Authors :
Presenting Author: Ludovica Corona, PhD – The University of Texas at Arlington
Hmayag Partamian, PhD – The University of Texas at Arlington
Cynthia Keator, MD – Cook Children's
Linh Tran, MD – Cook Children's Health Care System
Saleem Malik, MD – Cook Children's Health Care System
Dave Shahani, MD – Cook Children's Health Care System
M. Scott Perry, MD – Cook Children’s Physician Network
Christos Papadelis, PhD – Cook Children's Health Care System
Rationale:
Invasive studies used normative power maps to detect abnormalities linked to the epileptogenic zone (EZ) and outcomes in patients with drug-resistant epilepsy (DRE). Yet, power may fail to capture key functional interactions, leading to inaccurate EZ localization. Functional connectivity (FC) may better map these networks by identifying brain regions whose resection leads to seizure freedom. We aim here to show that FC deviations from normative maps of typically developing (TD) children identify more accurately the EZ than power. We hypothesize that patients with DRE show higher FC than TD children, epileptic hemispheres/lobes have higher FC than non-epileptic ones, and regions with abnormal FC better localize the EZ than power.
Methods:
We analyzed resting-state high-density EEG and magnetoencephalography data from 48 TD and 48 children with DRE (Fig. 1A). Electromagnetic source imaging was applied on artifact-free data to reconstruct source time-series in 166 regions of interest (ROIs) (Fig. 1B). We computed undirected FC [Amplitude Envelope Correlation (AEC) and corrected imaginary Phase Locking Value (ciPLV)] and power at the source-level across bands and kept the strongest values over time (Fig. 1C-D). For patients with DRE, FC and power |z-scores| were computed using normative maps (Fig. 1E) and their global values as baseline to detect abnormal ROIs. Abnormal and non-abnormal subnetworks were defined via z-scores and characterized by global efficiency, node strength, and centrality metrics (Fig. 1F). For patients with focal DRE (fDRE), we defined EZ hemisphere/lobes/ROIs via surgical notes and grouped them by outcome (good: Engel 1; poor: Engel >1). We compared global FC and power between groups, and FC and power between epileptogenic and non-epileptogenic hemispheres of patients with fDRE. FC and power were also compared across EZ, ipsilateral non-EZ, and contralateral non-EZ lobes. FC global efficiency was compared between abnormal and non-abnormal subnetworks of patients with fDRE and those of TD. Finally, we compared FC and power between EZ and non-EZ ROIs in good- and poor-outcome patients.
Results:
Global FC was higher in patients with DRE than TD in several bands (p<