Single-channel EOG artifact removal using fixed frequency EWT and GMETV filter

Area of Science:

• Biomedical Engineering
• Neuroscience
• Signal Processing

Background:

• Portable electroencephalogram (EEG) systems offer user-friendly, wearable solutions for healthcare.
• Electrooculogram (EOG) artifacts, particularly from eye blinks, significantly hinder accurate diagnosis in single-channel EEG (scEEG).
• Effective artifact removal is crucial for reliable scEEG data analysis.

Purpose of the Study:

• To develop and validate an automated method for removing EOG artifacts from scEEG signals.
• To enhance the accuracy of brain signal analysis in wearable EEG systems.
• To provide an effective preprocessing tool for clinical and research applications.

Main Methods:

• An automated artifact removal technique combining Fixed Frequency Empirical Wavelet Transform (FF-EWT) with a Generalized Moreau Envelope Total Variation (GMETV) filter.
• Identification of artifact-contaminated components using kurtosis (KS), dispersion entropy (DisEn), and power spectral density (PSD) metrics.
• Validation using both synthetic and real-world EEG datasets.

Main Results:

• The FF-EWT+GMETV method successfully suppressed EOG artifacts while preserving essential low-frequency EEG information.
• Synthetic data showed lower Relative Root Mean Square Error (RRMSE) and higher Correlation Coefficient (CC).
• Real EEG recordings demonstrated improved Signal-to-Artifact Ratio (SAR) and Mean Absolute Error (MAE).

Conclusions:

• The proposed FF-EWT+GMETV technique offers a robust solution for EOG artifact removal in scEEG.
• This advancement significantly improves the quality of brain signal analysis.
• The method is a valuable preprocessing tool for clinical diagnostics and neuroscience research.