Heart Rate Variability Estimated by Heart Instantaneous Frequency in Low-Constrained Tasks

Monitoring cardiac condition by electrocardiograms (ECG) to obtain heart rate (HR) and heart rate variability (HRV) is beneficial for health management. HR and HRV are usually measured by R R intervals which require detecting R-peaks in ECG waveforms. However, R-peaks can be un-detectable due to artifacts caused by motion and respiration. Heart instantaneous frequency (HIF), a technique for estimating HR and HRV by ECG without detecting R-peaks, with an activity-adapted method is applied and evaluated by this study. The ECG data were recorded from 14 healthy individuals (all males, age = 23.3 +/- 1.1 y/o, height = 175.3 +/- 7.7 cm, weight = 71.5 +/- 14.1 kg) in sleeping, free-indoor-activities, outdoor walking, outdoor jogging and post-exercise recovery by a portable device at a sampling rate = 200 Hz. Three-axial acceleration was also recorded on the chest simultaneously with ECG for identifying activity intensity. Little constraint was set on the tasks. Any regular pattern and speed of motion were allowed. The results obtained by R R intervals and HIF are highly correlated the highest correlation coefficients revealed in HR, low frequency power of HRV, and high frequency power of HRV are 0.987, 0.958 and 0.794 respectively. Higher correlation coefficients and smaller relative errors are shown by applying the first harmonic frequency of ECG than other frequency bands to estimate HIF during jogging and recovery. The findings would be useful in developing automatic processing method for portable cardiac monitoring.