Identifying proximity to white matter language tracts with gradient-based intraoperative electrical mapping.

Abstract

Intraoperative identification of language white matter tracts (WMTs) is challenging, as these tracts are visually imperceptible. This study aimed to assess whether proximity to the language WMTs can be determined intraoperatively by correlating direct electrical stimulation (DES) intensity with the distance to language tracts as defined by preoperative diffusion tensor imaging (DTI)-based tractography. Twenty-eight patients undergoing awake craniotomy for diffuse glioma resection participated in the study. All patients received preoperative language assessments and DTI-based language tract reconstruction. Subcortical DES was applied along the tumor cavity border using bipolar or monopolar stimulation, with DES locations registered for offline analysis. A positive linear correlation was found between the distance from the stimulated point to the closest language WMT and the subcortical DES electrical threshold (r = 0.57). Stimulation that evoked interference had a significantly lower intensity (mean 6.93, SD 3.82; n = 21) than noninterfering cases [mean 15.06, SD 7.4; n = 11; t(30) = 3.2, p < 0.001]. Tumor pathology, volume, and associated edema did not significantly affect the distance-intensity correlation or likelihood of language interference. Only the bipolar stimulation correlation remained significant following separate analysis of the bipolar and monopolar methods. These findings suggest that intraoperative threshold-based electrical mapping can feasibly assess language tract proximity, supporting maximal tumor resection while minimizing language deficits.