Clinical study of an integrated sensor system for detection and classification of obstructive sleep apnea (OSA)

Abstract

Obstructive Sleep Apnea (OSA) is traditionally diagnosed via Polysomnography (PSG), which relies on multiple wired sensors in an unfamiliar hospital setting. In this study, a compact home-sleep-test system is proposed, integrating a fringing-field capacitive sensor for wireless respiratory-effort monitoring system(, and a nasal airflow sensing system (2 cmx2 cm) with an connected 2 cm x 1 cm temperature sensor (both wired to the processing unit). A customized signal-processing algorithm was developed to denoise both channels and automatically identify apnea and hypopnea events. Validation with subjects (n = 31) demonstrated performance metrics (SN = 0.846, SP = 0.944, Precision = 0.917, and Accuracy = 0.903, ) in classifying OSA severity. By combining novel capacitive fringing-field sensing and temperature-based airflow measurement into a largely wireless wearable, a practical and accurate alternative to traditional PSG for at-home OSA detection is offered.