Abstract:ObjectiveWe describe a novel method to spatially map interictal epileptiform discharges (IEDs) through voxel-wise functional connectivity analysis of the functional magnetic resonance imaging (fMRI) portion of simultaneous electroencephalography (EEG)-fMRI data. This method measures the local synchronicity of fMRI signals associated with IED and, in contrast to conventional methods, does not require modeling of neural activities or hemodynamic response. MethodsSimultaneous EEG-fMRI was performed on six patients with focal epilepsy. IED events were detected from the EEG data. The fMRI data was subdivided into time segments of 20s in length, and then reorganized into one set of concatenated time series containing the IED events and many sets without IEDs. Local degree centrality (LDC), a metric of functional connectivity, was computed for each brain voxel to summarize its signal correlations to brain voxels within 14mm of physical distance. This computation was repeated for each set of concatenated time series, yielding one whole-brain LDC map for time with the IED events and many maps for time without IED. A statistical score was computed for each voxel to detect the voxels with significant LDC value differences associated with IEDs. The fMRI data were also processed separately by conventional methods for comparison. ResultsIn all six patients, regions with significant LDC increase during IEDs were concordant in location to both simultaneous EEG and the epileptogenic focus determined from separate clinical studies. In contrast, results from the conventional methods were concordant in only three patients. SignificanceWe show that for focal epilepsy, voxel-wise functional connectivity analysis of EEG-fMRI data may improve IED localization and EEG concordance compared to the conventional analysis. This new analytic method may improve the robustness of interictal EEG-fMRI as a technique for mapping the epileptogenic focus, and helps study the local synchronization aspect of the epileptic network.

Improved mapping of interictal epileptiform discharges with EEG-fMRI and voxel-wise functional connectivity analysis.