Mental fatigue analysis by measuring synchronization of brain rhythms incorporating enhanced empirical mode decomposition

Abstract

A new and effective approach for mental fatigue analysis is presented here. Empirical mode decomposition (EMD), as a fully adaptive and data-driven method for analyzing nonlinear and nonstationary time series, is presented for measuring the synchronization of the brain rhythms from different brain lobes. The EMD algorithm is applied to a desired channel and each time one of the extracted intrinsic mode functions (IMFs) is considered as one of the brain rhythms. This IMF can be filtered by an adaptive line enhancement (ALE) algorithm. The superiority of using ALE to conventional filtering has been tested using simulated signals. Then, by applying Hilbert transform to several enhanced IMFs from different parts of the brain, the changes in linear and non linear synchronization levels are estimated for determination of the fatigue state.