Off-line removal of ocular artifacts from event-related potentials using a multiple linear regression model

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Event-related potentials (ERPs) are crucial for understanding brain responses.
• Eyeblinks and eye movements introduce significant artifacts into EEG recordings, distorting ERPs.
• Existing methods struggle to accurately differentiate and correct ocular artifacts from cortical signals.

Purpose of the Study:

• To develop and validate a novel algorithm for correcting ocular artifacts in EEG.
• To accurately isolate and remove eyeblink and eye movement potentials from electroencephalography (EEG) signals.
• To improve the reliability of event-related potentials (ERPs) analysis in the presence of ocular artifacts.

Main Methods:

• A multiple linear regression model with random regressors was employed for artifact correction.
• A propagation factor was calculated for each event-related EEG record.
• Segmentation of EEG records into 2.56-second intervals allowed for distinct artifact factor calculation.
• The algorithm was executed off-line, applying corrections only when electrooculography (EOG) significantly influenced ERPs.

Main Results:

• The algorithm effectively corrects for eyeblink and eye movement potentials in EEG.
• It prevents incorrect propagation factor calculation when ocular and cortical potentials overlap.
• Distinct propagation factors for eyeblinks and eye movements were calculated within single trials.
• Corrections were selectively applied based on the significant influence of EOG on ERPs.

Conclusions:

• The developed algorithm provides a robust method for correcting ocular artifacts in EEG.
• This technique enhances the accuracy of event-related potential (ERP) and cortical DC shift analysis.
• The off-line, data-driven approach ensures reliable artifact removal for improved neuroscience research.