Multifractal analysis of normal RR heart-interbeat signals in power spectra ranges

Abstract

Power spectral density is an accepted measure of heart rate variability. Two estimators of multifractal properties: Wavelet Transform Modulus Maxima and Multifractal Detrended Fluctuation Analysis are used to investigate multifractal properties for the three strongly physiologically grounded components of power spectra: low frequency (LF), very low frequency (VLF) and ultra low frequency (ULV). Circadian rhythm changes are examined by discrimination of daily activity from nocturnal rest. Investigations consider normal sinus rhythms of healthy 39 subjects which are grouped in two sets: 5-hour wake series and 5-hour sleep series. Qualitative arguments are provided to conjecture the presence of stochastic persistence in LF range, loss of heart rate variability during night in VLF range and its increase in ULF.