Detection of Arrhythmic cardiac signals from ECG recordings using the Entropy-Complexity plane

Abstract

The aim of this paper was to analyze in the Entropy-Complexity plane (HxC) a set of time series coming from ECG of humans, with the objective to discriminate recordings from two different groups of patients: normal sinus rhythm and cardiac arrhythmias.
The time series dataset used in this paper were ECG recordings obtained from PhysioNet. These were 47 long-term signals of patients with diagnosed cardiac arrhythmias and 18 long-term signals from normal sinus rhythm patients. The HxC plane was built with the Shannon’s Entropy on x-axis and the statistical complexity on y-axis. As the information content about the time series is conveyed in the form of a probability distribution function (PDF), the methodology proposed by Bandt & Pompe (2002) was applied to compute it.
Average values of statistical complexity and normalized Shannon entropy were calculated and analyzed in the HxC plane for each time series, where the average values of complexity from ECG time series of patients with diagnosed arrhythmias were bigger than normal sinus rhythm group. On the other hand, the Shannon entropy average values for arrhythmias patients were lower than the normal sinus rhythm group. These notorious characteristics made possible to discriminate different spaces of both groups of signals in the HxC plane. The results were analyzed through a multivariate statistical test hypothesis.
The methodology proposed has a remarkable conceptual simplicity, computational speed, and robustness to noise. It shows a promissory efficiency in the detection of cardiovascular pathologies.