Vital signs detection using non-contact radar and singular value decomposition

Background: Recently the intricate progression of the pandemic has presented doctors and healthcare professionals with numerous challenges in conducting medical assessments. Consequently, non-contact health monitoring techniques have garnered significant interest, because this method does not require to contact with the measured person, in the context of an ongoing epidemic. Objective: Within this research, a system for measuring breathing rate and heart rate using a 24 GHz radar was proposed, eliminating the need for direct patient contact. Materials and methods: This study aims to introduce the Singular Value Decomposition (SVD)-based algorithm to separate breathing and heart rate patterns stemming from body surface movements due to vital signs. Experimental results were validated on signals of 12 healthy participants. They were instructed to remain motionless on a bed equipped with a radar beneath it. The radar signals were captured and transmitted to the proposed algorithm. Simultaneously, the participant had a contact sensor put on body to obtain reference signals. Results: The results show high performance with 95 percent limits of agreement on Bland-Altman analysis between the radar and reference measurements of respiration rate and heart rate are 1.65 and 5.51 bpm, respectively. Conclusion: The results of respiratory rate and heart rate derived from the proposed radar signal method with those obtained from a respiratory belt and ECG demonstrated a high level of agreement. This research lays the groundwork for future studies on the use of non-contact radar sensors to measure human vital signs.