AI identifies health risks from routine sleep study data

A novel AI model can use information collected during routine sleep studies to identify patients’ long-term health risks, according to a study published today in Nature Communications. Developed by a multidisciplinary research team, the model uncovered hidden sleep patterns linked to higher odds of heart disease, cognitive decline and death. 
 
The findings also suggest that other routine medical tests may contain substantially more information than is currently being extracted in clinical practice. AI picked out signals in standard overnight sleep study data that are not captured by conventional summary measures alone. 

The research revealed there are patient subtypes with sharply different long-term health risks. Patients in the highest-risk group had twice the mortality risk over the next five years compared to those in the lowest-risk group. This distinction was not captured by the standard clinical measure used to assess sleep apnea severity, the apnea-hypopnea index. 

Each year in the United States, an estimated 1 to 4 million studies are performed in sleep labs, typically to evaluate sleep apnea. While these studies collect rich data on each patient’s brains, lungs, muscles and hearts, clinicians historically have narrowed their focus to a small subset of that information to grade sleep apnea severity.  

“For decades we have distilled an overnight sleep study into a handful of summary measures,” said sleep medicine specialist Dr. Reena Mehra,  professor of medicine at the University of Washington School of Medicine and the study’s senior author. “AI gives us the opportunity to move beyond those summaries and learn from the full richness of sleep physiology.” 

The model was developed by a team of sleep physicians, AI researchers, data scientists and neuroscientists brought together through the Discovery Accelerator, a 10-year research partnership between Cleveland Clinic and IBM. The program is aimed at quickening the pace of discovery in life sciences through AI and quantum computing.  

Using data from the Cleveland Clinic Sleep Signals, Testing, and Reports Linked to Patient Traits (STARLIT) registry, the researchers grouped patients into five risk categories. The model predicted outcomes well for both men and women, while the apnea hypopnea index has historically performed better in men. The findings were independently confirmed in a nationwide patient cohort.  
 
“Modern AI lets us recover much more of the information contained in a night’s worth of sleep physiology, revealing clinically meaningful patient groups with very different long-term health risks,” said the corresponding author, Jeffrey L. Rogers, who is a global research leader at IBM and an adjunct neurosurgery professor at the Yale School of Medicine. 

The model could also help researchers better understand how sleep affects health. Instead of relying on standard measures, it uses AI to spot subtle physiological patterns invisible to the naked eye that can help predict risk for heart disease, neurological disorders and death, opening the door to earlier, more personalized care. 

“Nearly 70 million Americans live with chronic disorders of sleep and wakefulness, affecting daily functioning and overall health. This discovery offers a more personalized approach to sleep medicine by potentially expanding the value of routine sleep testing and reinforcing the key role sleep plays in chronic disease,” said Matheus Lima Diniz Araujo, an applied computer scientist in health care who is a sleep researcher at the Cleveland Clinic. 

“Sleep is increasingly recognized as a critical component of health, yet the physiological information captured during sleep remains largely underused,” said Erhan Bilal, founder of Enkira, previously an IBM researcher and a lead author of the study. 

“As these methods continue to be validated in prospective studies, they have the potential to transform the sleep study from primarily a diagnostic test into a richer source of information about an individual's future health and may accelerate discoveries about the relationships between sleep physiology and chronic disease," Mehra said. 

The research team also included Carl Saab and Kristen Beck from IBM; and Catherine Heinzinger, Samer Ghosn and Nancy Foldvary-Schaefer from the Cleveland Clinic Foundation. 

The study was supported by the Cleveland Clinic-IBM Discovery Accelerator Program and a grant from the National Heart Lung and Blood Institute (1R21HL170206-01).

For details about UW Medicine, please see our About page.