{"id":16030,"date":"2026-01-20T08:00:00","date_gmt":"2026-01-20T07:00:00","guid":{"rendered":"https:\/\/holistic.news\/en\/?p=16030"},"modified":"2026-02-05T11:54:40","modified_gmt":"2026-02-05T10:54:40","slug":"one-night-of-sleep-is-enough-ai-can-reveal-hidden-disease-risk","status":"publish","type":"post","link":"https:\/\/holistic.news\/en\/one-night-of-sleep-is-enough-ai-can-reveal-hidden-disease-risk\/","title":{"rendered":"One Night of Sleep May Be Enough. AI Can Reveal Hidden Disease Risk"},"content":{"rendered":"\n

Groundbreaking AI Estimates Future Disease Risk<\/h2>\n\n\n\n

Researchers at Stanford University have developed a foundation model known as SleepFM<\/strong>. Using data from just one night of polysomnography, the system can estimate long-term risk for more than 100 medical conditions. This breakthrough in AI disease prediction allows scientists to identify elevated risk for illnesses such as prostate cancer, breast cancer, Parkinson\u2019s disease, dementia, heart attacks, and even all-cause mortality.<\/p>\n\n\n\n

SleepFM analyzes data from polysomnography\u2014a comprehensive sleep examination that records brain activity, heart rhythm, muscle movement, oxygen levels, airflow, and other physiological signals.<\/p>\n\n\n\n

\n

\u201cWe record an extraordinary number of signals during a sleep study. It\u2019s an incredibly data-rich process,\u201d<\/p>\n<\/blockquote>\n\n\n\n

-emphasizes study co-author Dr. Emmanuel Mignot, quoted by ScienceDaily<\/a><\/em>.<\/p>\n\n\n\n

The researchers trained the model on a massive dataset comprising 585,000 hours of sleep recordings from 65,000 individuals<\/strong>, aged 2 to 96. Researchers then linked these data with electronic health records spanning up to 25 years of follow-up<\/strong>, allowing them to compare sleep patterns with real medical outcomes over time.<\/p>\n\n\n\n

How SleepFM Works: AI Learns the \u201cLanguage of Sleep\u201d<\/h2>\n\n\n\n

SleepFM functions as a foundation model<\/strong>, similar in concept to large language models<\/a>\u2014except instead of text, it learns patterns from biological signals.<\/p>\n\n\n\n

Each sleep recording is divided into five-second segments, treated as the \u201cwords\u201d of sleep. The model then learns complex relationships between multiple signal channels, including EEG, EKG, respiration, pulse oximetry, and muscle activity.<\/p>\n\n\n\n

The most valuable information comes from subtle mismatches between signals<\/strong>. For example, the system detects when brain activity indicates deep sleep while the cardiovascular system behaves as if the body is awake. These discrepancies often correlate with elevated long-term health risks.<\/p>\n\n\n\n

\n

\u201cSleepFM learns the language of sleep. It doesn\u2019t explain its reasoning in plain English, but we\u2019ve developed interpretation techniques that let us understand what patterns the model focuses on when estimating disease risk,\u201d<\/p>\n<\/blockquote>\n\n\n\n

– explains co-author Dr. James Zou.<\/p>\n\n\n\n

Impressive Accuracy Across Multiple Diseases<\/h2>\n\n\n\n

The AI model evaluated more than 1,000 disease categories<\/strong> and identified 130 conditions<\/strong> where sleep data alone could estimate risk with meaningful accuracy. In these cases, the model achieved a C-index above 0.8<\/strong>, meaning its predictions aligned with real medical outcomes in over 80 percent of cases.<\/p>\n\n\n\n

The strongest results were observed for cancers, cardiovascular disease, neurological disorders, pregnancy complications, and psychiatric conditions. Reported performance included:<\/p>\n\n\n\n