Improving electroencephalographic source localization of epileptogenic zones with time-frequency analysis

Area of Science:

• Neuroscience
• Medical Imaging
• Epileptology

Background:

• Accurate localization of the epileptogenic zone is crucial for effective epilepsy treatment.
• Current methods for noninvasive functional neuroimaging, such as electroencephalography (EEG), require advanced analysis techniques for precise localization.
• Space-time-frequency analysis of EEG signals offers potential for improved spatial resolution.

Purpose of the Study:

• To evaluate the efficacy of time-frequency (TF) analysis followed by EEG source localization in improving the detection and identification of epileptogenic activity.
• To compare the accuracy of brain electromagnetic tomographic (BET) images generated from TF domain data versus time-domain data.
• To assess the concordance of BET-derived localization with clinical findings in patients with refractory frontal lobe epilepsy (FLE).

Main Methods:

• Seventeen patients with refractory FLE underwent video EEG recording.
• Time-frequency (TF) analysis was applied to identify early epileptogenic EEG changes.
• Bayesian model averaging (BMA) was used to compare BET images from the TF domain and time domain.

Main Results:

• TF analysis successfully identified the initial epileptogenic EEG changes.
• BET images derived from the TF domain provided more concordant and delimited localization of epileptogenic zones.
• TF analysis revealed subtle EEG features, including lateralization, which were less apparent with time-domain analysis.

Conclusions:

• EEG TF analysis significantly improves source localization accuracy for the epileptogenic zone.
• This enhanced localization provides a clearer delineation of seizure origins compared to traditional methods.
• Validated EEG TF analysis holds promise as a valuable noninvasive tool for clinical epileptogenic zone localization.