Automatic removal of various artifacts from EEG signals using combined methods.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Electroencephalography (EEG) signals are crucial for studying brain activity.
• EEG data often contains artifacts from sources like muscle activity (electromyography, EMG) and eye movements, which can obscure neural signals.
• Accurate artifact removal is essential for reliable EEG analysis.

Purpose of the Study:

• To propose a novel and robust method for the automatic removal of various artifacts from EEG signals.
• To enhance the quality and reliability of EEG data for subsequent analysis.
• To introduce a new peak detection algorithm for identifying specific artifact types.

Main Methods:

• Canonical Correlation Analysis (CCA) was used to separate electromyography (EMG) artifacts.
• Independent Component Analysis (ICA) was applied to decompose EMG-free EEG signals for ocular artifact removal.
• Support Vector Machine (SVM) classifier, utilizing spectral and topographic features, identified eye movement artifacts. A novel peak detection algorithm was developed for eye blink artifact identification.

Main Results:

• The proposed method successfully separated EMG artifacts from EEG signals.
• Ocular artifacts, including eye blinks and movements, were effectively identified and removed using ICA and SVM.
• Comparisons of EEG data before and after artifact removal demonstrated the method's efficacy.

Conclusions:

• The developed method offers a promising solution for comprehensive artifact removal from EEG signals.
• The integration of CCA, ICA, and SVM with a novel peak detection algorithm provides robust artifact suppression.
• This technique can significantly improve the signal-to-noise ratio of EEG recordings.