Multiscale and moving average entropy: Statistical dependence in stressed subjects

Abstract

The complexity quantification of physiologic signals is a field of considerable importance since loss of complexity has been proposed as a generic feature of pathological systems. This paper investigates entropy of the preprocessed time series of different laboratory animals-normal and borderline hypertensive rats. The systolic blood pressure and heart rate were recorded before the stress, during the first 20 min exposure to stress, after the first exposure and after the last exposure to stress. The preprocessing included high-pass filtered, normalized and centralized original time series, its moving averaged and multiscaled counterparts, as well as surrogate data and Gaussian samples. © 2009 IEEE.