Related Experiment Video
Updated: Jul 19, 2026
10:23
Equipment Setup and Artifact Removal for Simultaneous Electroencephalogram and Functional Magnetic Resonance Imaging for Clinical Review in Epilepsy
Published on: June 23, 2023
A fully automated correction method of EOG artifacts in EEG recordings.
A Schlögl1, C Keinrath, D Zimmermann
1Institute of Human-Computer Interfaces, Graz University of Technology, Krenngasse 37/IV, A-8010 Graz, Austria. alois.schloegl@tugraz.at
Summary
This study introduces a fully automated method to reduce electrooculography (EOG) artifacts in electroencephalography (EEG) recordings. The validated regression-based approach significantly improves artifact detection and correction, offering a viable alternative to manual methods.
Area of Science:
- Biomedical Signal Processing
- Neuroscience
- Signal Artifact Correction
Background:
- Electroencephalography (EEG) is crucial for brain activity monitoring.
- Electrooculography (EOG) artifacts commonly contaminate EEG data, compromising analysis.
- Manual artifact identification is time-consuming and often insufficient.
Purpose of the Study:
- To present and validate a fully automated method for reducing EOG artifacts in EEG.
- To compare the effectiveness of the automated method against manual artifact detection.
Main Methods:
- A regression analysis-based correction method was developed.
- The method was applied to 18 multi-channel EEG recordings.
- Two independent experts performed blinded evaluations of corrected and raw EEG data.
Main Results:
- Manual review identified EOG artifacts in 5.9% of raw data, with 4.7% corrected.
- The automated method identified additional EOG artifacts in 1.9% of data.
- The proposed method achieved an 80% reduction in EOG artifacts.
Conclusions:
- The automated EOG artifact reduction method is a viable and efficient option.
- It significantly outperforms manual identification and rejection of EOG artifacts.
- The method is available for offline and online analysis via the open-source BioSig library.
