Effectiveness of Online Ocular Artifact Removal from Electroencephalogram Signal during Neurofeedback Training

Electroencephalography (EEG) is an electrophysiological, non-invasive method that can record the electrical signals of the brain by placing electrodes on the scalp. Neurofeedback training (NFT) is a learning-based, non-invasive intervention method that uses the Brain-Computer Interface (BCI) to regulate the performance of the brain. Artifacts in EEG are signals that are not associated with brain activities, and these signals may affect the NFT process. So, it is important to remove these artifacts from EEG signals. In this paper, a neurofeedback system is designed to perform the real-time Electrooculography (EOG) artifact removal using Independent Component Analysis-Regression (ICA-REG). The NFT-BCI system is designed by using the EMOTIV Insight headset to collect raw EEG signals, openvibe for processing the EEG signal, and the Unity3D application for the interface of the BCI system. The designed BCI system was used to perform alpha NFT for 6 subjects for 6 sessions and analyze the EEG data recorded from the subjects. Although no significant effect on NFT has been found between using the raw and clean EEG signals, high correlation and low RMSE between the raw and artifact-cleaned EEG data indicates the effectiveness of the proposed artifact removal process.