Graph Measures of Node Strength for Characterizing Preictal Synchrony in Partial Epilepsy

Area of Science:

• Neuroscience
• Clinical Electrophysiology
• Computational Neurology

Background:

• The rapid discharge on electroencephalography (EEG) signifies seizure onset, often with decreased brain synchrony.
• Preictal periods can show increased synchrony, quantifiable using network measures, potentially indicating seizure origin.

Purpose of the Study:

• To compare preictal synchrony network measures with the epileptogenicity index (EI) for quantifying seizure onset patterns.
• To evaluate the effectiveness of node strength measures in detecting regions with high EI values.

Main Methods:

• Analysis of 24 seizures from 12 patients using stereotaxic EEG (SEEG).
• Computation of pairwise nonlinear correlation (h(2)) and node strength (IN, OUT, TOT) in the preictal period.
• Receiver operating characteristic (ROC) analysis to assess the detection capacity of strength measures against EI, using signals filtered in the 15-40 Hz band.

Main Results:

• Node strength measures (OUT and TOT) showed the best correspondence with EI in the 15-40 Hz band.
• ROC analysis performance improved when considering seizures with visible preictal synchrony.
• Preictal connectivity graph strength effectively detected regions with high EI.

Conclusions:

• Measuring preictal connectivity graph strength provides valuable clinical information for identifying the epileptogenic zone.
• Network analysis of preictal EEG synchrony offers a promising method for localizing seizure origins.