A Selective EOG Removal Method for EEG Signals: The Multi-thresholding Technique

EOG is one of the major artifacts in EEG signal processing. There are varieties of methods have been proposed that aim to eliminate the influence of Occular artifacts on the EEG signals. However, the problem is the trade-off between their performance of removing EOG artifact and their simplicity. In this study, we propose a simple and reliable method but giving a good performance. The idea of this method is to use a multi-threshold technique to target EOG contaminated parts in the signal then selectively subtract it out in order to get a corrected signal with a minimum alteration on the uncontaminated parts. In this study, we used triple-threshold, both in time and frequency domain, to target the contaminated parts (or EOG artifact component). The result shows that besides its simplicity, this method also reliable and effective when selectively removed some typical EOG artifacts like blinks or eye movements without altering other clean parts in the EEG signals. More than that, our method is also able to extract the estimated EOG artifact component from the EEG signal. The need for this method is only one single prefrontal EEG channel, no need for an EOG reference channel for the input. The source code of this method is freely available to download in the form of a MATLAB function by request. We encourage the researchers to give it a try.