VIDEO: Machine learning may be effective in early detection of REM sleep behavior disorder

Machine learning may be an effective tool for screening individuals who develop REM sleep behavior disorder and subsequent neurodegenerative conditions, Yoav N. Nygate, said in this Healio video.

"It’s estimated that over 80% of the patients with [REM sleep behavior disorder] will at some point develop some form of neurodegenerative condition,” Nygate, an artificial intelligence engineering manager at EnsoData, a software company based in Madison, Wisconsin, said in an interview during SLEEP 2023. “Which means that RBD may be used as some form of an early effective risk marker for neurodegeneration.”

Nygate and fellow researchers from Rush University Medical Center sought to employ machine learning to evaluate individuals who may show signs of REM sleep behavior disorder (RBD), including subgroups of those with Parkinson’s disease.

Nygate added that future endeavors would be focused on the robustness of the machine models, particularly for those with sleep apnea, and the performance on real-world data sets in clinical settings.