Noncontact Cardiac Parameters Estimation Using Radar Acoustics for Healthcare IoT

• Published in: IEEE internet of things journal Vol. 11; no. 5; pp. 7630 - 7639
• Main Authors: Rong, Yu, Lenz, Isabella, Bliss, Daniel W.
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.03.2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Acoustic measurements; Acoustic waves; Acoustics; Cardiac parameters; Feasibility studies; Filter design (mathematics); Health care; Heart rate; heart rate (HR); Heart rate variability; heart sound (HS); Internet of medical things; MIMO; Motion perception; Parameter estimation; Phonocardiography; Radar; Radar detection; Radar measurements; radar signal processing; Remote sensing; Respiration; Sensors; Signatures; Stethoscope; Wireless sensor networks
• ISSN: 2327-4662; 2327-4662
• DOI: 10.1109/JIOT.2023.3317670

Novel use of radar acoustics for noncontact cardiac parameter estimation is presented as a potentially invaluable addition to wireless sensor networks in Healthcare IoT. Here, radar acoustics refers to the radar captured high-frequency mechanical motion, beyond 20 Hz, on the skin surface induced by heart sound (HS) acoustic waves. Conventional radar-based heart rate (HR) detection methods rely on detecting the heartbeat motion, around 1 Hz, while the unaddressed respiration motion coupling issue prevents consistent heartbeat estimation accuracy in the presence of stronger respiration interference, below 1 Hz. Another issue common to prior methods is the nonadaptive spectral filter design for separating heartbeat signal. As a consequence, these methods cannot be generalized to measurement spaces with high heartbeat variability as demonstrated in this study. To effectively address these limitations, a new HR detection method using radar is proposed to derive a high-fidelity pulse template from the smoothed envelogram of HS measurements. What is more, the feasibility study of radar HS-based HR variability analysis is performed, in which inter-beat-internal is derived from the first and second HS signatures. To that end, automatic identification of HS signatures and signal denoising techniques are developed accordingly. Multiple validation experiments are conducted on human subjects. Our results imply that the new methodology effectively improves remote cardiac sensing performance using the radar technology and presents an exciting opportunity to expand Healthcare IoT, improving patient experiences and outcomes.