The labels are from Stanford VPNL's vcAtlas (cytorchitectonic atlas of the human ventral visual system).
Does this mean that my mri_label2vol command which I posted earlier looks okay?
To doublecheck my files in freeview, I did a simple
freeview -v $FSLDIR/data/standard/MNI152_T1_2mm.nii.gz \ $wd/MPM_rh.FG2.label.nii \ -l $VC/MPM_rh.FG2.label &
where *.label.nii is the output from the mri_label2vol command I ran. I couldn't figure out how you load the ROI NIFTI mask as an ROI so that you can see it on top of the structural template -- the -l switch seems to only take label formats?
In the attached screenshot freesurfer_prob_1.png, you see the yellow ROI corresponding to one label (fusiform gyrus) from the vcAtlas, which appears to be incorrectly rotated by 90deg.
The second attached screenshot freesurfer_prob_1.png shows the output from mri_label2vol as a white cluster, which if all were right, I would assume would overlap with the yellow cluster corresponding to the original atlas label...but I guess the volumetric ROI looks roughly in the right place, but still not quite right? Why does it look so different in mricron?
Sorry for flood of newbie Qs.... Thank you! Gina
------------------------------ FWD'd msg ------------------ Douglas N. Greve Wed, 28 Mar 2018 10:18:40 -0700
what is your freeview command line? In the mri_label2vol command below, where does the label come from?
On 2018-03-28 19:11, gj wrote:
Hi,
I loaded the 2mm MNI152 template and the mri_label2vol outputs in mricron, and I saw that the labels are completely outside the template. So, I know now to use freeview rather than the deprecated tkmedit so I can also load the original atlas labels to check against (using the 2mm MNI152 template as background) -- in any case, I don't even see the labels, I guess because the ROIs/masks lie outside the display. I'm also not sure what the equivalent of tkmedit's overlay-reg is for freeview? I assume my not loading one is why I get the error, "Label coordinate out of bound" when I try and load the original atlas labels (in fsaverage space) on the MNI152 template?
Thank you! Gina
P.S. Can't find a straightforward way to reply with the entire thread when receiving only a digest from the listserv -- reply to the digest takes the freesurfer-request address rather than the list address and the button on the website for archives only gives the option of replying directly to the person who posted)...what am I missing?
Douglas Greve dgreve at mgh.harvard.edu Tue Mar 27 14:01:16 EDT 2018
how are you verifying that it is not in the right place? You should view it on the 2mm MNI152 using freeview. Also, your label should be a label in fsaverage space.
On 3/27/18 9:52 AM, gj wrote:
Hi,
Newbie question....I want to take FreeSurfer labels (from an atlas in fsaverage space) and convert it into a volumetric NIFTI mask. I ran the following command, but I must be missing something because the output is not in the right space (far too posterior):
$FREESURFER_HOME/bin/mri_label2vol \--label $ll \ --temp $FSLDIR/data/standard/MNI152_T1_2mm.nii.gz \ --reg $FREESURFER_HOME/average/mni152.register.dat \ --fillthresh .5 \ --o $wd/${ll}.nii
Am I perhaps supposed to first create the volumetric file without mapping into MNI space and then use tkregister2?
Thanks in advance for any help or advice! Gina
