Parametric phase diffusion analysis of irregular oscillations

Neuroscience Institute, Georgia State University - Atlanta, GA, 30303 USA

Received: 8 July 2014
Accepted: 26 August 2014

Abstract

Parametric phase diffusion analysis (ΦDA), a method to determine variability of irregular oscillations, is presented. ΦDA is formulated as an analysis technique for sequences of Poincaré return times found in numerous applications. The method is unbiased by the arbitrary choice of Poincaré section, i.e. isophase, which causes a spurious component in the Poincaré return times. Other return-time variability measures can be biased drastically by these spurious return times, as shown for the Fano factor of chaotic oscillations in the Rössler system. The empirical use of ΦDA is demonstrated in an application to heart rate data from the Fantasia Database, for which ΦDA parameters successfully classify heart rate variability into groups of age and gender.