Relating Cortical Atrophy in Temporal Lobe Epilepsy with Graph Diffusion-Based Network Models.

Area of Science:

• Neuroscience
• Computational Biology
• Medical Imaging

Background:

• Mesial temporal lobe epilepsy (TLE) exhibits characteristic patterns of seizure origination and spread.
• Neuronal atrophy distribution in TLE, observable via MRI, mirrors these epileptogenic activity patterns.
• Both activity and atrophy spread appear to correlate with white matter connectivity.

Purpose of the Study:

• To model the networked spread of activity and atrophy in TLE using first principles.
• To compare two network diffusion models: one where atrophy follows hyperactivity, and another of progressive degeneration.
• To validate these models against gray matter atrophy distributions in TLE patient cohorts.

Main Methods:

• Developed two first-order network diffusion models for TLE.
• Modeled atrophy distribution as a consequence of epileptogenic activity propagation.
• Modeled atrophy as a progressive degenerative process.
• Validated models against volumetric gray matter atrophy data from TLE-MTS and TLE-no cohorts.

Main Results:

• Network models closely reproduced observed regional gray matter atrophy distributions.
• High correlations were found between model predictions and measured atrophy (R = 0.586 for TLE-MTS, R = 0.283 for TLE-no).
• The atrophy spread model demonstrated superior performance compared to the hyperactivity spread model.

Conclusions:

• Network diffusion models effectively capture atrophy patterns in TLE.
• The atrophy spread model is more accurate than the hyperactivity spread model for TLE.
• These findings support potential clinical applications in predicting atrophy spread, identifying seizure onset zones, and surgical planning for TLE patients.