Accuracy and robustness of an air-ultrasound method for non-contact heart rate and heart rate variability measurements

Abstract

Heart rate (HR) and heart rate variability (HRV) are critical indicators of cardiovascular health. Non-invasive techniques are essential for monitoring vital signs and indirectly measuring brain activity. This study evaluates an air-ultrasound technique for non-contact measurement of HR and HRV. We designed a system to capture skin motions related to cardiac activity, with displacements in the micrometer range. The study aimed to improve the accuracy of cardiac parameter measurement over previous methods by including respiratory tracking. The air-ultrasound method was tested in four young adult males, showing mean absolute errors in interbeat intervals compared to electrocardiogram R-R intervals, ranging from 1.5 to 11 ms, affected by individual variability and conditions. Incorporating respiratory phase information reduced errors by about 30%. Comparable or superior accuracy to contact methods like pulse oximetry, ballistocardiography, and seismocardiography was achieved with a 0.5-meter stand-off. Despite a limited, homogeneous sample in age, sex, and body mass index, the findings show significant variability in skin motion waveforms, necessitating adaptive strategies. The results indicate that this air-ultrasound technique is promising for neuroscientific, behavioral, and clinical applications, where non-contact cardiac monitoring is advantageous.