Hi MRtrix team,
I’m using the -seed_dynamic option in tckgen to generate whole-brain tractograms and had a few questions about best practices when not using ACT. In our study, we are specifically including small brainstem substructures as part of the connectome. These regions are often not well handled by standard 5TT segmentations, which is why we believe that ACT might not feasible for our current pipeline. When reading about dynamic seeding, I’ve come across some ambiguity in the documentation and in the literature:
-
The MRtrix documentation says that
-seed_dynamic fod.mif“must not provide any other seeding mechanism,” but it doesn’t mention that a mask is required. Published studies that have incorporated dynamic seeding (e.g., Smith et al., 2015, Yeh et al., 2019), have not specified if a mask has been provided when not applying ACT/5tt anatomical segmentations. -
My main question: Is it recommended (or functionally necessary) to include a white matter mask with
-seed_dynamiceven when not using ACT? Will SIFT2 adequately compensate for the lack of anatomical constraints in the seeding step?
Any clarification on what happens behind the scenes when a mask is not provided would be greatly appreciated!
Thanks so much for your help and for all the fantastic tools and documentation!
— Monica