Electromagnetic Source Imaging Predicts Surgical Outcome in Children with Drug Resistant Epilepsy
Abstract number :
1.11
Submission category :
2. Translational Research / 2C. Biomarkers
Year :
2022
Submission ID :
2204033
Source :
www.aesnet.org
Presentation date :
12/3/2022 12:00:00 PM
Published date :
Nov 22, 2022, 05:23 AM
Authors :
Rupesh Chikara, PhD – University of Texas at Arlington; Saeed Jahromi, MS – Bioengineering – The University of Texas at Arlington; Eleonora Tamilia, PhD – Fetal-Neonatal Neuroimaging and Developmental Science Center – Boston Children’s Hospital, Harvard Medical School; Joseph Madsen, MD – Division of Epilepsy Surgery, Department of Neurosurgery – Boston Children’s Hospital, Harvard Medical School; Steve Stufflebeam, MD – Athinoula Martinos Center for Biomedical Imaging – Massachusetts General Hospital, Harvard Medical School; Phillip Pearl, MD – Division of Epilepsy and Clinical Neurophysiology, Department of Neurology – Boston Children’s Hospital, Harvard Medical School; Christos Papadelis, PhD – Jane and John Justin Neurosciences Center – Cook Children's Health Care System
Rationale: Noninvasive diagnostic tools, such as magnetoencephalography (MEG) and high-density electroencephalography (HD-EEG), are becoming of high importance in the presurgical evaluation of patients with drug resistant epilepsy (DRE). These tools have different sensitivity profiles for detecting and localizing epileptiform foci. MEG has higher spatial resolution for sources that are tangential to the plane of the cortical surface, but it is almost blind to radially oriented sources and sources from deep brain areas. Contrarily, HD-EEG can detect and localize both radially oriented as well as tangentially oriented sources. Despite their complementary properties, MEG and HD-EEG are rarely recorded simultaneously or analyzed into a combined source imaging solution. Here, we assess the accuracy and clinical utility of combined electromagnetic source imaging in localizing interictal epileptiform discharges (IEDs) and predict surgical outcome in children with DRE.
Methods: We retrospectively analyzed MEG and HD-EEG data, which were simultaneously recorded, from 23 children (12 females, mean age: 12.91±4.07; range: 5–18 years) with DRE associated with focal cortical dysplasia who underwent intracranial EEG and surgery. We dichotomized our patients into good (Engle 1; 14 patients) and poor (Engel ≥2; 9 patients) outcome. We initially identified IEDs commonly seen in both MEG and HD-EEG (Figure 1A). We constructed a realistic head model from each patient’s pre-operative MRI (Figure 1B). We then localized the underlying generator of IEDs using the equivalent current dipole (ECD) and an in-house dipole clustering method (Figure 1D). Source localization was performed for the EMSI, ESI, and MSI. For ECDs and clustered dipoles, we calculated the distance from the clinically defined seizure onset zone (DSOZ) (Figure 1E) and resection (DRES) (Figure 1F). Finally, we estimated receiver operating characteristic (ROC) curves for predicting outcome for the EMSI, MSI, and ESI (Wilcoxon signed-rank test).
Results: EMSI presented shorter DSOZ (15.18±9.06 mm, Figure 2A) compared to individual modalities [ESI: 25.04±16.20 mm, p< 0.01; MSI: 23.37±8.98 mm, p< 0.05] for dipole clustering in patients with good outcome. Similarly, EMSI revealed shorter DRES (8.56±6.24 mm; Fig. 2c) compared to individual modalities [ESI: 18.47±17.32 mm, p< 0.05; MSI: 15.51±10.11 mm, p< 0.02). No significant differences in DSOZ or DRES between modalities were observed for poor outcome patients (Figures 2B, 2D). For all modalities, clustering showed lower DRES compared to ECDs for patients with good outcome (p < 0.05; Figure 2C). ROC analysis revealed higher (0.81, Fig. 2f) than ESI (0.70) and MSI (0.61). ECDs showed lower AUC compared to clustering (Figures 2E, 2F).
Translational Research