MEG language lateralization in partial epilepsy using dSPM of auditory event-related fields

Area of Science:

• Neuroscience
• Medical Imaging
• Epilepsy Research

Background:

• Determining hemispheric language dominance is crucial for epilepsy surgery planning.
• Magnetoencephalography (MEG) methods for language dominance vary, impacting results.
• Previous studies often used dipole modeling of event-related fields (ERFs).

Purpose of the Study:

• To evaluate if distributed source modeling with MEG can replicate prior language dominance findings.
• To compare MEG-based language laterality with functional MRI (fMRI) in epilepsy patients.

Main Methods:

• Analyzed MEG data from 45 epilepsy patients performing an auditory word-recognition task.
• Utilized dynamic statistical parametric mapping (dSPM) for source imaging of auditory ERFs.
• Calculated language laterality indices (LIs) in four regions of interest (ROIs) and compared with fMRI results.

Main Results:

• The parietal region (angular and supramarginal gyri) showed the most lateralized MEG responses.
• MEG-fMRI concordance for language dominance was highest in the parietal ROI (69%).
• Discordances were noted, with MEG sometimes favoring atypical language lateralization in patients with right-hemispheric seizure origins.

Conclusions:

• Distributed source modeling (dSPM) of MEG ERFs provides language laterality estimates comparable to previous methods.
• MEG-fMRI concordance is highest in the parietal region for auditory word recognition tasks.
• Epileptic network laterality may influence MEG language laterality estimates, warranting further investigation.