Related Experiment Video
Updated: Aug 9, 2025
High Density Event-related Potential Data Acquisition in Cognitive Neuroscience
Published on: April 16, 2010
RELAX part 2: A fully automated EEG data cleaning algorithm that is applicable to Event-Related-Potentials
N W Bailey1, A T Hill2, M Biabani3
1Central Clinical School Department of Psychiatry, Monash University, Camberwell, VIC, Australia; School of Medicine and Psychology, The Australian National University, Canberra, ACT, Australia; Monarch Research Institute Monarch Mental Health Group, Sydney, NSW, Australia.
The RELAX pipeline effectively cleans electroencephalography (EEG) artifacts for Event-Related Potentials (ERPs), even with low-frequency filtering. This automated method improves data consistency and reliability in ERP research.
Area of Science:
- Neuroscience
- Signal Processing
Background:
- Electroencephalography (EEG) records neural activity, often analyzed as Event-Related Potentials (ERPs).
- Non-neural artifacts contaminate EEG data, complicating ERP analysis and requiring artifact cleaning.
- Existing automated artifact cleaning methods often exclude low frequencies vital for ERPs (<1 Hz).
Purpose of the Study:
- To evaluate the RELAX (Reduction of Electroencephalographic Artifacts) pipeline for cleaning EEG data intended for ERP analysis.
- To assess RELAX's performance compared to established artifact cleaning pipelines.
Main Methods:
- Compared multiple RELAX pipeline versions against four standard EEG cleaning pipelines.
- Evaluated cleaning performance using artifact cleaning metrics and ERP condition-based variance.
- Utilized a Go-Nogo task dataset for comparative analysis.
Main Results:
- RELAX, particularly with Multi-channel Wiener Filtering (MWF) and wavelet-enhanced independent component analysis (wICA_ICLabel), demonstrated superior artifact cleaning.
- These RELAX configurations yielded highly dependable ERP estimates.
- RELAX maintained high cleaning performance even with data high-pass filtered at 0.25 Hz, suitable for ERPs.
Conclusions:
- RELAX offers effective artifact cleaning for ERP research, preserving crucial low-frequency information.
- The pipeline is user-friendly, implemented in MATLAB via EEGLAB, and available on GitHub.
- RELAX is recommended for enhancing objectivity and consistency in ERP studies.
