Removal of EOG artifacts using ICA regression method

Abstract

The present study shows a hybrid method of removing Electrooculogram (EOG) artifacts. The Independent Component Analysis (ICA) method is used in conjunction with the regression method. The ICA estimates the EOG recordings from the Electroencephalogram (EEG) channels to be used to compute the propagation coefficients for the regression method. This ICA regression method is applied on to a set of EOG contaminated EEG signals. The result is compared to the typical regression method. It was found that the ICA regression method performs better than the typical regression method in removing artifacts especially the ones that contain eye blinks. © 2008 Springer-Verlag.