Abstracts

Rationale: Bottom-of-sulcus dysplasia (BOSD) is a highly localized form of type II focal cortical dysplasia in which neuroimaging, pathological and genetic abnormalities are maximal in the bottom of a single sulcus, tapering to a normal gyral crown. There is evidence that corticectomy limited to the dysplastic sulcus is sufficient for seizure control. This study investigates the distribution of epileptogenicity in BOSD using automated spike and HFO detection and phase amplitude coupling (PAC) methods applied to intraoperative electrocortiography (ECoG).

Methods: 25 patients with BOSD who underwent one-stage resective surgery guided by neuroimaging and ECoG between 2013-2021 were included. Intraoperative ECoG (sampling rate 2000 Hz) was performed under isoflurane and remifentanil anaesthesia with a depth electrode in the dysplastic sulcus and a strip electrode over the gyral crown (GC) and adjacent gyrus (AG) (Fig.1). The location of strip electrode channels (bipolar) was assigned as GC and/or AG using intraoperative photographs projected onto a 3D gyral surface render of the pre-operative MRI. The location of depth electrode channels was assigned to the bottom-of-sulcus (BOS), top-of-sulcus (TOS) or outside of the sulcus using a probabalistic method which determined all possible trajectories between the surface entry point and voxel coordinates within a manual segmentation of the sulcus on MRI. Delphos (Detector of ElectroPhysiological Oscillations and Spikes) software (PloS one.2017;12:e017402) was used to calculate the rate of spiking, ripples and fast ripples (FR) in each location. PAC was calculated at delta, theta, alpha and beta phases with gamma and ripple amplitude, using previously described methods (Scientific Reports.2020;10:14654). Generalised linear and linear mixed models were used to compare each neurophysiologic feature among the five locations.

Results: Across all channels (n=526), the mean occurrence rates (per min) were spikes 10.33 (SE: 0.75), ripples 1.87 (SE: 0.17) and FR 0.91 (SE: 0.18). A significant main effect of location on spike rate, ripple rate and mean PAC in gamma amplitude, particularly for theta/gamma coupling, was seen (all p< 0.05). All three features were maximal at the BOS, with spike rate and mean PAC being significantly greater in the BOS than all other locations (Fig. 2A, 2C). Additionally, spike rate and mean theta/gamma PAC were significantly greater in the TOS channels compared to GC channels (