Ocular artifact suppression for eeg cleaning leading to the brain computer paradigm

Abstract

Brain Computer paradigm or Interface (BCI) establishes a link between the human brain and the computer where an Electro- encephalography (EEG) signal is data of the human brain. EEG signals are a very popular brain activity recording technique. These EEG rhythms are usually impure due to external and internal noise. Recognizing and eliminating ocular artifacts from an EEG signal is very challenging. Different types of research have already been done on identifying and suppressing ocular artifacts. The proposed method suppresses ocular artifacts from electroencephalography signals, producing an artifact-free, clean EEG signal that gives better accuracy. At first, the fast ICA method is applied, followed by wavelet thresholding and stationary wavelet transform, and finally, the artifact is subtracted from the contaminated signal. It is designed to ensure improved accuracy in ocular artifact removal, but it can also suppress muscular and cardiac artifacts. At first, it was checked on a synthetic signal, and then it was checked on the BCI Competition IV dataset, where it suppressed artifacts successfully. It is feasible to evaluate the suggested method's performance metrics and compare these results with those of the EMD and ICA methods to achieve higher accuracy by computing the ratio of signal to noise (SNR) and average square error (MSE). This research emphasizes ocular artifact suppression and minimal signal loss.