NI13
Analysis of White Matter Tracts of Resting State Cognitive Networks in Multiple Sclerosis

Thursday, June 2, 2016
Exhibit Hall
Rahil M Dharia, BSc , School of Medicine, Wayne State University, Detroit, MI
Esmaeil Davoodi-Bojd, PhD , Neurology, Henry Ford Hospital, Detroit, MI
Quan Jing, PhD , Neurology, Henry Ford Hospital, Detroit, MI
Lian Li, PhD , Neurology, Henry Ford Hospital, Detroit, MI
Mirela Cerghet, MD, PhD , Neurology, Henry Ford Hospital, Detroit, MI
PDF


Background: Cognitive impairment affects 40-70% of patients with MS and these impairments are seen in multiple domains of cognition including attention, memory, speed of information processing. In a previous study using resting-state fMRI we found abnormal connectivity links for most patients with MS in default mode network (DMN) as well as other cognitive networks (attention, memory, language, visual-spatial memory). 

Objectives: To analyze the white matter damage and topological organization of white matter tracts in specific brain regions responsible for cognition in Multiple Sclerosis (MS) patients and compared with healthy controls. 

Methods: MRI measurements of DTI, T1, T2, and T2 FLAIR were acquired by GE clinical 3T system of 32 healthy subjects and 20 MS patients. DTI images were preprocessed in MATLAB to calculate the fractional anisotropy (FA) maps. The FA maps then were analyzed by TBSS toolbox in to co-register and create the skeleton map of the white matter (WM) for each case. For each patient, a 5×5×5 kernel was centered on each skeleton point and a one tail t2-test was used to calculate the p-value of the hypothesis testing that the FA values inside that kernel are smaller than the corresponding points in all normal cases. HAMMER package was used to segment each T1-weighted image into 94 regions and then aligned to the DTI space. In each region of each patient’s brain segment, the p-values were calculated and the region was considered as significant if the averaged p-value was smaller than 0.05. The percentile of the MS patients having significant FA reduction in each region was calculated. In the next phase, T1-weighted images of the whole brain were segmented in 164 regions using FreeSurfer and fiber tracking between connectivity nodes was assessed. 

Results: Out of 94 regions, in 46 regions none of the 20 patients showed significant FA reduction while in 17 regions (anterior limb of internal capsule, frontal lobe WM, globus pallidus, hippocampal formation, lateral occipito-temporal gyrus, parietal lobe WM, temporal lobe WM, uncus, corpus callosum), 10 or more patients showed significant FA reduction.  On second year follow up MRI 4/8 (50%) showed significant FA further reduction in frontal lobe WM, globus pallidus, hippocampal formation, right lateral occipito-temporal gyrus, occipital lobe WM and uncus. DTI connectivity in DMN and cognitive networks was decreased, while fMRI connectivity showed mixed results.

Conclusions: MS patient have significant white matter damage and reduction in various regions in the brain involved in cognitive networks. We report that both anatomical and functional connectivity inside default mode network and cognitive networks showed significant differences compared with healthy subjects. For MS patients with multiple scans over time, further progression of white matter damage was found.