An Ingenious Heart Rate Estimation Method in Harmonic Background for FMCW Radar
Xuefeng Zhou, Shisheng Guo, Guolong Cui, Meiqiu Jiang, Luyuan Shi
Abstract
The practicality of noncontact vital signs monitoring based on frequency-modulated continuous wave (FMCW) radar in fields such as medical care and health services has become increasingly prominent in recent years. This article investigates the issue of heart rate (HR) susceptibility to respiratory harmonics in FMCW radar vital signs detection. We introduce a vital signs model incorporating harmonics and propose a method to estimate HR by enhancing harmonic components. Specifically, singular spectrum analysis (SSA) is initially applied to reconstruct the signal, suppressing noise components. Subsequently, the impact of respiratory harmonics on HR detection is assessed. Based on this evaluation, consideration is given to employing the first-order backward difference and variable harmonic product spectrum (VHPS) to enhance the third harmonic of the heartbeat signal, indirectly obtaining HR. Finally, the effectiveness of the proposed method is validated using real data, achieving an average accuracy of 98.51% in HR detection with a root means square error (RMSE) as low as 1.98 BPM.