EEG preprocessing for synchronization estimation and epilepsy lateralization

Area of Science:

• Neuroscience
• Biomedical Engineering
• Signal Processing

Background:

• Epilepsy diagnosis often relies on electroencephalogram (EEG) analysis.
• Accurate lateralization of epileptic zones is crucial for effective treatment.
• Synchronization analysis in EEG signals is a key area for understanding brain activity.

Purpose of the Study:

• To evaluate synchronization estimation for lateralizing epileptic zones in EEG recordings.
• To assess the impact of preprocessing techniques on synchronization analysis.
• To introduce a novel global synchronization index for hemispheric analysis.

Main Methods:

• Utilized EEG recordings for synchronization analysis.
• Applied preprocessing steps including artifact/noise cancellation (blind source separation), wavelet denoising, and classification.
• Developed a new global synchronization index based on the cross-power spectrum, estimated per cerebral hemisphere.

Main Results:

• The proposed synchronization index demonstrated high accuracy in lateralizing epileptic zones.
• Correct lateralization was achieved in over 90% of cases after applying preprocessing.
• The study confirmed the significant effect of preprocessing on synchronization analysis outcomes.

Conclusions:

• The novel global synchronization index, combined with effective preprocessing, offers a reliable method for epileptic zone lateralization using EEG.
• This approach enhances the diagnostic utility of EEG in epilepsy management.
• Further research can explore the application of this index in other neurological conditions.