Computational protocol: White matter microstructure pathology in classic galactosemia revealed by neurite orientation dispersion and density imaging

Similar protocols

Protocol publication

[…] Pre-processing of the data included estimating susceptibility induced distortions. From the pairs of images with reversed phase-encode directions (and thus distortions going in opposite directions), the susceptibility-induced off-resonance field was estimated (similar to Andersson et al ; topup of FMRIB Software Library [FSL] version 5.0, Smith et al ). Further, eddy current-induced distortions and subject motion were estimated, and all distortions were corrected (using FSL's eddy). B-vectors were rotated to account for the corrections (using Python; one shell of the corrected DWI data (b = 1000 s/mm2), diffusion tensors were estimated using a linear fitting algorithm (dtifit, FSL). Diffusion Tensor Imaging ToolKit (DTI-TK) was used for tensor-based spatial normalization of the volumes to an iteratively optimized population-specific template (Zhang et al ) ( The algorithm applies a deformable registration to the tensor images, resulting in improved registration compared to standard FA-based registration algorithms (Wang et al ; Keihaninejad et al ). The resulting normalized maps were averaged and high-resolution FA maps (1 mm iso-voxel) were derived. The mean FA map was thinned to create a mean FA skeleton, representing the centres of all WM tracts common to all participants (tract based spatial statistics [TBSS], FSL, Smith et al ). After skeleton evaluation, each participant's aligned data was then projected onto this skeleton using calculated distance maps, and the resulting data were fed into the statistical analysis.In parallel, NODDI was applied to the pre-processed data ( The NODDI tissue model distinguishes between three compartments: 1) intra-neurite space, representing the neurites and modelled as restricted diffusion (sticks, incorporating orientation dispersion); 2) extra-neurite space, surrounding the neurites, modelled as hindered, but not restricted diffusion (anisotropic Gaussian diffusion); and 3) cerebral spinal fluid (CSF), modelled as isotropic Gaussian diffusion. The main resulting parameters are: neurite density index (NDI), derived from the intra-neurite volume fraction (high in WM, low in GM); and orientation dispersion index (ODI), quantifying variation of neurite orientation (ranging from 0 for perfectly coherently oriented structures, to 1 for isotropic structures; this index is typically high in GM, low in WM). The output scalar maps from NODDI were normalized to the —already defined— study-specific common group space using the transformation fields as calculated per participant during the tensor-based registration. Then, the normalized NDI and ODI data were projected onto the —already calculated— mean FA skeleton using the original distance maps.On the skeletonised FA, NDI and ODI maps, permutation-based statistics were carried out (as implemented by randomise in FSL; 5000 permutations). First, a design was used having group as a between-subjects factor and age as a covariate. Second, correlations were calculated with age, several disease variables (i.e. age at onset of the diet, GALT enzyme activity) and available behavioural outcomes (visual and verbal working memory, sustained attention, voice onset times in a language production task) were examined across the skeleton (age only) and within regions of interest (ROIs). In the skeleton approach, p-values were corrected by means of the threshold-free cluster enhancement (TFCE) option (Smith and Nichols ). In the ROI approach, no correction for multiple testing was applied (due to the small sample size), but the effect size of the correlations was taken into account (only large correlations [r >0.5] were considered as relevant; Cohen ). An alpha of 0.05 (corrected for multiple comparisons where applicable) was used as the significance level. […]

Pipeline specifications

Software tools FSL, DTI-TK
Application Diffusion magnetic resonance imaging analysis
Organisms Homo sapiens
Diseases Galactosemias, Metabolism, Inborn Errors, Leukoencephalopathies