Abstracts

Rationale: Temporal lobe epilepsy is the most frequent site of origin of focal seizure. However, often time its clinical manifestation is quite similar to extra temporal onset epilepsy1. In recent years, there has been a rapid development in computational studies, including electrical source localization (ESL) methods2, which may help with initial localization. Our aim is to predict the value of ictal onset and initial propagation pattern using scalp EEG source imaging (ESI) analysis for defining the localization and extent of epileptogenic zone, in comparison to intracranial (IC) EEG findings in patient with temporal lobe epilepsy.

Methods: A consecutive cohort of patients with temporal lobe epilepsy, who underwent scalp and IC EEG monitoring at our institution between January 2011-December 2020 were reviewed. Ictal EEG onset and propagation were categorized into 4 groups (Group 0 unable to assess; 1 anterior temporal; 2 posterior temporal; and 3 orbitofrontal region). The onset and propagation sequence were identified for each patient, blinded of the clinical data. IC EEG data were also reviewed independently of the clinical and scalp EEG data. EEG was recorded using the Neuvo system, with a sampling rate of 1000Hz and reviewed with a filter band 1-70Hz. EEG was displayed in a common reference average montage with the removal of bad electrodes. A segment of the containing a typical seizure pattern is then co registered with the image data (Figure 1). Electrodes with high impedance and prominent contamination are removed from the average and from the source analysis. The EEG source data sets were analyzed using Curry 9 software. Filters are applied to narrow the band containing the frequency of the rhythmic burst. Localization is performed at the earliest point where the signal to noise ratio was at least 10 and the confidence ellipsoid is not greater than 5 cm³ within a sub lobar region. Current density is performed at 95% of the current strength at the exact time of the calculated dipole source. Seizure onsets were identified by noting the earliest sustained rhythmic change from baseline EEG activity and early propagation was determined by the direction of the dipole movement as approaching the peak of the ictal spike.

Results: A total of 43 patients were reviewed. Nine were excluded due to lack of scalp EEG data. In the remaining 34, twenty patients (65%) matched ESI scalp and IC ictal onset, and 18 patients (53%) matched the initial propagation region. Anterior temporal onset was associated with a positive predictive value of 77% to concordant scalp ESI onset. Of note, seven out of eight (87.5%) of cases that mismatched both ictal onset and early propagation region involved either orbitofrontal region or posterior temporal region.

Conclusions: Scalp EEG source analysis may have a good predictive value in the localization of ictal onset and early propagation with anterior temporal localization. Orbitofrontal and posterior temporal ictal onset or propagation are not well localized using scalp EEG source imaging.

Figure 1: This figure is showing ictal onset from the right anterior temporal region which propagates into orbitofrontal area.

Funding: Please list any funding that was received in support of this abstract.: None.