A few pointers:
1. Try using a fill threshold of 0 (this will take any label point that is in the label. setting the fill thresh to 0.5 means that a voxel needs to be filled half way to be considered for the ROI).
2. I would suggest using nii or mgz as an output format (we have not used bshort in a long time, where did you see it?)
3. The resulting label will only be a thin layer on the 2d sheet. To add "thickness" to it, you can use --proj. Run mri_label2vol with --help to get copious amounts of documentation.
4. The labelvoxvol won't make much difference if you use a 0 fill thresh
5. The resulting ROI should be binary unless you are using the --label-stat flag.
6. Why do you have --label specified twice?
doug
Arslan, Ali wrote:
Hi, I'm trying to create a ROI of V1 from the labels (for each hemisphere) I defined. I was able to create an volume with the command below
mri_label2vol --label $SUBJ/label/lh.v1.ali.label --label $SUBJ/label/rh.v1.ali.label --temp $SUBJ/mri/T1.mgz --reg $SUBJ/stats/register.dat --fillthresh 0.5 --o $SUBJ/ROI/test_v1_T_000.bshort
What I want are these
- This command gives me an ROI which includes the gray matter but I
want a smoother ROI which also has some nearby white matter voxels so that I make sure ROI doesn't leave any V1 voxel out for my functional analysis later on. I looked at the --proj flag for this but couldn't find any explanation about it. Would it help me about this? Where can I find more information about it? Would changing --labvoxvol make sense?
- The resulting ROI is not binary and I would like to make it so
(actually left and right hemispheres are in different densities, e.g. one seems lighter and the other one is darker in my viewer). Can I output the volume as binary from mri_label2vol or should I convert it later on (and how)?
Best, Ali _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer