Hi Alex, Yes. This is a very important step! DIPY provides a function that takes in a gradient table and a sequence of affines and produces a corrected gradient table: https://github.com/dipy/dipy/blob/master/dipy/core/gradients.py#L605 Cheers, Ariel On Mon, Dec 13, 2021 at 6:23 AM Alex Crimi via DIPY <dipy@python.org> wrote:

Hi there,

not sure this is specific to Dipy but maybe it is. As someone pointed out to me, if we register DWI to something else before running the tractography the values in the bvec files should also be rotated. There is even a paper explaining that this is critical: The B-matrix must be rotated when correcting for subject motion in DTI data - PubMed (nih.gov) <https://pubmed.ncbi.nlm.nih.gov/19319973/>

What is your experience with this? Do you know any easy transformation for this? We need to do this as to carry out the diffusion harmonization (see my previous mail) we need to register all volumes to each other.

Best, Alex _______________________________________________ DIPY mailing list -- dipy@python.org To unsubscribe send an email to dipy-leave@python.org https://mail.python.org/mailman3/lists/dipy.python.org/ Member address: arokem@gmail.com

Hi Alex, Yes. This is a very important step! DIPY provides a function that takes in a gradient table and a sequence of affines and produces a corrected gradient table: https://github.com/dipy/dipy/blob/master/dipy/core/gradients.py#L605 Cheers, Ariel On Mon, Dec 13, 2021 at 6:23 AM Alex Crimi via DIPY <dipy@python.org> wrote:

Hi there,

not sure this is specific to Dipy but maybe it is. As someone pointed out to me, if we register DWI to something else before running the tractography the values in the bvec files should also be rotated. There is even a paper explaining that this is critical: The B-matrix must be rotated when correcting for subject motion in DTI data - PubMed (nih.gov) <https://pubmed.ncbi.nlm.nih.gov/19319973/>

What is your experience with this? Do you know any easy transformation for this? We need to do this as to carry out the diffusion harmonization (see my previous mail) we need to register all volumes to each other.

Best, Alex _______________________________________________ DIPY mailing list -- dipy@python.org To unsubscribe send an email to dipy-leave@python.org https://mail.python.org/mailman3/lists/dipy.python.org/ Member address: arokem@gmail.com