Related Experiment Video
Updated: Jun 28, 2026
10:41
Using Electroencephalography Measurements and High-quality Video Recording for Analyzing Visual Perception of Media Content
Published on: May 26, 2018
Ocular reduction in EEG signals based on adaptive filtering, regression and blind source separation.
S Romero1, M A Mañanas, M J Barbanoj
1Department of Automatic Control (ESAII), Biomedical Engineering Research Center, Universitat Politecnica de Catalunya (UPC), Barcelona, Spain. sergio.romero-lafuente@upc.edu
Annals of Biomedical Engineering
|November 6, 2008
Summary
Blind source separation (BSS) effectively reduces electrooculographic (EOG) artifacts in electroencephalography (EEG) recordings. This method offers superior performance compared to regression and adaptive filtering for analyzing brain activity.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Quantitative electroencephalography (EEG) is vital for diagnosing neural dysfunction and assessing drug effects.
- Bidirectional contamination from electrooculography (EOG) in EEG data can lead to erroneous conclusions.
- Objective evaluation of ocular artifact reduction methods is limited.
Purpose of the Study:
- To quantitatively evaluate the performance of different ocular artifact reduction techniques using simulated data.
- To compare regression analysis, adaptive filtering, and blind source separation (BSS) for EOG artifact removal.
- To assess the impact of filtering EOG references on cerebral high-frequency components.
Main Methods:
- Simulated EEG data with EOG contamination was used for evaluation.
- Methods assessed included regression analysis, adaptive filtering (with RLS algorithm), and blind source separation (BSS) based on second-order statistics.
- Performance was quantitatively measured by similarity, agreement, and errors in spectral variables between original sources and corrected EEG recordings.
Main Results:
- Errors in artifact reduction were primarily located in anterior brain regions (frontopolar, lateral-frontal) and were prominent in delta and theta frequency bands.
- Filtered versions of time-domain regression and adaptive filtering demonstrated effective ocular reduction.
- BSS based on second-order statistics exhibited the highest similarity indexes and the lowest spectral errors.
Conclusions:
- Blind source separation (BSS) is a highly effective method for ocular artifact reduction in EEG.
- The choice of artifact reduction method significantly impacts the accuracy of spectral analysis in EEG.
- Further research should focus on optimizing BSS techniques for robust EEG analysis.
