Mount Sinai receives NIH grant for AI tool for OSA

NEW YORK – The Icahn School of Medicine at Mount Sinai has been awarded a $3.32 million, four-year grant from the National Heart, Lung and Blood Institute of the National Institutes of Health (NIH) to study a new artificial intelligence (AI)-based tool that has the potential to predict cardiovascular event risk and treatment response among individuals with obstructive sleep apnea (OSA). The findings from this study could help clinicians make more informed recommendations about OSA treatment for their patients. The primary tool for assessing OSA severity is the apnea-hypopnea index, which counts the incidents when an individual stops breathing and the periods when their breathing is shallow due to blockage of the airways. However, the tool does not accurately predict cardiovascular disease risk, and reducing a patient’s apnea-hypopnea index through continuous positive airway pressure (CPAP) therapy—the primary treatment for OSA—does not always result in benefits. “There is no reliable tool for physicians to assess which sleep apnea patients are at highest risk for cardiovascular events, or which patients are likely to respond to, or even be harmed by, CPAP therapy,” said Girish N. Nadkarni, MD, MPH, chair of the Windreich Department of Artificial Intelligence and Human Health, director of the Hasso Plattner Institute for Digital Health, and Irene and Dr. Arthur M. Fishberg Professor of Medicine at the Icahn School of Medicine at Mount Sinai, and chief AI officer for the Mount Sinai Health System. “That makes it difficult to determine which individuals should be escalated for treatment.” The AI-based predictive tool was developed using data from epidemiology studies involving individuals who had and did not have an OSA diagnosis. The tool uses a technique known as transformer-based neural networks, or transformers, to analyze data from polysomnograms—sleep studies that collect data based on up to 20 parameters, including heart rate, muscle movements, and brain activity. The volume and complexity of the data gathered during these tests is such that these parameters have often been assessed in isolation in making decisions about CPAP therapy among OSA patients.