Dear all,
I would like to perform the Voronoi random parcellation in my neonatal surfaces, I’ve been looking for how to do it but I didn’t find anything. Any clue about how I could do it? Thanks in advance.
Manuel
Hi @mblesac,
There is some code on @EmmaR’s GitHub for this: GitHub - ecr05/Random_forests_for_cortical_imaging_data: scripts for training random forest models models on HCP/dHCP
The voronoi parcellation on there goes up to 150 parcels per hemisphere. Does that work?
BTW, how is the surface registration going?
Best wishes,
Logan
Hi @lzjwilliams,
Thanks for your reply. I will have a look into this and see if I manage to run it.
Best regards,
Manuel
Hi @lzjwilliams,
I have difficulties with this code, how I should run it? This will produce a parcellation in native space of 150 rois comparable across subjects?
Best regards,
Manuel
I think @lzjwilliams meant that there is a 150 voronoi parcel label file in that repo. There is Matlab code to generate the files somewhere but technically its not mine to share - plus it would be better in Python but that’s not something I can deliver any time soon.
Hi @EmmaR,
Thanks for your reply. I got it now, so basically this is a 150 roi parcellation in template space.
Best regards,
Manuel
@EmmaR, the file for the Voronoi parcellation is a merged file, and the granularity goes up to 150.
@mblesac, I’ll try to get back to you ASAP
Hello @mblesac ,
I am also looking for a way to get random parcellation on the surface.
In my search I came across this github repo GitHub - miykael/parcellation_fragmenter: Fragments FreeSurfer parcellation annotation in N-equal sized parcels that might be helpful. It says it will subdivide pre-parcellated bigger regions, but it should work on the whole surface as well.
Having said that, I have not had time to test it yet.
Would be happy if we could share the progress and advise on this 
Best, Andrea