Abstracts

Rationale: Several techniques, most notably electrical cortical stimulation mapping (CSM), can be used to delineate functional areas prior to resective brain surgery. However, current techniques that attempt to establish the connectivity between these functional areas (i.e., diffusion tensor imaging (DTI), resting-state fMRI, or cortico-cortical evoked potentials (CCEPs)) are qualitative, difficult to apply, and/or not reliable. In this study, we developed an automated and quantitative method called SIGNI (Stimulation-Induced Gamma-based Network Identification) that definitively and automatically establishes connectivity by evaluating cortical population-level responses to electrical cortical stimulation. This connectivity information could supplement identification of functional areas and may be useful for reducing postoperative functional deficits. Methods: We recruited 8 human subjects with intractable epilepsy that underwent surgical implantation of subdural electrodes (1-3 mm exposed diameter, 3-10 mm pitch) over temporal, frontal, and parietal cortices. We delivered one train of approximately 200 electrical pulses (2 Hz, 4-15 mA, 0.3 ms pulse width) to each of 63 electrode pairs while recording electrocorticographic (ECoG) signals at all other locations with high sampling frequency. Our SIGNI technique automatically removes electrical artifacts with a novel procedure, determines those locations at which broadband gamma activity (70-170 Hz, an established indicator of cortical population-level activity) significantly increases with stimulation, and visualizes the resulting connectivity on a three-dimensional model of the brain. Results: We verified the efficacy of our method by investigating the effect of stimulation amplitude on evoked cortical activity (r²=0.4, p<0.05) and how distance from the stimulated location affects the timing of the evoked activity (r²=0.4, p<0.05). Conclusions: Our results demonstrate that cortical connectivity can be rapidly established in an intraoperative or extraoperative setting during functional mapping with subdural electrodes. The additional information our method provides may guide surgical planning, particularly in cases where pathologic tissue neighbors eloquent cortex, and may improve postoperative outcomes and reduce functional loss. SIGNI may also prove useful for basic science investigations of the basis of cortical function and communication. Funding: This work was supported by the NIH (EB018783 and MH109429), the U.S. Army Research Office (W911NF-14-1-0440) and Fondazione Neurone.