A Novel Method for Estimating Heart Rate Variability Through a Multiple-Input Multiple-Output FMCW Radar

Abstract

In this manuscript, the problem of assessing the heart rate variability (HRV) of a single subject using a colocated multiple-input multiple-output radar of frequency modulated continuous wave type is investigated. The proposed solution exploits beamforming to acquire multiple measurements from different points on the body of the monitored subject. These measurements are combined using a selection strategy that aims at reducing the impact of random body movements while enhancing the overall signal-to-noise ratio. The resulting signal is then filtered using a physiology-inspired filter. Our results demonstrate that the proposed estimation method can accurately identify the instants at which heartbeats occur and achieve accuracy similar to a regular electrocardiogram in terms of specific HRV metrics referring to interbeat intervals.

Keywords

• Biomedical engineering
• biomedical signal processing
• electronic healthcare
• millimeter wave radar
• radar remote sensing
• signal denoising