Figure 2.

Regions of interest definition in diffusion space and DTI Probabilistic tractography of left CST fiber. The ROIs from MNI space were transformed into native diffusion space. Initially, the DTI images were rigidly registered to the individual T1-weighted image using FSL’s Linear Image Registration Tool (FLIRT), employing mutual information as the cost function to generate the FA_2T1 matrix. The individual T1 image was then normalized to MNI space through both linear (FLIRT) and nonlinear registration (FNIRT, FSL’s Nonlinear Image Registration Tool). Subsequently, the co-registered DTI image in structural space was warped using the transformation field obtained from the T1 to MNI normalization. The transformation matrix (FA_2T1) and warp-fields (T1_2MNI warp) were inverted using inverse and invwarp command, respectively. These inverse transformations were applied to the ROIs in MNI space to obtain the corresponding ROIs in native diffusion space. We used the BEDPOSTX (Bayesian Estimation of Diffusion Parameters Obtained using Sampling Techniques) to calculate the distribution of fiber orientations at each brain voxel. To ensure that only the white matter fibers were tracked, tractography for each hemisphere was conducted twice, once with the left CP as the seed mask and the left PLIC as the termination mask, and vice versa (i.e., the left PLIC = seed mask; the left CP = termination mask, the left CST tract). For each hemisphere and species, we applied a threshold to include only those voxels that received at least 3.08 × 10^-5 percent of the total streamlines emitted from the ROI masks used to trace each tract (a total of 25,000 times the number of voxels in the mask). Furthermore, tracts in diffusion space were binarized and combined by overlapping operation (overlapped-tract) for each subject. Finally, we masked diffusion maps with the binarized overlapped-tracts to compute the mean value of FA within the final-tract mask in diffusion space.