On the Heart-Rate Signal Provided by the Zephyr BioHarness 3.0

The BioHarness 3 system (BH3) by Zephyr is a wearable cardiac sensor specifically designed for training optimization of professional athletes. BH3 records the electrocardiogram (BH3_ECG) and the heart-rate signal (BH3_HRS). Specifically, BH3_HRS is neither the popular tachogram nor the direct not-uniformly sampled heart-rate sequence as function of time. Consequently, the aim of the present study is to gain more insights on BH3_HRS that, if found reliable, would allow a future evaluation of the possibility of a clinical use of the sensor for cardiac risk evaluation. Data were acquired from an amateur athlete (male, 25 years old) during a 5-min rest followed by a 10 min exercise. R-peak detection was performed on BH3_ECG, and the obtained heart-rate signal (HRS) was low-pass filtered using the following six filters: 3-, 4-, and 5-sample averages and 0.30 Hz, 0.35 Hz, and 0.40 Hz 6th order Butterworth low-pass filters. The filtered HRSs were then compared to BH3_HRS in terms of correlation coefficient (ρ), mean square error (MSE), resting heart-rate variability (HRV) and exercise maximum heart rate. Results indicate that the HRS closest to BH3_HRS was obtained with the 3-point average (ρ=0.9688-0.9991, MSE=0.45-0.47 mV2; comparable resting HRV and exercise maximum heart rate).