AI Model Could Predict Approximately 130 Diseases Using Data From a Single Night's Sleep

Scientists from Stanford University have created the first artificial intelligence (AI) model that could help predict the risk of developing approximately 130 diseases using information gathered from a single night of sleep, according to a study published Tuesday in the journal Nature Medicine.

The model, SleepFM, has been trained on nearly 600,000 hours of sleep data from approximately 65,000 participants. It combines brain, heart, muscle, and respiratory signals collected through polysomnography (PSG), the "gold standard" for sleep analysis, the researchers stated in the article.

This research represents the first instance of using AI to analyze sleep data on such a large scale.

The new tool could identify risks for diseases with high mortality rates, such as dementia, myocardial infarction, heart failure, chronic kidney disease, stroke, and atrial fibrillation.

"SleepFM produces latent sleep representations that capture the physiological and temporal structure of sleep and enable accurate prediction of future disease risk," the text explains.

The study emphasizes that artificial intelligence overcomes the challenges of analyzing the vast amount of data obtained from polysomnography.

"From an AI perspective, sleep is relatively understudied," stated James Zou, PhD, an associate professor of biomedical data science and co-author of the study, on the Stanford School of Medicine website. Stanford is one of the United States' elite academic centers, located in California.