Clinical Value of Magnetoencephalography and Functional Magnetic Resonance Imaging for the Prediction of the Surgery Outcome
Abstract number :
2.2
Submission category :
5. Neuro Imaging / 5B. Functional Imaging
Year :
2022
Submission ID :
2204847
Source :
www.aesnet.org
Presentation date :
12/4/2022 12:00:00 PM
Published date :
Nov 22, 2022, 05:26 AM
Authors :
Christophe Grova, PhD – Concordia University, McGill University; Edouard Delaire, MSc – PhD student, Concordia University; Chifaou Abdallah, MD – Multimodal Functional Imaging Lab, Biomedical Engineering Dpt, McGill University, Montreal, Canada; Aura Gonzalez Ramirez, MD – Multimodal Functional Imaging Lab, Biomedical Engineering Dpt, McGill University, Montreal, Canada; Jeff Hall, MD, PhD – Montreal Neurological Institute and Hospital, Neurology and Neurosurgery Dpt, McGill University, Montreal, Canada; Francois Dubeau, MD – Montreal Neurological Institute and Hospital, Neurology and Neurosurgery Dpt, McGill University, Montreal, Canada; Birgit Frauscher, MD, PhD – Analytical Neurophysiology Lab, McGill University, Montreal, Canada; Jean Gotman, PhD – Montreal Neurological Institute and Hospital, Neurology and Neurosurgery Dpt, McGill University, Montreal, Canada; Eliane Kobayashi, MD, PhD – Montreal Neurological Institute and Hospital, Neurology and Neurosurgery Dpt, McGill University, Montreal, Canada
Rationale: Magnetoencephalography (MEG) and simultaneous Electro-encephalography-functional magnetic resonance imaging (EEG-fMRI) are powerful tools to localize the epileptic focus1. In our recent work2 comparing MEG and fMRI to intracranial stereo-EEG, we found that EEG-MEG more accurately localizes the primary-irritative zone, whereas EEG-fMRI was more sensitive to the seizure-onset zone. Here, our objective was to compare EEG-MEG and EEG-fMRI localization to the surgical cavity and to evaluate how each modality could predict surgical outcomes. _x000D_
Methods: A total of 30 patients with focal drug-resistant epilepsy who underwent MEG, EEG-fMRI, and subsequent open brain surgery between 2008 and 2019 were included. All patients had available postoperative imaging and a post-operative follow-up > 1 year: 13 seizure-free (SF - Engel Ia), 17 non-seizure-free (NSF -Engel Ib-IV). For MEG, we applied Magnetic source imaging maps (MSI) using the Maximum Entropy on the Mean framework and consensus map to obtain the cluster exhibiting the largest number of spikes.4 Similarly, we selected only the fMRI cluster exhibiting the largest absolute t-value (positive or negative BOLD response).3 For each patient, the surgical cavity was visually identified and manually drawn on postoperative data and considered as ground truth for the epileptogenic zone (EZ). MSI and fMRI results were quantitatively assessed with the surgical cavity projected along the cortical surface, using two metrics: the minimum Distance Localization Error (DLE) between the maximum MSI/fMRI localization and the closest edge of the cavity, the spatial dispersion (SD) quantifying the spatial spread of MSI/fMRI localization around the cavity. Based on the spatial overlap with the cavity, we finally evaluated the performance of each modality to predict the surgical outcome based on sensitivity and precision._x000D_
Results: A total of 27 patients had MSI results, whereas EEG/fMRI resulted in significant BOLD responses in 18 patients (Fig. 1-2). For MSI, SD and DLE were found significantly smaller for SF patients when compared to NSF (p < 0.05). For fMRI, DLE and SD values were not able to discriminate between SF and NSF patients. Overall, we also found smaller DLE and SD values for MSI when compared to fMRI. When considering spatial overlap with the surgical cavity to predict the surgical outcome, MSI reached a precision and sensitivity of 63% and 56% whereas fMRI exhibited precision and sensitivity of 33%. _x000D_
Neuro Imaging