Optimization of Concurrent EEG-fMRI to Improve Detection of Epileptogenic Zone in People with Epilepsy
Abstract number :
2.061
Submission category :
3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year :
2021
Submission ID :
1825672
Source :
www.aesnet.org
Presentation date :
12/9/2021 12:00:00 PM
Published date :
Nov 22, 2021, 06:44 AM
Authors :
Po-Tso Lin, MD - Neurological Institute, Taipei Veterans General Hospital, Taipei, Taiwan; Chien-Chen Chou, MS – Department of Neurology – Neurological Institute, Taipei Veterans General Hospital, Taipei, Taiwan; Yen-Cheng Shih, MD – Department of Neurology – Neurological Institute, Taipei Veterans General Hospital, Taipei, Taiwan; Cheng-Jia Lee, PhD – Department of Neurosurgery – Neurological Institute, Taipei Veterans General Hospital, Taipei, Taiwan; Hsiang-Yu Yu, MD – Department of Neurology – Neurological Institute, Taipei Veterans General Hospital, Taipei, Taiwan; Wen-Jui Kuo, PhD – Institute of Neuroscience, National Yang Ming Chiao Tung University, Taipei, Taiwan
Rationale: Concurrent electroencephalogram and functional magnetic resonance imaging (EEG-fMRI) has been used to assist presurgical localization of the seizure focus in people with epilepsy with around 70% accuracy. This study was aimed and set up to improve the detection rate of seizure focus by using an optimized concurrent EEG-fMRI protocol.
Methods: An optimized concurrent EEG-fMRI protocol was used to record EEG and the Blood Oxygenation Level Dependent (BOLD) signal from 14 patients (mean age: 27 years; range: 12-44 years). The optimized protocol employed a fast-fMRI sequence (sampling rate= 10 Hz) with a spare arrangement which allowed a time window of 1.9 sec for EEG recording without radio frequency noise (acquiring one fMRI volume in 0.1 seconds repeated every 2 seconds). Two types of seizure-related EEG patterns were identified, including the spike and rhythmic slow waves. Onsets of the two seizure-related EEG patterns were labeled and used as regressors for modeling the BOLD activities from fMRI data. In order to be able to effectively model the temporal fluctuation of the seizure-related BOLD activity, both the canonical hemodynamic response function (HRF) and its temporal derivatives were incorporated in the SPM statistical model. The BOLD results were validated by the epileptogenic zone determined by intracranial EEG studies or resected brain substrate. Each hemodynamic map was required 10 contiguous voxels and p< 0.001 for the individual analysis.
Neurophysiology