EPILEPTIC NETWORKS USING SCALP AND INTRACRANIAL EEG-FMRI AND POSTSURGICAL OUTCOME
Abstract number :
1.187
Submission category :
5. Neuro Imaging
Year :
2013
Submission ID :
1750217
Source :
www.aesnet.org
Presentation date :
12/7/2013 12:00:00 AM
Published date :
Dec 5, 2013, 06:00 AM
Authors :
U. J. Chaudhary, S. Perani, D. Carmichael, R. Rodionov, S. Vulliemoz, R. Thornton, M. Pugnaghi, C. Micallef, A. McEvoy, C. Scott, B. Diehl, M. Walker, J. Duncan, L. Lemieux
Rationale: We aim to determine the distribution and assess the potential significance of spike-related haemodynamic changes using scalp and intracranial EEG-fMRI and their relation to postsurgical outcome. Methods: Fourteen patients with refractory focal epilepsy undergoing presurgical evaluation were offered scalp [1] and intracranial EEG-fMRI [2]. Spikes were identified and classified according to their spatiotemporal distribution. For patients who did not have spikes during scalp-EEG-fMRI, topographical map correlation for spikes was calculated between scalp-EEG inside and outside the scanner. Spike-related BOLD changes were assessed over the whole-brain using general linear modelling. Anatomical localization, distribution and concordance of changes in BOLD maps for different spike-types were compared with location of the invasively-defined irritative-zone (IZ) and their relationship to postsurgical-outcome. We evaluated the association between different spike-types, BOLD-changes in deep brain structures, degree-of-concordance [1] (Entirely-concordant: all BOLD-clusters in the IZ; Concordant-plus: global statistical maximum BOLD cluster in the IZ; Some-concordance: one of the BOLD cluster in the IZ) and postsurgical outcome.Results: 11/14 patients had scalp-EEG-fMRI. Thirteen different spike-types were identified on scalp EEG inside and outside the scanner in 9/11 patients and significant BOLD-changes (Figure 1) were seen for 12/13 spike-types. Spike-related BOLD-maps had a degree-of-concordance with the IZ for 6/12 spike-types and were discordant for 6/12. All 14 patients had intracranial-EEG-fMRI and 53 different spike-types were identified in 13/14 patients. Significant BOLD-changes were seen for all spike-types (Figure 1) and maps had a degree-of-concordance with the IZ for 31/53. Spike-related BOLD-changes were seen in the IZ, healthy neocortex and in the deep brain structures. Deep brain changes were associated with discordant BOLD maps (rs = 0.3, p < 0.05). 11/14 patients underwent surgery and postsurgical-outcome was assessed at a minimum duration of 3months (range:3-45). Combined-interictal-BOLD-map (for all spike-types in an individual patient) revealed BOLD-clusters nearest to the IZ within the resected area in some patients and at a distance of 0.2 to 6cm from the resection-margin in others. Spike-related BOLD-maps for different spike-types had a higher degree of concordance with the irritative zone for patients with ILAE class-I postsurgical-outcome (Scalp-EEG-fMRI: rs = 0.8, p < 0.05; intracranial-EEG-fMRI: rs = 0.4, p < 0.05) as compared to patients with ILAE class-IV postsurgical-outcome (Figure 2). Patients having combined-interictal-BOLD-maps with a degree-of-concordance with the IZ for both scalp and intracranial-EEG-fMRI had ILAE class-I postsurgical-outcome.Conclusions: Scalp and intracranial-EEG-fMRI can map spike-related BOLD-changes in the neocortex and deep brain structures for different spike-types and their distribution and degree-of-concordance with the IZ may explain postsurgical-outcome. 1. Chaudhary et al, Brain, 2012 2. Carmichael et al, Neuroimage, 2012
Neuroimaging