Patch-Type Heart Rate Variability Analysis with Artificial Intelligence for Detection of Obstructive Sleep Apnea
Hsu YS, Lin YC, Kuo YE et al., Nature and Science of Sleep, February 5, 2026
A new comfortable chest patch using AI and heart monitoring can detect moderate-to-severe sleep apnea with over 80% accuracy - better than previous methods and without bulky equipment.
Researchers developed an AI-powered patch-type heart rate variability (HRV) analyzer that achieved 81.4% accuracy in screening moderate-to-severe OSA (AHI≥15). The device introduces a novel Cardiovascular Hypopnea Index (CVHI) using ECG amplitude features and nonlinear analysis, significantly outperforming demographic-based screening (73%) and previous ECG-based methods (70.6%). The comfortable, wearable device can continuously monitor for up to one week without significantly disrupting sleep.
Traditional sleep studies are cumbersome and often delayed. This wearable option could expand access to OSA screening and help identify high-risk patients faster. For patients avoiding in-lab studies due to discomfort or inconvenience, patch-type devices offer a practical alternative with strong accuracy.
