Abstracts

Rationale: In EEG and MEG analysis for pre-surgical evaluations, equivalent current dipoles (ECD) are routinely used to model spike and seizure onsets. An ECD represents in reality at least 4 cm (MEG [1]) or 10 cm (EEG [2]) of activated cortex, and these dipoles are not necessarily located within the generator cortex. A critical task in interpreting ECD results is estimating which area of nearby cortex is involved. Cortical Current Density Reconstructions (CDRs), that take into consideration both neuronal orientations and spatial coupling of nearby neurons, provide images of the most likely source cortex and thus can help to interpret ECD findings and resolve associated ambiguities. The commonly observed posterior displacement of MEG ECDs relative to their EEG counterparts in anterior temporal spike recordings [3] serves as an example of how CDR analysis can resolve apparent differences. Methods: 150 patients were recorded simultaneously with 306 channels of MEG (Elekta Triux) and 25 channels of EEG (standard 10-20 plus 3 subtemporal channels bilaterally). Spikes of similar voltage and/or magnetic field topography were averaged. Realistic boundary element head models were created based on patients' individual MRIs. ECDs and cortical standardized Low Resolution Electromagnetic Tomography (sLORETA) CDRs were calculated simultaneously for both EEG and MEG data. CDRs used cortex orientation constraints and spatial coupling of neighboring areas to account for the spatially extended nature of epileptiform discharges [4]. Regularization was adjusted according to the Chi-criterion, so that the fit deviation matched the amount of noise in the data. All data analysis was performed using the Curry 8 software (Compumedics, Charlotte, NC, USA). Results: Ten patients were identified with EEG ECDs in the temporal lobe tip that had counterpart MEG ECDs posteriorly displaced by two or more centimeters. This source location ambiguity was resolved in all cases by cortical CDR that co-localized both EEG and MEG generator cortex to the tip of the temporal lobe. Conclusions: Both ECDs and cortical CDR can provide models of brain sources that explain resultant EEG and MEG. Cortical CDR, however, can take the orientation of existing cortical areas and their spatial coupling into account, thus delivering a more realistic source model of EEG and MEG. Cortical CDR yields extended patches of cortex whose simultaneous activity explains the given data. These cortical activations do not necessarily co-localize with ECD results. Situations where dipole results are ambiguous regarding source cortex or where EEG and MEG dipoles do not co-localize can be resolved by using CDR methods. [1] Oishi M, Otsubo H, Kameyama S, Morota N, Masuda H, Kitayama M. Epilepsia 2002, 43:1390?"5 [2] Tao JX, Ray A, Hawes-Ebersole S, Ebersole JS. Epilepsia 2005, 46:669-76 [3] Ebersole JS, Ebersole SM. J. Clin Neurophysiol 2010, 27:360-71 [4] Wagner M, Fuchs M, Kastner J. in: BIOMAG 2002. Eds. Nowak H et al., VDE, Berlin 2002, 804-6 Funding: n/a