A skin-interfaced wireless wearable device and data analytics approach for sleep-stage and disorder detection

• Published in: Proceedings of the National Academy of Sciences - PNAS Vol. 122; no. 23; p. e2501220122
• Main Authors: Du, Yayun, Gu, Jianyu, Duan, Shiyuan, Trueb, Jacob, Tzavelis, Andreas, Shin, Hee-Sup, Arafa, Hany, Li, Xiuyuan, Huang, Yonggang, Carr, Andrew N., Davies, Charles R., Rogers, John A.
• Format: Journal Article
• Language: English
• Published: United States National Academy of Sciences 10.06.2025
• Subjects: Adult; Biomarkers; Chronic conditions; Data Analytics; Energy efficiency; Female; Heart rate; Heart Rate - physiology; Humans; Machine Learning; Male; Monitoring, Physiologic - instrumentation; Monitoring, Physiologic - methods; Polysomnography - instrumentation; Respiration; Respiratory Rate; Sleep; Sleep disorders; Sleep Stages - physiology; Sleep Wake Disorders - diagnosis; Sleep Wake Disorders - physiopathology; Tracking; Wearable Electronic Devices; Wearable technology; Wireless Technology - instrumentation; Wrist; sleep; wearable technologies; machine learning
• ISSN: 0027-8424; 1091-6490; 1091-6490
• DOI: 10.1073/pnas.2501220122

Accurate identification of sleep stages and disorders is crucial for maintaining health, preventing chronic conditions, and improving diagnosis and treatment. Direct respiratory measurements, as key biomarkers, are missing in traditional wrist- or finger-worn wearables, which thus limit their precision in detection of sleep stages and sleep disorders. By contrast, this work introduces a simple, multimodal, skin-integrated, energy-efficient mechanoacoustic sensor capable of synchronized cardiac and respiratory measurements. The mechanical design enhances sensitivity and durability, enabling continuous, wireless monitoring of essential vital signs (respiration rate, heart rate and corresponding variability, temperature) and various physical activities. Systematic physiology-based analytics involving explainable machine learning allows both precise sleep characterization and transparent tracking of each factor’s contribution, demonstrating the dominance of respiration, as validated through a diverse range of human subjects, both healthy and with sleep disorders. This methodology enables cost-effective, clinical-quality sleep tracking with minimal user effort, suitable for home and clinical use.