Abstracts

Rationale: Electric source imaging (ESI) of interictal epileptic discharges using high-density EEG is a prospectively validated method to help localizing the epileptogenic zone in candidates for epilepsy surgery. However, its accuracy relative to the preoperative gold standard of intracranial EEG (icEEG) has not been reported. Furthermore, it is debated whether using interictal activity, as ESI does, in order to localize the source of ictal activity is adequate. Methods: Retrospective study of 35 candidates to epilepsy surgery who underwent icEEG monitoring. ESI was performed using 128- or 256-electrode EEG, a 3-shell spherical model with anatomical constraints (SMAC) and a distributed linear inverse solution (LAURA). Results: The median distance from the ESI maximum to the nearest intracranial electrode involved in an irritative zone (IZ) was 19 mm (interquartile range 12-23 mm), that to the most irritative electrode 30 mm (26-48 mm), that to the nearest electrode in the seizure onset zone (SOZ) 22 mm (14-30 mm). The ESI-to-nearest IZ and ESI-to-maximal IZ distances were strongly correlated with the ESI-to-SOZ distance (p<0.0001 and p<0.01 respectively). There were no significant differences in ESI accuracy in patients with medial temporal lobe, lateral temporal lobe or extratemporal epilepsy (Figure 1). The ESI maximum was included in the resected brain volume more often in patients with favorable postoperative outcome (9 of 18 patients in Engel classes I and II) than in those with unfavorable outcome (1 of 8 patients in Engel classes III and IV; difference in proportions 0.375, 95% confidence interval 0.0278 to 0.6667). Furthermore, the maximal IZ electrode was more often within the resected brain volume in the patients with favorable outcome (15 of 18 patients) than in those with unfavorable outcome (3 of 8; difference in proportions 0.4583, 95% confidence interval 0.0833 to 0.8194, p < 0.05). Conclusions: Put together, our results indicate that the localization of interictal spikes by ESI closely corresponds to the localization of irritative zones by intracranial EEG; that irritative zones are often part of the seizure onset zone or lie close to it; and that localizing interictal spikes by ESI helps localizing the seizure onset zone. Additionally, including the ESI maximum in the resected brain volume is correlated with favorable postoperative outcome. Importantly, ESI performs accurately regardless of the epilepsy subtype.