Design of a microcontroller-based real-time heart rate variability measurement system using a low-complexity r-peak detection algorithm

The development of a real-time heart rate variability (HRV) measurement system is reported based on microcontrollers. Previous studies have analyzed autonomic nervous system changes using an offline method to obtain HRV parameters; therefore, researchers cannot monitor the changes as they occur and cannot perform biofeedback. To calculate real-time HRV parameters, we first developed a low-complexity R-peak detection algorithm to determine the RR intervals. This algorithm only highlights the R peak position; thus, it can quickly calculate RR intervals and easily be implemented in a microcontroller. Experimental results show that the rate of heartbeat error of our algorithm is less than 1 bpm. In addition, a high correlation coefficient (r > 0.99) exists between the measured and actual heart rates in the tested range of 20-200 bpm. The HRV parameters of each subject were calculated using Bland-Altman statistical analysis and had a narrow LoA, and all parameters exhibited good correlation (r > 0.91). Thus, the results provide evidence that the system can generate adequately reliable HRV parameters. Importantly, the system can generate real-time HRV parameters; thereby facilitating autonomic nervous system research to elucidate the modulation of and changes in sympathetic and parasympathetic neural activities.