With some experimentation, I found that mri_label2vol tends to create worse volumes than recon-all's aparc2aseg regardless of the surface you use. I ended up getting good results by making a binarized mask from the segmented volume in the T1w folder, dilating it, and multiplying by a binarized WM mask to constrain the dilation to WM.
John On Fri, Feb 13, 2015 at 3:36 PM, John Plass <[email protected]> wrote: > Hi Matt, > Thanks for your help. Unfortunately, our analysis requires that we use > other software that only accepts volumetric masks. Basically, I'd just like > to mimic as closely as possible what was done in the pre-proccessing > pipeline to create the segmented .nii.gz volumes in each subject's T1w > folder, but selecting for a specific label of interest and projecting into > the white matter 1 mm. mri_label2vol assumes that the labels derived from > the annot files are defined for the ?h.white surface, but I wanted to see > if this is actually the case. Should I use, e.g., ?h.white.hires instead? > > John > > On Fri, Feb 13, 2015 at 2:31 PM, Glasser, Matthew <[email protected] > > wrote: > >> We recommend you run tractography directly from the surfaces using >> probtrackx2. You can use these surfaces if you want your other volume >> (i.e. non-cortical) ROIs in native subject space: >> >> >> ${StudyFolder}/${Subject}/T1w/fsaverage_LR32k/${Subject}.${Hemisphere}.white.32k_fs_LR.surf.gii >> >> Or these if you want them in MNI space: >> >> >> ${StudyFolder}/${Subject}/MNINonLinear/fsaverage_LR32k/${Subject}.${Hemisphere}.white.32k_fs_LR.surf.gii >> >> We recommend you stop tracking at the pial surface to avoid crossing >> sulci and that you use waystop masks in the subcortical greymatter to avoid >> connections going to sub cortex and then looping back around to cortex >> (which creates false positives). >> >> Peace, >> >> Matt. >> >> From: John Plass <[email protected]> >> Date: Friday, February 13, 2015 at 2:10 PM >> To: "[email protected]" <[email protected]> >> Subject: [HCP-Users] Projecting Label into White Matter >> >> Hello, >> I have a few questions about using HCP structural data to create >> waypoints/masks for tractography. In the past, we've created masks of the >> white matter adjacent to a label of interest with Freesurfer's >> mri_label2vol, using the "proj" argument to project from the label into the >> adjacent white matter. I'd like do this with the HCP data, but its a bit >> unclear to me what would be the proper files to use: >> >> 1. I believe this command uses the ?h.white surface by default. The >> extended preprocessed structural data contains multiple white matter >> surfaces (?h.white, ?h.white.hires, lh.white.deformed lh.white.prehires). I >> am not sure which to use. Which surface were the labels defined on? >> >> 2. What is the best .mgz file to use as a template volume? >> T1.hires.norm.mgz appears to be a good match for the T1 nii.gz file, but >> I'd like to be sure I'm using the best file. >> >> Any help would be greatly appreciated! >> >> John Plass >> >> _______________________________________________ >> HCP-Users mailing list >> [email protected] >> http://lists.humanconnectome.org/mailman/listinfo/hcp-users >> >> >> ------------------------------ >> >> 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 mailing list [email protected] http://lists.humanconnectome.org/mailman/listinfo/hcp-users
