A unique AI model developed by Stanford University researchers and their colleagues could one day be used to predict your risk of more than 100 health conditions, without you even needing to be awake.
As detailed in a recently released paper, the SleepFM AI model analyzes a comprehensive suite of physiological recordings to predict a person’s future risk of dementia, heart failure, and all-cause mortality – based on a single night of sleep.
SleepFM is a foundation model, like ChatGPT, trained on a vast dataset of nearly 600,000 hours of sleep data gathered from 65,000 participants. As ChatGPT learns from words and text, SleepFM learns from 5-second increments of sleep data from recordings from various sleep clinics.
Sleep clinicians collected this data through an extensive, if uncomfortable, technique called polysomnography (PSG). This ‘gold standard’ of sleep studies uses various sensors to track activity in the brain, heart, and respiratory system, as well as leg and eye movements, during states of unconsciousness.
“We record an amazing number of signals when we study sleep,” says Emmanuel Mignot, sleep medicine professor at Stanford and the paper’s co-senior author. Read More
Source: Science Alert, Ivan Farkas
Scientect 