How create a connectome between voxels?

Hi all,

I am new and learning MRtrix from Introduction to MRtrix — Andy’s Brain Book 1.0 documentation (andysbrainbook.readthedocs.io).

Finally, I created a connectome that represents the number of streamlines connecting different parts of the brain.

Similarly, can I create a connectome that represents the number of streamlines connecting different voxels of the whole brain? How can I get it specificly?

Thanks,

Ricardo

Hi, you might find this paper helpful: High-resolution connectomic fingerprints: Mapping neural identity and behavior - ScienceDirect. And the code: GitHub - sina-mansour/neural-identity.

Best,
Steven

Sina’s work linked above specifically looks at vertex connectivity, which we still do not have native support for.

If you’re specifically interested in voxel connectivity, the requisite question is: do you only want endpoint-endpoint connectivity between voxels (acknowledging that surface vertices are a superior manifold for such), or do you want intersection-based connectivity between all voxels, primarily those containing white matter? The latter is IMO closer in concept to the fixel-fixel connectivity matrix used in the FBA pipeline rather than “connectomics”.

Theoretically you could do either by generating a “parcellation” image where each voxel contains a unique integer value, and generating a connectome from that (and indeed, utilising the -assignment_all_voxels option would get you the latter of the cases described above). However I expect this to immediately run out of RAM on any system you could attempt it on, as we do not yet support sparse connectivity matrix generation.

Cheers
Rob

Thanks for your answer.