Improved Multiscale Permutation Entropy Measure for Analysis of Brain Waves
Young-Seok Choi · International Journal of Fuzzy Logic and Intelligent Systems · 2017
This paper presents a novel multiscale entropy measure for analyzing brain rhythms, called electroencephalograms (EEGs), with an aim of unveiling the underlying dynamics of EEG over multiple time scales.This work is inspired by the known fact that neurological signals such as EEG has distinct dynamics over different temporal scales.To reflect the nonlinear and nonstationary nature of EEGs, the recently developed empirical mode decomposition is incorporated, allowing an EEG to be decomposed into its inherent multiple temporal scale elements, referred to as intrinsic mode functions (IMFs).By computing the permutation entropy of each IMF in a time-dependent manner and averaging them over multiple scales, it yields a data-adaptive multiscale permutation entropy (DAMPE) measure for analyzing the brain waves.Simulation and experimental results show that the proposed DAMPE is efficient to reveal the dynamical changes over multiple scales.