I am trying to generate DTI image dataset using a flourescence image dataset. I have been successful so far doing so. Now I need to generate multi-shell MRI-DTI image data. Can you please help me to know the format in which these kind of matrices are stored? So that I can store the multi-shell matrix in a proper form to use for tractography?
Assuming that you’re working with the MRtrix-format gradient table, the format is simply:
- One row for each volume in the image series
- 4 values along each row, separated by commas, that represent (in this order):
- X component of unit vector of diffusion sensitisation direction
- Y component of unit vector of diffusion sensitisation direction
- Z component of unit vector of diffusion sensitisation direction
b-value (in s.mm^-2)
Note that the directions are defined with respect to ‘scanner’ / ‘real’ space, and are not defined relative to the orientations of the image axes (as encoded in the transformation contained within the image header).
So if you are generating multiple pseudo-DWI ‘shells’ based on images with differential contrast, make sure that all of the pseudo-DWI volumes within each shell have the same b-value. Also, make sure you use b-values of greater than 10, as this is the threshold used to detect b=0 images, that are presumed to have no orientation dependence.