EEGs Signals Artifact Rejection Algorithm by Signal Statistics and Independent Components Modification

Abstract

Electroencephalography (EEG) signals acquire a lot about the brain functionality, its different patterns employed in brain diseases recognition, and recently used in Brain Computer Interface (BCI) systems. Automatic recognition of these patterns gains a lot of attention nowadays. These EEG signals are contaminated with artifacts like eye and muscle movement artifacts. Fine tuning of these signals and automatic rejection of artifacts prior to feature extraction is straightforward. In this paper, a novel method for artifact cancelation based on signal statistics with modification of independent sources extracted by Independent Component Analysis (ICA) of EEG signals is suggested. Visual inspection of the reconstructed signals shows the validity of the proposed method in artifact rejection. Moreover, this method did not require any extra information channel attached with EEG signals.