Assessing Microstructural Substrates of White Matter Abnormalities: A Comparative Study Using DTI and NODDI

Abstract

Neurite orientation dispersion and density imaging (NODDI) enables more specific characterization of tissue microstructure by estimating neurite density (NDI) and orientation dispersion (ODI), two key contributors to fractional anisotropy (FA). The present work compared NODDI- with diffusion tensor imaging (DTI)-derived indices for investigating white matter abnormalities in a clinical sample. We assessed the added value of NODDI parameters over FA, by contrasting group differences identified by both models. Diffusion-weighted images with multiple shells were acquired in a group of 8 healthy controls and 8 patients with an inherited metabolic disease. Both standard DTI and NODDI analyses were performed. Tract based spatial statistics (TBSS) was used for group inferences, after which overlap and unique contributions across different parameters were evaluated. Results showed that group differences in NDI and ODI were complementary, and together could explain much of the FA results. Further, compared to FA analysis, NDI and ODI gave a pattern of results that was more regionally specific and were able to capture additional discriminative voxels that FA failed to identify. Finally, ODI from single-shell NODDI analysis, but not NDI, was found to reproduce the group differences from the multi-shell analysis. To conclude, by using a clinically feasible acquisition and analysis protocol, we demonstrated that NODDI is of added value to standard DTI, by revealing specific microstructural substrates to white matter changes detected with FA. As the (simpler) DTI model was more sensitive in identifying group differences, NODDI is recommended to be used complementary to DTI, thereby adding greater specificity regarding microstructural underpinnings of the differences. The finding that ODI abnormalities can be identified reliably using single-shell data may allow the retrospective analysis of standard DTI with NODDI.

Dissociable diffusion MRI patterns of white matter microstructure and connectivity in Alzheimer’s disease spectrum

“…Whereas these measures have been suggested as sensitive features of WM microstructure, changes in their values may not be attributed to changes in specific tissue microstructure substrates23. NODDI and RSI are complementary approaches based on multi-compartment models of diffusion2343, and may provide microstructural indices with higher specificity compared to conventional diffusion MRI metrics2344. However, the application of such multi-compartment models on the single-shell data included in this study, although technically feasible45, is suboptimal23, and future work should strongly consider additionally including complementary diffusion measures derived using multiple-shell diffusion MRI data.…”

Age-related microstructural and physiological changes in normal brain measured by MRI γ-metrics derived from anomalous diffusion signal representation

“…A registration pipeline similar to that proposed by Timmers et. al (Timmers et al, 2016) was used. Briefly, a population-specific template was obtained with DTI-TK software (available on http://www.nitrc.org/projects/dtitk).…”

“…The threshold limit value of this skeleton was set to 0.4 in order reduce bias due to cross subject variability of the WM tracts. Finally, the participant-specific transformation fields, obtained during the tensor-based transformation, were used to normalize all the other diffusion metrics used in this study as specified by Timmers et al in supplementary methods (Timmers et al, 2016).…”

Characterizing White Matter in Huntington's Disease

“…The higher shell value of b = 2000 was excluded as the diffusion signal at this b-value is no longer suitable for modeling diffusion tensors. 35 The tensor was fit to the combined shell data using fits a diffusion tensor model at each voxel (DTIFIT) and the output FA, MD, RD, and AD maps were included in the TBSS analysis. For the analysis of NODDI metrics, diffusion data from both acquisitions were combined using all shells.…”