Synchronization Clusters of Interictal Activity in the Lateral Temporal Cortex of Epileptic Patients: Intraoperative Electrocorticographic Analysis

Abstract

SUMMARYObjective: Drug-resistant temporal lobe epilepsy (TLE) can be treated by tailored surgery guided by electrocorticography (ECoG). Although its value is still controversial, ECoG activity can provide continuous information on intracortical interactions that may be useful to understand the pathophysiology of TLE. The goal of this study is to characterize local interactions in multichannel ECoG recordings of the lateral cortex of TLE patients using three synchronization measures and to link this information with surgical outcome. Methods: Intraoperative ECoG recordings from 29 TLE patients were obtained using grids of 20 electrodes (4 × 5) covering regions T1, T2, and T3 of the lateral temporal lobe. Linear correlation, mutual information, and phase synchronization were calculated to quantify lateral intracortical interactions. Surrogate data files were generated to test results statistically. Results: By distributing locally the interactions between the electrodes, we characterized the spatial patterns of ECoG activity. We found clusters of synchronized activity at specific areas of the lateral temporal cortex in most patients. Methodologically, linear correlation and phase synchronization performed better than mutual information for cluster discrimination. ROC analysis suggested that surgical removal of sharply defined synchronization clusters correlated with seizure control. Conclusions: Our results show that synchronous intraoperative ECoG activity emerges from specific cortical areas that are highly differentiated from the rest of the temporal cortex. This suggests that synchronization analysis could be used to functionally map into the temporal cortex of TLE patients. Moreover, our results suggest that these sites might be involved in the circuits that participate in clinical seizures.

Hyper-Synchronization, De-Synchronization, Synchronization and Seizures

“…If, at is believed (Kandel et al, 2000), the spread of seizures follows normal cortical circuits, it would be expected that seizure propagation will be desynchronized, at least at the onset stage, due to the synchronization inhomogeneity of the cortical substrate. Moreover, interictal synchronization inhomogeneities have been reported in the lateral cortex (Ortega et al, 2008) in TLE patients. Therefore, it is highly expected that during the seizure onset, the seizure spreading, both in the mesial and in the lateral side of the temporal lobe, display a desynchronized activity, at least through the intracortical propagation.…”

“…Cross-correlation offers both properties. Moreover, the discovery that cross-correlation, although a linear statistic, perform as well as nonlinear synchronization measures when applied to neurophysiological data (Netoff & Schiff, 2002;Quian Quiroga et al, 2002: Ortega et al, 2008 justifies in most cases its first use, at least as an initial exploratory task. As is well-known in the field of time series analysis, it is always mandatory to employ a full battery of methods to estimate an underlying quantity.…”

“…The whole same procedure is able to be extended in the case of synchronization between time series. When dealing with multivariate records, firstly, a measure of synchronization is chosen, typically Pearson coefficient (Zhou et al,2006), Equation (13), though other measures are also used (Ortega et al, 2008;Ortega et al, 2010) to check the results' reliability, and then a transformation to distances is performed. Then, a favorite clustering algorithm is used in order to create the hierarchical organization, which will uncover the synchronization clusters present in the data.…”

New network and synchronization approaches in focal epilepsy research and treatment

“…In order to dig further into the above findings we have analyzed another sample of ECoG data from 20 drugresistant TLE patients [22]. In this case, however, a mesial strip of 8 electrodes was also located through the Sylvian fissure recording electrical activity in the mesial area of the temporal lobe.…”

Een hersentumor: lokale beschadiging of ziek netwerk?

“…de oorsprong) van de epilepsie niet wordt weggehaald, waardoor patiënten epileptische aanvallen blijven houden. Door met behulp van de netwerktheorie vast te stellen welke punten in het hersennetwerk het belangrijkst zijn voor het ontstaan en verspreiden van de aanvallen, kan de uitkomst van de operatie verbeterd worden (Ortega et al, 2008a;2008b).…”