Ocular artifact suppression from EEG using ensemble empirical mode decomposition with principal component analysis

Abstract

Signals associated with eye blinks (230–350 micro-volts) are orders of magnitude larger than electric potentials (7–20 micro-volts) generated on the scalp because of cortical activity. These and other such non-cortical biological artifacts spread across the scalp and contaminate the Electroencephalogram (EEG). We present here a novel approach for efficient detection and effective suppression of these artifacts using single channel EEG data by combining Ensemble Empirical Mode Decomposition (EEMD) along with Principal Component Analysis (PCA). We present a methodology for ocular artifact suppression, by performing EEMD on the contaminated EEG data segment to get the intrinsic mode functions (IMFs) and subsequent elimination of artifacts by automatic selection of particular principal components, which capture ocular artifact features after using PCA on IMFs.