I am working on the mice’s brain tractography. And I perform the CSD and tractrography according to the guidelines. But it seems weird about the results because the fiber length are much shorter than what I expected. And it more looks like dot-shaped, not the lines. Does anymore know the reason why it looks like that? Or is it normal for the mice?
Many thanks in advance!
The first thing to verify is whether your diffusion gradient orientations are correctly imported. You could do a quick dwi2tensor and then extract the tensor first eigenvector field using tensor2metric and display it in mrview using the vector plot tool. Do the tensor eigenvectors make anatomical sense?
The second thing to verify is whether the single fiber response function was obtained correctly. Is this using the response function that your colleague sent me by email earlier today? How did you obtain the response function? Regardless of how you obtained it, it is always a good idea to overlay the set of voxels that was used to estimate the response on your data set and see if the locations make sense (do they correspond to locations where you expect single fiber voxels?).
The third thing to verify is whether the fiber ODFs make sense. Do they point in the direction of the fibers? Do you see sharp single fiber populations in the voxels used for response function estimation? I don’t really see any obvious long fiber bundles popping up from your ODF field. This might point to using a suboptimal response function and/or interpreting the gradient directions incorrectly. Also, you provided no details about how you performed the deconvolution step?
Finally, it is important to know what tracking parameters you used in order to help you.
Dear Dr. bjeurissen,
Many thanks for your reply. I have solved this problem by flipping z direction. And I have an another question about connectome analysis. Does it will affect the results of connectome if the atlas is not in the same coordinates with the result of tractography?
BTW, I want to know how can I make the atlas and 4D-nifti file in the same coordinate space? Thanks!