Generation of synthetic RR interval sequences of healthy patients

Abstract

A method based on a heart model is developed to generate synthetic RR interval time series of a healthy subject. The model presented here makes no assumptions about the distribution of RR intervals. The main contribution of this paper is the introduction of variability associated in the RR interval series which arises from changes in the input from parasympathetic, sympathetic input as well as from the response of other components in the cardiovascular control circuitry. This is modelled by a time varying chaotic sequence using the improved logistic map multiplied by a constant parameter. In simulating a measured RR series, its mean heart rate is used and a set of ordinary differential equations solved for different values of this constant parameter. This parameter is chosen when the power present in the low frequency (0.04–0.15) Hz and high frequency (0.15–0.4) Hz spectral range of the measured series show minimal error with that present in the synthetic series. The synthetic series generated is devoid of artefacts and is closer to the cleaned version of the measured series. The complexity assessed using sample entropy indicates errors less than 10% between the two series. This indicates that the chaotic function used to model the variability is a suitable choice to give the complexity of a healthy RR series. The artefact free synthetic RR series of healthy subjects provides a useful reference to compare results from diseased subjects; and to test different artefact cleaning procedures on RR data from healthy subjects