VME-EFD : A novel framework to eliminate the Electrooculogram artifact from single-channel EEGs

Area of Science:

• Neuroscience
• Signal Processing
• Biomedical Engineering

Background:

• Electroencephalography (EEG) is crucial for diagnosing neurological disorders.
• EEG signals are often contaminated by electrooculography (EOG) artifacts from eye movements.
• Accurate artifact removal is essential for reliable EEG analysis.

Purpose of the Study:

• To introduce a novel VME-EFD framework for effective EOG artifact removal from EEG data.
• To enhance the accuracy of EEG-based neurological disorder diagnosis.
• To improve the performance of brain-computer interface (BCI) applications.

Main Methods:

• The proposed VME-EFD framework combines Variational Mode Extraction (VME) and Empirical Fourier Decomposition (EFD).
• EEG signals are decomposed by VME into EEG and EOG components.
• EFD analyzes EOG components based on energy and skewness to identify and remove artifact-level.

Main Results:

• VME-EFD demonstrated superior performance compared to existing methods in simulations.
• Lower Root Mean Square Error (RMSE) and improved spectral power difference (ΔPSD) in the alpha band were observed.
• Higher correlation coefficients (CC) indicate effective artifact removal while preserving EEG signal integrity.

Conclusions:

• The VME-EFD framework successfully removes EOG artifacts from EEG signals.
• The method preserves critical EEG features, especially in the alpha band.
• VME-EFD is highly suitable for improving the accuracy of BCI applications and neurological diagnostics.