Eye blink correction: a test on the preservation of common ERP components using a regression based technique

Abstract

Eye blinks are a pervasive problem in electroencephalography research as they contaminate the brain signal. This paper tests the merits of a software tool employing the regression-based Gratton method that claims to remove the detrimental effects of the eye blink and leaves the activity of the brain. The efficacy of the correction tool was tested on five common stimulus-locked Event Related Potential (ERP) components used in a standard Go/Nogo task. Results suggested that the ‘corrected’ data could be predicted from data containing no eye blinks, suggesting the tool does not distort the data to great extent. This effect was found significant for all components, except for the P3. The conclusion is that this tool distorts the data at acceptable levels, yet caution should be taken when interpreting later components, like the P3.

Keywords

• Eye blink correction
• ERP
• EEG
• Artifacts
• Go/nogo task