Abstracts

Rationale: Language mapping using event-related high-gamma (70-110 Hz) augmentation on electrocorticography (ECoG) has several advantages over the gold-standard electrical stimulation mapping via intracranial electrodes. The advantages of ECoG-based mapping include: (i) lack of stimulation-induced seizures, (ii) better sensitivity of localization of language areas in young children, and (iii) shorter patient participant time. Yet, ECoG-based language mapping may be currently underutilized among clinicians practicing epilepsy surgery. Here, we will publicly share our software analysis for language mapping with fellow investigators. Methods: We provide video and hands-on presentations to explain, point-by-point, our time-frequency analysis procedures. Results: We acquire ECoG data primarily using subdural platinum grid and strip electrodes. Video-ECoG is recorded on a 192-channel Nihon Kohden Neurofax 1100A Digital System with a sampling frequency at 1,000 Hertz. The average of the ECoG signals derived from the 5th and 6th intracranial electrodes is used as the original reference. ECoG signals are then re-montaged to a common average reference excluding channels classified as seizure onset zone and those affected by interictal spikes or artifacts. The onset of auditory or visual stimulus and patient's response are marked on ECoG recording offline for each trial. ECoG data are subsequently loaded in BESA EEG software, and reference time windows are specified for each naming task (e.g.: at 200-600 milliseconds prior to stimulus onset). Time-frequency analysis eventually determines at what channel and at what moment the amplitude of ECoG signals are augmented or attenuated in steps of 5 Hertz and 10 milliseconds. The aforementioned analysis procedures take up to 1 hour of investigator’s time based on user experience. Usage of batch function and automatic detection of stimulus onset can reduce the analysis time, but quality assurance by visual inspection is highly recommended. Bootstrap statistics are available to determine if amplitude modulation is statistically significant at each time-frequency bin and at each channel. Conclusions: The above-mentioned software analysis determines ‘when’ and ‘where’ high-gamma activity is augmented during a given task. We believe that this hands-on presentation will increase the transparency of our work and improve its familiarity to investigators considering implementation of ECoG-based language mapping in epilepsy presurgical evaluation. Funding: NIH Grant NS64033 (E. Asano)