Connectivity within regions characterizes epilepsy duration and treatment outcome

Area of Science:

• Neuroscience
• Medical Imaging
• Network Science

Background:

• Identifying reliable biomarkers for temporal lobe epilepsy (TLE), especially in early stages, is challenging.
• Current whole-brain connectome analyses use coarse parcellations, potentially missing localized changes crucial for TLE detection.

Purpose of the Study:

• To investigate high-resolution structural connectomes for detecting localized changes in TLE.
• To assess the relationship between local connectivity alterations and TLE disease duration.
• To evaluate the efficacy of local connectivity as a predictor of surgical outcomes in TLE patients.

Main Methods:

• Constructed high-resolution (~50,000 nodes) structural networks using diffusion tensor imaging in 33 TLE patients and 36 healthy controls.
• Analyzed local connectivity within brain regions and global connectivity between 68 cortical regions.
• Utilized NICARA software for MRI processing, network reconstruction, and visualization.

Main Results:

• Patients with TLE exhibited altered local connectivity (lower clustering coefficient, higher edge density) within regions, which was not observed in global networks.
• Local connectivity changes correlated with disease duration.
• Local connectivity provided superior prediction of surgical outcomes (95.39% accuracy) compared to global connectivity.

Conclusions:

• High-resolution analysis of intra-regional connectivity offers a more sensitive approach to identifying TLE biomarkers than traditional whole-brain connectomes.
• Intra-regional connectivity changes are associated with disease progression and can accurately predict surgical success in medically intractable TLE.