One Night of Sleep Analyzed by AI Could Predict Risk of 100+ Diseases
Data obtained during sleep can provide valuable clues about overall health.
Advanced analysis of this data, with the help of artificial intelligence (AI), could allow for the estimation of the risk of developing serious conditions, using recordings made during a single night’s sleep.
An artificial intelligence model called SleepFM can use data obtained during a single night’s sleep, in a specialized laboratory, to anticipate the risk of developing over 100 medical conditions, according to researchers.The model was trained on more than half a million hours of sleep data, collected from approximately 65,000 people.
How sleep data is collected
Table of Contents
Data was obtained through polysomnography, the gold standard method for sleep assessment. This involves using multiple sensors that record brain activity, heart activity, breathing, body movements, eye movements, and other physiological parameters. Researchers describe these recordings as a vast source of underutilized physiological information.
To leverage this large volume of data, the team developed the artificial intelligence model and trained it on 585,000 hours of recordings made in various sleep clinics.

Hauri’s Reported Appearance in a Controversial Television Program