|
I'd recommend using wb_command -surface-create–sphere to make a 10k sphere with as close to equal tile/triangle areas as one can get (though updated versions of FreeSurfer shouldn't have an issue with the GIFTI files). It doesn't sound like you need to
do any further registration—you just want to downsample the data.
You'll need to create your own CIFTI space with 10k surfaces instead of 32k surfaces. You'll need to decide if you want to keep the 2mm subcortical voxels (there are about 30k of them) or downsample those as well. You can extract the subcortical (and
cortical) data using wb_command -cifti-separate. To create the CIFTI space, you'll also need to use wb_command -cifti-separate to output the medial wall ROIs (-roi). You can use wb_commmand -metric-resample to create 10k versions of these using the 32k sphere
and your new 10k sphere. Then you can create a new CIFTI dense timeseries like this:
wb_command -cifti-create-dense-timeseries <ciftiout> -volume <downsampled label volume> <downsampled data volume> -left-metric <downsampled left surface data> -roi-left <downsampled left medial wall roi> -right-metric <downsampled right surface data> -roi-right
<downsampled right medial wall roi>
This will have created the new CIFTI space.
You can then use wb_command -cifti–resample with this new CIFTI space as the template CIFTI space. The current surface for left and right hemispheres is the 32k sphere we distribute with each subject in ${StudyFolder}/${SubjectID}/MNINonLinear/fsaverage_LR32k/${SubjectID}.${Hemisphere}.sphere.32k_fs_LR.surf.gii.
The new sphere is your 10k sphere from wb_command -surface-create-sphere. We don't yet have individual cerebellar surfaces, so we use voxels for the cerebellum, knowing this is not ideal.
You likely know that the 32k surface meshes were chosen because their average vertex spacing (2mm) matches the acquired fMRI data, so a 10k surface will be coarser than the acquired fMRI data.
Let me know if you have any further questions.
Peace,
Matt.
From: Chris Baldassano <[email protected]>
Date: Tuesday, October 8, 2013 7:28 PM To: "[email protected]" <[email protected]> Subject: [HCP-Users] wb_command -cifti-resample Hi all,
I'm trying to use -cifti-resample and I'm having a hard time understanding what the arguments are. After specifying the input and template cifti files, there are options for left-spheres, right-spheres, and cerebellum-spheres. How do I use these options?
The help text says to provide "a sphere with the same mesh as the current left/right surface" and "a sphere with the new left/right mesh that is in register with the current sphere" - what does this mean, and how are these spheres related to the surfaces in
the input and template files? Do these options override the surface information in the template? What is cerebellum-spheres, since I thought that subcortical areas are stored as volumes rather than surfaces?
My end goal is to downsample the input surface to ~10k nodes/hemisphere (for input to a very computationally-intensive model), so following the advice from this previous thread http://www.mail-archive.com/[email protected]/msg00039.html
I used freesurfer to decimate the left and right hemisphere 32k spheres. FYI this required manually editing the GIFTI files to have version "1.0" instead of "1", otherwise freesurfer chokes on reading them. Now I'm trying to figure out the next step - should
I be creating a new CIFTI template using these decimated spheres, or can I use the left/right spheres options?
Thanks!
Chris Baldassano
_______________________________________________
The materials in this message are private and may contain Protected Healthcare Information or other information of a sensitive nature. If you are not the intended recipient, be advised that any unauthorized use, disclosure, copying or the taking of any action in reliance on the contents of this information is strictly prohibited. If you have received this email in error, please immediately notify the sender via telephone or return mail. _______________________________________________ |
- [HCP-Users] wb_command -cifti-resample Chris Baldassano
- Re: [HCP-Users] wb_command -cifti-resample Glasser, Matthew
- Re: [HCP-Users] wb_command -cifti-resample Timothy Coalson
