Average Entropy: Measurement of disorder for cardiac RR interval signals

Abstract

We introduce average entropy (AE) as a disorder measure of heart rate time series. The AE measurement reflects both the probability distribution of all discrete values in a time series and the randomness of the series. The AE analysis is able to differentiate the congestive heart failure (CHF), the healthy, and the atrial fibrillation (AF) subjects for both short and long heart rate time series. The differentiation accuracy of AE is higher than 90%. In addition, since the meaning of "disorder" is easily confused with "complexity", we explore the relation between them by analyzing heart rate time series. Specifically, we demonstrate that the complexity measured by entropy of entropy (EoE) versus the disorder measured by AE plot of heart rate time series exhibits a distinct inverted U relation. (C) 2019 Elsevier B.V. All rights reserved.