Abstracts

Rationale: One-third of epilepsy patients suffer from seizures that are not controlled by medication, and seizure-free rates range from 30-70% for patients who undergo surgical resection. Surgical outcomes might be improved with more accurate localization of epileptogenic tissue. We hypothesize that high-density microelectrodes will more reliably detect localized interictal markers of the epileptogenic zone than conventional macroelectrodes.

Methods: We performed intra-operative micro-electrocorticogram (µ-ECoG) recordings in anesthetized (propofol/remifentanil/dexmedetomidine) epilepsy patients undergoing craniotomy and in non-epilepsy controls. Recordings were obtained using one of two liquid crystal polymer (LCP) microelectrode arrays: one device offered greater spatial coverage (256 channels, 1.72 mm pitch, 38x21 mm² coverage), and the other had increased contact density (244 channels, 0.76 mm pitch, 12x12 mm² coverage). Both devices featured contact diameters of 0.2 mm. µ-ECoG recordings were collected at 20 kHz and bandpass-filtered between 0 and 300 Hz using a multitaper filter. Channels with high impedance were excluded from analysis. A line-length detector was used to screen recordings for candidate events. A board-certified epileptologist (D.F.) and a trained reviewer (J.S.) labeled events as microseizures if they met three criteria: 1) paroxysmal start from a quiet baseline, 2) evolution in frequency over time, and 3) return to baseline.

Results: We performed intraoperative recordings (mean duration: 13.5 min, range 5.0-53.7 min) in seven patients with focal epilepsy (age: 38.7±6.2 yrs, sex: 4 male and 3 female) and two patients with Parkinson’s disease (age: 68.0±8.5 yrs, sex: 2 female). In two epilepsy patients, recordings from non-involved cortex were also collected. A total of 108 microseizures were detected in 6/7 epilepsy patients (Fig 1). One of the patients without epilepsy had two microseizures, while the other had none. No microseizures were identified in the recordings of non-involved cortex from epilepsy patients (Fig 2). As a group, the epilepsy patients had a significantly higher microseizure rate at 2.33/min compared to 0.01/min for the non-epileptic controls (Mann-Whitney U test, P=0.0303). By spatially averaging adjacent signals, we constructed virtual electrodes of varying diameters. We found a decrease in microseizures detected (64.7%, 84.3%, 89.2%) with increasing contact diameter (1.7, 2.4, 3.2 mm, respectively). By removing contacts, we simulated arrays with different spatial resolutions. We observed a decrease in microseizures detected (20.6%, 46.1%, 70.6%, 82.4%, 83.3%) with increasing contact pitch (1.08, 1.52, 2.15, 3.05, 4.31 mm, respectively).

Conclusions: Intraoperative recordings with a high-density µ-ECoG LCP array show that microseizures occur at a greater frequency in epilepsy patients compared to controls. Micro contact size and greater contact density enables improved detection of microseizures.

Funding: Please list any funding that was received in support of this abstract.: This work is supported by a CTSA grant (UL1TR002553), NIH U01 NS099697-01, NIH T32 GM136573-01, and Finding a Cure for Epilepsy and Seizures (FACES).