EEG is better when cleaning effectively targets artifacts.

Area of Science:

• Neuroscience
• Signal Processing

Background:

• Electroencephalography (EEG) is crucial for studying brain activity but is often contaminated by non-neural artifacts.
• Independent Component Analysis (ICA) is commonly used to remove artifacts by subtracting artifactual components, but this can inadvertently remove neural signals.

Purpose of the Study:

• To demonstrate that standard ICA artifact removal can artificially inflate effect sizes and bias source localization.
• To introduce and validate a novel targeted artifact reduction method for EEG data.

Main Methods:

• Developed a novel method for targeted artifact reduction focusing on specific artifact periods and frequencies.
• Tested the method across diverse EEG systems and cognitive tasks, including Go/No-go and N400 paradigms.

Main Results:

• The targeted artifact reduction method effectively cleaned artifacts from EEG data.
• This novel approach successfully reduced artificial inflation of effect sizes and minimized source localization biases.
• Neural signals were better preserved compared to standard ICA artifact removal.

Conclusions:

• Targeted artifact cleaning in EEG pre-processing, as implemented in the RELAX pipeline, improves data reliability and validity.
• This method mitigates biases and signal loss associated with traditional artifact component subtraction.
• The RELAX pipeline is freely available as an EEGLAB plugin.