Abstracts

THE EPILEPTIC NETWORK IN SPACE AND TIME: SIMULTANEOUS ELECTRICAL SOURCE IMAGING AND FMRI

Abstract number : 3.15
Submission category : 5. Human Imaging
Year : 2008
Submission ID : 8579
Source : www.aesnet.org
Presentation date : 12/5/2008 12:00:00 AM
Published date : Dec 4, 2008, 06:00 AM

Authors :
Serge Vulliemoz, R. Thornton, R. Rodionov, D. Carmichael, M. Guye, L. Spinelli, John Duncan, C. Michel and L. Lemieux

Rationale: EEG-correlated fMRI (EEG-fMRI) can localize epileptogenic activity and study epileptic networks. In its current form, this imaging technique is characterized by high spatial resolution and low temporal resolution. The combination of EEG-fMRI with EEG source imaging (ESI) may help distinguish between BOLD clusters associated with initiation vs propagation of interictal epileptic activity. Methods: EEG-fMRI data (3T MRI scanner, 32-channel EEG) acquisition and analysis were performed on 5 patients with cryptogenic focal epilepsy. Interictal discharges (IED) were recorded during EEG-fMRI as described previously (Salek-Haddadi et al., Brain Res. 2006). In summary, after correcting the EEG for MRI gradient and pulse artefacts, IED were identified by consensus of two expert observers and used to analyse the fMRI data using SPM5 (www.fil.ion.ucl.ac.uk/spm/). F-maps (FWE corrected) showed regional IED-related changes of Blood Oxygen Level-Dependent (BOLD). ESI was performed on the same set of IED and based on the distributed linear inverse solution algorithm LAURA (Michel et al., Clin. Neurophys., 2004) using an individual realistic head model based on the SMAC approach (Spinelli et al., Brain Topogr. 2000). In some patients, EEG was also recorded outside the scanner. Results: There was good agreement between the ESI results based on EEG recorded inside and outside the scanner. ESI during fMRI: At the first period of stable EEG topography (IED onset), there was good anatomical concordance between ESI and one BOLD cluster, thus localising the initiation of the IED. The distance between maxima of ESI and nearest BOLD cluster was less than 22 mm. In later time frames, ESI maxima were in close proximity to other BOLD clusters than the one identified at IED onset. Conclusions: Simultaneous ESI and EEG-fMRI analysis allows identification of the BOLD changes related to spike initiation and propagation. The combination of ESI and fMRI may provide new insights into the organisation of epileptic networks, particularly in cases where there is good spatial concordance.
Neuroimaging