Poster E16, Saturday, August 18, 3:00 – 4:45 pm, Room 2000AB
Structural connectivity across stimulation-defined critical language areas
Brian H. Silverstein1, Eishi Asano1,2, Yasuo Nakai2,3, Jeong-won Jeong1,2;1Wayne State University, Detroit, MI, USA, 2Children’s Hospital of Michigan, Detroit, MI, USA, 3Wakayama Medical University, Wakayama, JPN
Data from electrical stimulation mapping (ESM) have identified areas critical to language function in the left temporal and frontal lobes, and studies of electrocorticography (ECoG) information flow suggest the areas are functionally connected. Yet, the anatomical pathways, especially between the posterior temporal lobe and both the inferior precentral gyrus (iPCG) and the inferior frontal gyrus (IFG), are unclear. We utilized diffusion tensor imaging (DTI) tractography to investigate the white matter tracts connecting ESM-defined language areas. We hypothesized that the posterior middle and superior temporal gyri (pMTG; pSTG) would project directly to the IFG and iPCG, with a bias towards the iPCG, supporting the ECoG findings. Three Tesla DTI scans were performed on 65 neurosurgical patients (33 males; age: 11.8±SD 3.7 y/o). Our group-level ESM probability map was generated via ESM-based determinations of cortical function in vivo from 100 neurosurgical patients with left hemispheric (LH) language dominance. From ESM and anatomy, we generated 8 LH ROIs in 3 domains: Receptive aphasia: pSTG and pMTG; Expressive aphasia: IFG, anterior iPCG (aiPCG), and two middle frontal gyrus (MFG) sites; Speech arrest: posterior iPCG (piPCG) and middle precentral gyrus (mPCG). All ROIs were used as seeds and targets for performing anatomically-constrained probabilistic tractography. The fiber counts between each pair of ROIs were then normalized by ROI volume. It was found that the pSTG and pMTG are directly connected with both the IFG and piPCG, with a bias towards the piPCG vs. the IFG (p<.0001). Fiber count did not correlate with age or gender (p>.05, r<.30). Notably, these connections followed the arcuate pathway, rather than the uncinate. Across all 8 ROIs examined, the pMTG and pSTG showed similar connectivity profiles; both had the strongest connections to the iPCG, followed by the IFG, and the pSTG/pMTG, with less connections to the mPCG and MFG. Our results demonstrate direct structural connectivity between the pMTG/pSTG and all other ESM-identified language areas. The fiber bundles directly connecting the pSTG with the IFG and piPCG can, in part, support the reciprocal neural interactions which facilitate processing of speech sounds reported in ECoG and fMRI studies of auditory speech perception and articulation. Likewise, the observation of dense interconnecting fibers between 8 ROIs in the frontal and temporal lobes greatly extends a language model beyond the Geschwind model.
Topic Area: Perception: Auditory