Real-time monitoring of spontaneous resonance in heart rate variability.

The resonant characteristic of heart rate variability is usually generated using biofeedback and the external pacing of breathing, which is typically around 6 breaths/min (0.1 Hz), although the exact frequency varies between individuals. It was hypothesized that the actual resonant characteristic of heart rate actually depends on the current psychophysiological state of the subject; therefore, the real-team evaluation of this form of resonance is important for a variety of biofeedback applications. This paper presents an analysis of the spontaneous resonance of heart rate variability generated during singing and non-paced slow breathing after breathing exercises. Two methods for automatic analysis and characterization of heart rate variability resonance in real-time have been presented. The first method uses FFT-based spectral analysis, while the second method calculates the period and amplitude of the RR interval variation during each resonant cycle. The proposed methods were tested on a 45-min record from a chanting session and compared with manually annotated and measured periods.