Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Differences in White Matter Fiber Density and Fiber-bundle Cross-section in Children who Stutter compared to Controls: a Fixel-based Analysis
Poster Session D, Saturday, October 26, 10:30 am - 12:00 pm, Great Hall 3 and 4
Hasini Weerathunge1, Fiona Höbler1, Mike Angstadt1, Andrew Yahn2, Ho Ming Chow3, Soo-Eun Chang1; 1Department of Psychiatry, University of Michigan, Ann Arbor, Michigan, United States of America, 2Functional MRI Laboratory, University of Michigan, Ann Arbor, Michigan, United States of America, 3Department of Communication Sciences & Disorders, University of Delaware, United States of America
Previous diffusion weighted MRI (DWI) studies in developmental stuttering have reported reduced fractional anisotropy (FA) in speech-related fiber pathways (Chow and Chang 2017; Chang et al., 2015). FA is a surrogate measure of white matter integrity and the most commonly reported measure of white matter microstructure. However, this voxel-averaged quantitative measure is not fiber-specific, and have poor interpretability, especially in brain regions with high probability of multiple fibers crossing at different orientations. These crossing fibers lead to an artificially lower FA value. With regards to children who stutter (CWS), inferior frontal gyrus (IFG) is one such location where FA has been found to be significantly different from age- and sex-matched children who do not stutter (CWNS). Fixel-based analysis is an advanced analysis method that can inform within-voxel microscopic fiber density and macroscopic morphology in a combined manner, while resolving crossing fiber-based variations in diffusion tensor imaging (DTI)-based measurements. In this study, we applied a fixel-based framework to investigate measures of fiber density (FD), fiber-bundle cross-section (FC), and fiber density and cross-section (FDC). We conducted a whole-brain tractography analysis to investigate the white matter structural connectivity differences in CWS compared to age- and sex-matched CWNS. We expected the group differences in fixel-based measures to be consistent with previous findings in white matter morphology studies to provide confirmation and further clarity on potential biological bases of white matter microstructural differences in CWS compared to CWNS. A total of 96 high quality DTI scans from 45 CWS and 51 CWNS between 3 and 10 years of age were retrospectively analyzed for this cross-sectional study. DWI data (Refer to Chow and Chang 2017 for sequence parameters) were processed using MRtrix3 (version RC3; Tournier et al., 2019) according to the fixel-based analysis pipeline provided via MRtrix3 documentation. For fixel-based analysis, tissue-specific response functions for white matter, grey matter, and cerebrospinal fluid were generated using a single-shell multi-tissue response function estimation method (Dhollander et al., 2016, 2018). Multi-tissue constrained spherical deconvolution algorithm (CSD; Tournier et al., 2004, 2007) was used to estimate the orientation of fiber(s) in each voxel. Fixel-based measures were analyzed voxel-wise to examine group differences using a General Linear Model (GLM). The connectivity-based fixel enhancement (CFE) approach was used to generate connectivity-based smoothing and statistical inference at the whole-brain level (Raffelt et al., 2015). Non-parametric permutation tests (5000 iterations) were used to identify differences in the fixel-based measures between CWS and CWNS, and family-wise error (FWE)-correction was applied to generate corrected p values (pFWE < .005). Preliminary results showed that CWS exhibited increased FC relative to CWNS in brain foci previously identified to show FA differences in developmental stuttering, including the bilateral IFG, bilateral motor, right prefrontal, medial premotor and superior cerebellar regions. The FD measure and the combined fiber density cross-section measure however did not show statistically significant differences between the groups. The fixel-based measures provide initial evidence that the nature of the white matte morphology differences observed in stuttering may be related to fiber bundle cross-section variations rather than fiber density.
Topic Areas: Disorders: Developmental, Methods