Abstract WMP22: Reduced White Matter Integrity in Acute Stroke Patients With Language Deficits: a Tract-based Spatial Statistics Study

Abstract

Introduction: We examined white matter integrity in acute stroke Patients with Language Deficits. Hypothesis: Microstructural abnormality in specific white matter tracts would show significant reduction in stroke patients with language deficits. Methods: Stroke patients with language deficits (N = 10, mean age = 56.3 years, within 7 days from stroke onset, normed verbal fluency score of less than -1.5), and age- and gender-matched healthy controls without language deficits (N = 10, mean age = 55.6 years, normal verbal fluency score) were scanned on GE750 3T MRI scanners. 3D T1-weighted whole brain BRAVO and DTI data (56 directions, b value = 1000 s/mm2) were collected. Verbal fluency test was administered to all subjects. Fractional anisotropy (FA) maps were generated from a tensor-model fit of eddy current corrected DTI data using FSL software. The skeletonized and fully non-linearly registered FA data were obtained using Tract-Based Spatial Statistics (TBSS) implemented in FSL, and then they were used for voxel-wise statistical analysis across subjects. Corona radiata (CR) (anterior, superior, posterior), cingulum, superior longitudinal fasciculus (SLF), and uncinate fasciculus of both right and left hemisphere from Johns Hopkins University (JHU)-ICBM-DTI-81 white matter labels atlas were used to identify significantly affected white matter tracts revealed by TBSS. Student t-tests were used to compare group means for normally distributed data and Mann-Whitney U tests to compare group means for non-normal data. Statistical analysis was performed using StatPlus package. Significance was set at p <.05. Results: There was a significant reduction of FA in the left SLF, left superior CR, and both right and left posterior CR of stroke patients with language deficits compared to healthy controls. A slight trend toward significant reduction of FA was also seen in right superior CR (p = 0.089) of stroke patients with language deficits. No significant group differences in FA value were seen in uncinate fasciculus and cingulum. Conclusions: Stroke patients with language deficits showed significant reduction in FA values in functionally relevant white matter tracts such as the left SLF, left superior CR, and both right and left posterior CR.