Validation of noncontact cardiorespiratory monitoring using impulse-radio ultra-wideband radar against nocturnal polysomnography

Abstract

Polysomnography (PSG) is a standard diagnostic test for obstructive sleep apnea (OSA). However, PSG requires many skin-contacted sensors to monitor vital signs of patients, which may also hamper patients' sleep. Because impulse-radio ultra-wideband (IR-UWB) radar can detect the movements of heart and lungs without contact, it may be utilized for vital sign monitoring during sleep. Therefore, we aimed to verify the accuracy and reliability of the breathing rate (BR) and the heart rate (HR) measured by IR-UWB radar. Data acquisition with PSG and IR-UWB radar was performed simultaneously in 6 healthy volunteers and in 15 patients with suspected OSA. Subjects were divided into 4 groups (normal, mild OSA, moderate OSA, and severe OSA) according to the apnea-hypopnea index (AHI). BRs and HRs obtained from the radar using a software algorithm were compared with the BRs (chest belt) and the HRs (electrocardiography) obtained from the PSG. In normal and in mild OSA, BRs (intraclass correlation coefficients R [ICCR] 0.959 [0.956-0.961] and 0.957 [0.955-0.960], respectively) and HRs ([ICCR] 0.927 [0.922-0.931] and 0.926 [0.922-0.931], respectively) measured in the radar showed excellent agreement with those measured in PSG. In moderate and severe OSA, BRs ([ICCR] 0.957 [0.956-0.959] and 0.873 [0.864-0.882], respectively) and HRs ([ICCR] 0.907 [0.904-0.910] and 0.799 [0.784-0.812], respectively) from the two methods also agreed well. The IR-UWB radar could accurately measure BRs and HRs in sleeping patients with OSA. Therefore, IR-UWB radar may be utilized as a cardiopulmonary monitor during sleep.