Conservative method for vertical electrooculogram attenuation based on local suppression of ongoing EEG artifact
Dimitri Marques Abramov1, Paulo Ricardo Galhanone1, Vladimir V Lazarev1
1Laboratory of Neurobiology and Clinical Neurophysiology, National Institute of Women, Children, and Adolescents Health Fernandes Figueira (IFF), Fundação Oswaldo Cruz (FIOCRUZ), Rio de Janeiro, RJ, Brazil.
Plos One
|July 18, 2024
Summary
Blinks create artifacts in electroencephalogram (EEG) analysis. A new FilterBlink algorithm effectively removes vertical electrooculogram (VEOG) artifacts, preserving crucial electroencephalogram (EEG) and event-related potential (ERP) data.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Blinking causes significant artifacts in electroencephalogram (EEG) analysis, particularly affecting event-related potentials (ERPs).
- Vertical electrooculogram (VEOG) artifacts, originating from eye blinks, manifest as positive potentials spreading towards posterior EEG channels.
- Existing methods like linear regression and Independent Component Analysis (ICA) for VEOG suppression can lead to information loss.
Purpose of the Study:
- To develop and validate a novel algorithm, FilterBlink, for effective attenuation of VEOG artifacts in EEG signals.
- To assess the performance of FilterBlink in preserving electroencephalogram (EEG) and event-related potential (ERP) data during artifact removal.
- To provide a straightforward and effective method for reducing blink-related artifacts without significant distortion.
Main Methods:
- Statistically identifying VEOG positions in frontopolar EEG channels.
- Generating 'blink templates' through EEG averaging for each channel.
- Subtracting blink templates from EEG segments when the correlation exceeds a threshold L, validated using a computational model.
Main Results:
- FilterBlink recovered 90-98% of original EEG and ERP signals, particularly effective at Fp1 and Fz channels (L=0.1).
- The algorithm successfully reduced VEOG interference with both ongoing EEG and ERPs, especially in anterior medial leads.
- No significant effect of VEOGs or the algorithm was observed on the mid-central channel (Cz).
Conclusions:
- FilterBlink offers a straightforward and effective solution for attenuating VEOG artifacts in EEG data.
- The method preserves essential EEG signal and embedded ERP information, outperforming existing techniques in minimizing data loss.
- This algorithm is a valuable tool for improving the accuracy of ERP analysis in neuroscience research.
