Computing DTI for motion corrected dMRI data#

Here, we compute the diffusion tensor metric images. SCT relies on the excellent dipy library for computing dMRI metrics [Garyfallidis et al., Front Neuroinform 2014].

sct_dmri_compute_dti -i dmri_moco.nii.gz -bval bvals.txt -bvec bvecs.txt
Input arguments:
  • -i : The input dMRI image.

  • -bval : A text file containing a b-value for each volume in the dMRI image, indicating the diffusion weightings for each of the volumes in the dMRI image.

  • -bvec : A text file with three lines, each containing a value for each volume in the input image. Together, the the three sets of values represent the (x, y, z) coordinates of the b-vectors, which indicate the direction of the diffusion encoding for each volume of the dMRI image.


You can also supply the -method restore option to estimate the tensors using “RESTORE: robust estimation of tensors by outlier rejection” [Chang, Magn Reson Med 2005].

Output files/folders:
  • dti_FA.nii.gz : Fractional anisotropy (FA) diffusion tensor image.

  • dti_AD.nii.gz : Axial diffusivity (AD) diffusion tensor image.

  • dti_MD.nii.gz : Mean diffusivity (MD) diffusion tensor image.

  • dti_RD.nii.gz : Radial diffusivity (RD) diffusion tensor image.

Now that we have the diffusion tensor images, we can move on to extracting DTI metrics from specific regions of the image.