Abstracts

Rationale: This study evaluates the utility of 3D localization of interictal spike activity in electroencephalographs (EEG) superimposed on magnetic resonance imaging (MRI) in a pediatric population. Source localization analyses selected spikes by calculating the inverse problem that produces a dipole source. Functional connectivity analysis was performed to validate propagation of spikes. 3D software based on the CURRY platform was adapted for analyzing scalp EEG data and reconstructing superimposed in pre-surgical evaluation cases. The results of EEG source localization (ESL) together with the respective connectivity maps were assessed in direct relation to intracranial EEG data. Methods: ESL is performed using EEG, MRI, and the 3D electrode data. This includes intermediate steps: (1) determining and storing 3D coordinates of EEG electrodes; (2) 3D cranial shape analysis in 3D for correlation with data from imaging; (3) obtaining SNR; (4) Applying Principal Component Analysis (PCA) for the reduction of redundancy ; (5) performing Independent Component Analysis (ICA) to obtain EEG independent components related to spike generators; (6) calculating white matter volume for 3D rendering ; and (7) assessing EEG connectivity maps at different stages. Results: The ESL was performed in 22 patients with refractory focal epilepsy. Scalp interictal EEG recordings from a 21-electrode were analyzed. The spatiotemporal modeling was represented by moving and rotating dipoles. ESL results were constrained to the cortical gray matter and based on the midway point of the spike's upswing phase. The results showed high concordance between ESL location and the suggested epileptogenic zone. EEG connectivity maps were also obtained to validate the propagation of the given source and to assess the degree of frequency association between two different time series (Figure 3). This measure can reveal subtle aspects of the network dynamics of the brain which complement the results obtained by previous analyses. A typical clinical is shown in Figure 1. Figures 2 and 3 shows the brain connectivity maps in relation to the 3D source results. Conclusions: This study evaluated 3D source localization of epileptic spikes and their EEG-derived connectivity maps in focal epilepsy. This approach represents cerebral sources in an accurate and non-invasive way, optimizing the placement of intracranial electrodes in surgical candidates. Attention must be paid to the ICA components selected to reconstruct the EEG data, so only sources related to the spike generators are utilized in the 3D source analysis and to the flow and direction of information given by the EEG connectivity maps. EEG functional connectivity may help describe 3D source propagation, with subsequent improvement in localization of the epileptogenic zone.