Abstracts

Seizure Onset Zones from Stereo-electroencenphalography Map to a Common Brain Network

Abstract number : 2.333
Submission category : 9. Surgery / 9C. All Ages
Year : 2023
Submission ID : 637
Source : www.aesnet.org
Presentation date : 12/3/2023 12:00:00 AM
Published date :

Authors :
First Author: Zhanqi Hu, MD – Boston Children's Hospital

Presenting Author: Catharina Gout, MD – Boston Children's Hospital

Catharina Gout, MD – Localization Laboratory, Laboratory for Translational Neuroimaging, Computational Radiology Laboratory, Fetal-Neonatal Neuroimaging and Developmental Science Center and Departments of Neurology, Radiology and Pediatrics – Boston Children's Hospital; Juliana Wall, BA – Localization Laboratory, Laboratory for Translational Neuroimaging, Computational Radiology Laboratory, Fetal-Neonatal Neuroimaging and Developmental Science Center and Departments of Neurology, Radiology and Pediatrics – Boston Children's Hospital; Krzysztof Sadowski, MD, PhD – Research Fellow, Localization Laboratory, Laboratory for Translational Neuroimaging, Computational Radiology Laboratory, Fetal-Neonatal Neuroimaging and Developmental Science Center and Departments of Neurology, Radiology and Pediatrics, Boston Children's Hospital; Michelle Chiu, MD, PhD – Fellow, Localization Laboratory, Laboratory for Translational Neuroimaging, Computational Radiology Laboratory, Fetal-Neonatal Neuroimaging and Developmental Science Center and Departments of Neurology, Radiology and Pediatrics, Boston Children's Hospital; Patrick Davis, MD, PhD – Localization Laboratory, Laboratory for Translational Neuroimaging, Computational Radiology Laboratory, Fetal-Neonatal Neuroimaging and Developmental Science Center and Departments of Neurology, Radiology and Pediatrics – Boston Children's Hospital; Eleonora Tamilia, PhD – Localization Laboratory, Laboratory for Translational Neuroimaging, Computational Radiology Laboratory, Fetal-Neonatal Neuroimaging and Developmental Science Center and Departments of Neurology, Radiology and Pediatrics – Boston Children's Hospital; Sylvain Rheims, MD, PhD – Department of Functional Neurology and Epileptology – Hospices Civils de Lyon and University of Lyon; Marc Guenot, MD, PhD – Department of Functional Neurosurgery – Hospices Civils de Lyon and University of Lyon; Scellig Stone, MD, PhD – Localization Laboratory, Laboratory for Translational Neuroimaging, Computational Radiology Laboratory, Fetal-Neonatal Neuroimaging and Developmental Science Center and Departments of Neurology, Radiology and Pediatrics – Boston Children's Hospital; Michael Fox, MD, PhD – Center for Brain Circuit Therapeutics, Departments of Neurology, Psychiatry and Radiology – Brigham and Women’s Hospital; Frederic Schaper, MD, PhD – Center for Brain Circuit Therapeutics, Departments of Neurology, Psychiatry and Radiology – Brigham and Women’s Hospital; Alexander Cohen, MD, PhD – Localization Laboratory, Laboratory for Translational Neuroimaging, Computational Radiology Laboratory, Fetal-Neonatal Neuroimaging and Developmental Science Center and Departments of Neurology, Radiology and Pediatrics – Boston Children's Hospital; Jurriaan Peters, MD, PhD – Neurology – Boston Children's Hospital

Rationale:
Lesion locations associated with epilepsy are part of a human brain network defined by functional connectivity to the basal ganglia, brainstem and cerebellum (Schaper et al, JAMA Neurology 2023). Lesions connected to this brain network pose an increased risk for epilepsy, regardless of etiology. Here, we investigated whether the seizure onset zone (SOZ), as identified by stereo-electroencephalography (sEEG) in patients undergoing epilepsy surgery, is connected to this network.

Methods:
A total of 10,056 contacts of sEEG electrodes in 67 cases (39 adults, 28 children) were localized in common atlas space by co-registration of the post-implantation CT scan with pre-operative MRI. Etiologies included focal cortical dysplasia or other malformations of cortical development (n=22), tuberous sclerosis complex (n=5), stroke (n=8), mesial temporal sclerosis (n=3), other lesional etiology (n=5), and MRI-negative (n=24). A board-certified clinical neurophysiologist classified each sEEG electrode contact as part of the SOZ or not. Recording locations were created by generating a 3 mm radius sphere around the MNI coordinate of each contact. To quantify the connectivity, sEEG electrode locations were run as a seed in a normative resting state fMRI connectome (n=1000). Connections significantly associated with SOZs (two-tailed FDR-corrected p < 0.05) were identified with voxel-wise nonparametric permutation analysis of linear models (PALM) using 2000 permutations. Finally, we computed the connection strength between SOZ and the network nodes by using region of interest-to-region of interest (ROI-to-ROI) connectivity.
Surgery