As Matt said, our 91k grayordinate space represents the cerebellum with voxels, not surfaces, so the cerebellum surface options of various commands should not be used with these files. They are there in the hopes that future data will support routine generation of individual cerebellar surfaces.
There isn't currently an easy way to have the command line tell you what label clusters fall inside of (you could parcellate the cluster ROIs by the atlas to get a "participation" of labels in each cluster, but you'd need to write some extra code to find which ones are nonzero, and extract the relevant label names). You can use wb_view to overlay them, and click on the cluster to get the label at that voxel/vertex, and toggle the label layer on/off, etc. Tim On Thu, Feb 2, 2017 at 11:53 AM, Xavier Guell Paradis <xavie...@mit.edu> wrote: > Dear Tim, > Thank you very much for your message. In the last step > (-cifti-weighted-stats), I am not sure which file to use for the > -cerebellum-area-metric. I would like to obtain cerebellar cluster sizes in > mm^3. Would it be possible to use a cerebellar volumetric atlas such as > Cerebellum-MNIfnirt-maxprob-thr25.nii (this is an atlas available here > http://www.diedrichsenlab.org/imaging/propatlas.htm), so that the output > of -cifti-weighted-stats says in which cerebellar lobule(s) each cluster is > found? > > Thank you very much, > Xavier. > ------------------------------ > *From:* Timothy Coalson [tsc...@mst.edu] > *Sent:* Monday, January 30, 2017 6:18 PM > *To:* Xavier Guell Paradis > *Cc:* hcp-users@humanconnectome.org > *Subject:* Re: [HCP-Users] How to know cluster size (number of voxels) > after -cifti-find-clusters > > The -cifti-find-clusters command assigns a separate integer within each > cluster. You can use -cifti-label-import and then > -cifti-all-labels-to-rois to get each cluster in a separate map. Then, > -cifti-weighted-stats with -spatial-weights and -sum will give you mm^2 for > surface clusters and mm^3 for volume clusters. Unfortunately, it is not > easy to tell from the command line whether each cluster is on the surface > or in the volume. However, you could make a cifti file using the output of > -surface-wedge-volume and a volume file containing the voxel volume, and > use that in -cifti-weighted-stats with -cifti-weights and -sum to get both > surface and volume clusters in mm^3 (which assumes that surface clusters > are the full width of the ribbon). > > Tim > > > On Mon, Jan 30, 2017 at 4:26 PM, Xavier Guell Paradis <xavie...@mit.edu> > wrote: > >> Dear HCP experts, >> After using -cifti-find-clusters, is there a way to know the size of the >> clusters that the command has found? We know that the clusters will be >> larger than the specified volume-value-threshold, but is there a way to >> know the mm^3 or number of voxels of the clusters identified? >> >> Thank you very much, >> Xavier. >> >> _______________________________________________ >> HCP-Users mailing list >> HCP-Users@humanconnectome.org >> http://lists.humanconnectome.org/mailman/listinfo/hcp-users >> > > _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
