Abstracts

Rationale: Spikes are the main biomarkers of epilepsy in EEG signal. Yet, they suffer from low specificity since they can also be seen in non-epileptogenic regions. Intracranial EEG (iEEG) studies have shown that spikes propagate across large brain areas, and the onset of this propagation is a more precise epilepsy biomarker than areas of spread. However, iEEG has limited spatial resolution to map the exact path of spike propagation, and the relationship of this propagating activity to abnormalities in the underlying structural connections is still unclear. Here, we used electric source imaging to map the spatiotemporal propagation of spikes, localize its onset and spread, and evaluate the microstructural integrity of the corresponding brain regions in children with drug resistant epilepsy (DRE).

Methods: We analyzed iEEG data from 19 children with DRE who underwent surgery with good outcome. For each patient, we detected spikes and their propagation using an in-house automated algorithm (Fig. 1A) and mapped these propagations using dynamic Statistical Parametric Mapping (dSPM) with a time-sliding window approach (Fig. 1B-C). On each patient’s MRI, we defined three regions: onset, early spread, and entire spread zone (Fig. 1C), and quantified their spatial extent, propagation speed, and volume. We estimated the overlap of these regions with resection volume (%), and their mean distance from resection (d_RES) and clinically defined seizure onset zone (SOZ) (Fig. 1D) that served as gold-standards of the epileptogenic zone. To quantify microstructural abnormalities, we used voxel-based diffusion imaging (n=14 patients) and estimated the laterality ratio for fractional anisotropy (FA), mean diffusivity (MD), radial diffusivity (RD), and axial diffusivity (AD), defined as the ratio between the three aforesaid regions and their mirrored contralateral normal-appearing areas (Fig. 1E).

Results: We observed a spike propagation in all patients. The onset had a more focal spatial extent compared to the early (p=0.03) and entire spread zone (p=0.0003). The propagation speed was lower in the onset compared to the early (p=0.003) and entire spread zone (p=0.03). Higher overlap with resection was seen for the onset (50% [range: 24-82%]) compared to the early (42% [range: 10-74%]; p=0.01) and entire spread zone (34% [range: 7-73%]; p=0.01). The d_RES was shorter for the onset (5 mm [range: 4-11 mm]) compared to the early (6 mm [range: 4-21 mm]; p=0.006) and the entire spread zone (9 mm [4-29 mm]; p=0.002). The MD laterality ratio was higher for the onset and early spread zone compared to the entire spread zone (Fig. 2).

Conclusions: By using iEEG source imaging, we show that the onset of spike propagation is a reliable biomarker of the epileptogenic zone and is also indicative of underlying microstructural abnormalities in contrast to areas of spread. Such an interictal biomarker may augment the presurgical evaluation of children with DRE.

Funding: Please list any funding that was received in support of this abstract.: RO1NS104116-01A1 and R21NS101373-01A1 by NINDS.