A motion-tolerant heart rate detection method using bio-impedance and MUSIC algorithm

In this work, a motion-tolerant heart rate detection method using bio-impedance and Multiple Signal Classification (MUSIC) algorithm is proposed. Bio-impedance is an attractive method of measuring heart rate; however, its susceptibility to motion severely degrades the performance. As a solution, we propose a pre-processing method with MUSIC, which replaces motion artifact by random noise. To verify the proposed idea, four types of measurement were performed, which are FFT, FFT with pre-processing, MUSIC, and MUSIC with pre-processing. Experimental results show that the proposed MUSIC with preprocessing achieves the most accurate heart rate, with a mean error of 5.6 bpm while others have errors larger than 12.3 bpm.