Extraction and Evaluation of EEG Covariates and Their Influence on GLM Model: EEG covariates and their influence on GLM model

Abstract

This study aims at the identification of suitable approaches to dimension reduction methods for EEG covariate extraction for GLM analysis of fMRI time series. We present the results of anatomical and mathematical methods of dimension covariate reduction and their combinations. Individual models according to the used covariates showed that jPCA creates a lower number of significantly correlated voxels. Anatomical reduction balances the number of correlated voxels between mean and jPCA. The choice of covariates has a significant effect on the resulting GLM activations. The average allows generalization to explain a physiological activity, jPCA offers the ability to identify specific activations.