Related Experiment Video
Updated: Jun 17, 2026

Translational Brain Mapping at the University of Rochester Medical Center: Preserving the Mind Through Personalized Brain Mapping
Published on: August 12, 2019
Cortical stimulation for language mapping in focal epilepsy: correlations with tractography of the arcuate fasciculus
Beate Diehl1, Zhe Piao, Jean Tkach
1Department of Clinical Neurophysiology, The National Hospital for Neurology and Neurosurgery, London, United Kingdom. beate.diehl@uclh.nhs.uk
Purpose:
Diffusion tensor imaging (DTI) provides information about magnitude (diffusivity) and directionality (fractional anisotropy, FA) of water diffusion and allows visualization of major white matter tracts. The arcuate fasciculus (AF) connects anterior (Broca's) and posterior (Wernicke's) language areas. We hypothesized that essential language areas identified by direct cortical stimulation would colocalize with areas revealing subcortical connectivity via the AF.
Methods:
Fourteen patients with pharmacoresistant left hemispheric epilepsy and left hemisphere language dominance underwent invasive evaluations for localization of epileptogenicity and functional mapping. DTI and T1-weighted volumetric magnetic resonance imaging (MRI) scans were coregistered, and subdural grid electrodes identified on postimplantation computed tomography (CT) scans were also coregistered to the MRI scans. The AF was reconstructed from a region lateral to the corona radiata on the FA map. Colocalization, defined as <1 cm between the AF and the electrode positions delineating language cortex, was visually assessed with excellent reliability (Cronbach's alpha = 0.98).
Results:
A total of 71 subdural grid contacts were overlying language cortex. Nineteen contacts in eight patients were over Broca's area, 16 of which (84.2%) colocalized with the AF. Fifty-two contacts in 10 patients were over Wernicke's area, with colocalization in 29 patients (55.8%). Colocalization was significantly greater in anterior regions than in posterior regions [chi(2)(1) = 4.850, p < 0.05].
Discussion:
The AF, as visualized with DTI, colocalized well with anterior language areas, but less so with posterior language areas, inferring that the latter are more spatially dispersed.

