Related Experiment Video
Updated: Sep 11, 2026
Diffusion Tensor Magnetic Resonance Imaging in the Analysis of Neurodegenerative Diseases
Published on: July 28, 2013
Microstructural Characterization of Functional Network-Based Neurodegeneration in Multiple Sclerosis Using
Florence L Chiang1, Eva A Krijnen2, Diana A Hobbs2
1From the Department of Radiology (F.L.C., D.A.H.), University of Texas Southwestern Medical Center, Dallas, TX, USA; Department of Radiology (F.L.C., H.L., H.-H.L., A.B., S.Y.H.), Neurology (F.A.K., E.C.K.), Massachusetts General Hospital, Boston, MA, USA and Research Imaging Institute (P.T.F.), University of Texas Health Science Center at San Antonio, San Antonio, TX, USA. Florence.Chiang@UTSouthwestern.edu.
Background And Purpose:
Neurodegeneration is a key component of clinical disability in multiple sclerosis (MS). However, the underlying mechanism of localized gray matter (GM) atrophy in MS remains unknown. More recently, a network-based etiology has been postulated, which may be associated with clinical progression. The goal of this study was to determine whether GM microstructural abnormalities are organized across an atrophy-prone network.
Materials And Methods:
We leveraged high-gradient diffusion MRI (dMRI) to probe GM mesoscopically by using the SANDI (Soma and Neurite Density Imaging) biophysical modeling approach. The intra-soma signal fraction (fis) was computed, which is a putative biomarker of GM cytoarchitecture. Regions of interest (ROIs) defined by nodes in the Atrophy-based Functional Network (AFN) were used to sample the individual fis map. Group-wise comparisons of the nodal and aggregate fis were assessed using Mann-Whitney U tests, and the multivariate fis (principal component 1; PC1) using independent samples t-tests, with false discovery rate (FDR) correction. Association of PC1 with the Expanded Disability Status Scale (EDSS) score was assessed using partial Spearman's rank-order correlation, controlling for age and sex. The same approach was applied to examine relationships across all nodal pairs.
Results:
Participants included 38 MS (M/F: 11/27; age: 44 ± 11 years; EDSS: median 2.25, range: 1 - 7.5; disease duration: median 8.5 [IQR: 4.0, 14.0] years) and 35 age- and sex-matched healthy controls (HC; M/F: 15/20, p = 0.32; age: 39 ± 15 years; p = 0.13). fis was decreased in MS for the aggregate average and PC1 of all AFN nodes. Correlation of the EDSS and PC1 further showed that fis decreases as disease severity worsens (ρ = -0.44, p < 0.05). FDR-corrected covariance analysis exhibited medium-to-large effect sizes, with surviving correlations having ρ ≥ 0.39.
Conclusion:
Decreased fis in atrophy-prone GM of MS correlated with disease severity. Further, GM microstructural covariance suggests neuronal loss may relate in part to network effects. Network-based microstructural measures may therefore inform future development of quantitative methods for monitoring disease progression in MS.
