Online EEG artifact suppression for neurofeedback training systems

Online EEG artifact suppression system is a crucial function of real-time Brain Computer Interface (BCI) applications. EEG artifacts significantly affect the accuracy of feature extraction and data classification for estimating cognitive states in Neurofeedback Training (NFT) systems. The EEG artifacts derived from ocular and muscular activities are inevitable and unpredictable due to subject´s physical conditions. One of the most prominent techniques employed to suppress the EEG artifacts is Independent Component Analysis (ICA). This technique separates EEG signals into Independent Components (ICs) and then discriminates EEG artifacts from neurally generated brain signals. Nevertheless, the source separation of ICA algorithm is imperfect. The IC identified to be an artifact can include brain wave activities useful for state classification. The proposed method will elaborate on the ICs with a low-complexity wavelet transform to extract the useful neural signals from the artifact component in real-time. This suppression technique implemented in NECTEC´s Neurofeedback System for Attention Training was tested in pre-trial sessions with 10 healthy subjects and 5 MCI patients at Chulalongkorn Hospital. Experimental results prove the performance and accuracy of the proposed suppression algorithm of light and strong eye-blink artifacts.