores for conditions like all-cause mortality and dementia.
- Who
- Researchers from Stanford University and other institutions
- What
- Developed an AI model called SleepFM that predicts disease risk from sleep data
- When
- The study was published in Nature Medicine
- Where
- The research was conducted at Stanford University and other institutions
- Why
- To predict the risk of developing over 100 health conditions using sleep data
- AI Model Name
- SleepFM
- Developed By
- Researchers including Stanford University
- Training Data
- 600,000 hours of sleep data from 65,000 participants
- Published In
- Nature Medicine
- Diseases Predicted
- 130 with reasonable accuracy
- Strong Predictions
- Cancer, pregnancy complications, circulatory conditions, mental disorders
- C-Index Scores
- 0.75 or higher for 130 conditions
- Key Predictions
- All-cause mortality (0.84), dementia (0.85), myocardial infarction (0.81)
- Data Analyzed
- Brain recordings (EEGs), blood oxygen levels, heart rate, breathing
- Potential Impact
- Revolutionize preventative medicine, early intervention
Sources
Researchers Develop AI Model That Predicts Disease Risk From Sleep Data
Researchers develop AI model that predicts disease risk from sleep data
'Pillow Talk' With AI: This Health Snore-Cast Can Predict A Person’s Risk For 100 Ailments
