Abstracts

Rationale: The increased spatiotemporal resolution offered by high-density electrocorticography (HD-ECoG) has been shown to have both improved detection of dipolar sources and efficacy in brain-machine interface applications in animal models and human subjects. However, little data exist regarding its utility for identifying epileptiform discharges during intraoperative monitoring and planning. We report our preliminary findings using HD-ECoG in lesional patients. Methods: 24 consecutive patients (13 female) undergoing HD-ECoG during awake craniotomy for resection were studied. Recordings were conducted with 8x8 grids spaced 5 mm with 3 mm electrode diameter and stereotactic depth electrodes with 0.81 mm contacts with 5 mm inter-electrode distance. Results were compared to pre-operative scalp EEG recordings. The presence of epileptiform discharges, after-discharges, seizures, and lateralized periodic discharges (LPDs) were judged by two of the co-authors (WOT and AF). Results: Lesions were neoplastic in 21; there was one patient each with encephalomalacia, cavernous malformation, and mesial temporal sclerosis. Epileptiform discharges during HD-ECoG were identified in 16 (67%). 14 patients (58%) had preoperative clinical seizures. Among these, epileptiform discharges were present in four (17%) on scalp EEG, and in those four plus three more on HD-ECoG (50%). After-discharges were present during electrocortical stimulation in 5/24 (21%), three of whom did not have preoperative seizures. One patient without clinical seizures had both after-discharges and a subclinical seizure identified on HD-ECoG. LPDs were identified in four patients (17%) during resection, three of whom had clinical seizures. Only one patient with LPDs had discharges on scalp EEG. LPDs were identified on as few as two contacts (1 cm). Frequencies ranged from 0.5 – 2.0 Hz. All patients with LPDs had a single structural cerebral lesion, and the LPDs were unifocal in each patient. One patient with left mesial-temporal “LPDs-Plus” (LPDs on depth recording in addition to spiking on grid) reported subjective improvement in memory immediately following resection that was later confirmed on neuropsychological testing. Overall, abnormal findings on the ictal-interictal spectrum were identified in four of ten (40%) patients without preoperative clinical seizures or epileptiform features on scalp EEG. The ECoG findings were judged with 100% concordance by two of the authors. Conclusions: Abnormal epileptiform features were identified in 40% of patients undergoing intraoperative HD-ECoG during awake craniotomy for resection of a structural lesion without preoperative seizures or epileptiform features on scalp EEG. LPDs have not been described on HD-ECoG in the context of awake craniotomy. Their presence may suggest a focal field potential with restricted synchrony to a sub-centimeter network of the brain. The increased spatiotemporal resolution offered by HD-ECoG is sensitive for the detection of subclinical epileptiform features not identifiable by standard EEG methods. Funding: n/a