This one did not overlay correctly either (see attached). This is the same mov volume/gray scale/extracted mask volume that I used as well. What does this tell us?
Best, Kareem
Quoting Douglas N Greve greve@nmr.mgh.harvard.edu:
OK, now check whether it overlays on the FSL template correctly in tkmedit:
tkmedit -f fslgrayscaleimage.nii.gz -overlay TP.nii.gz -fthresh 0.5
where fslgrayscaleimage.nii.gz is the gray scale image that you used as the mov volume in bbregister and also used as the gray scale image in fslview
doug
kwa1@rice.edu wrote:
Hmm, the label seems to be in the right area but larger and translated down (see attached picture). The bbregister I ran used the FS target to take the FSL standard space to and the registration looked fine. Any ideas why this happened?
Quoting Douglas N Greve greve@nmr.mgh.harvard.edu:
Sorry, I forgot that you were going to the surface, so dont use mri_label2vol. After #1, run tkmedit yoursubject orig.mgz -reg register.dat -overlay TP.nii.gz -fthresh 0.5 -surfs
you should see the mask in red. It should also intersect the surface. If not, then something is wrong.
doug
kwa1@rice.edu wrote:
Okay following that, here is what I did:
- bbregister the FSL anatomical contrast to the FS space of that
subject (using the 12 dof as suggested). Checking registration seemed to be fine.
- did mri_vol2surf of the FSL-derived mask (because when I
extract masks from fslview they are extracted into volumes), then I convert that surface to a label with mri_cor2label.
- I extracted the registration.dat file generated by bbregister
(with the 4x4 affine matrix) and ran the mri_label2vol using the registration matrix on the label developed in (2).
- Ran (2) again to get the registered label file.
When I mapped the label onto the subject's pial surface it did not map on. Any ideas why, though?
Additionally, if I wanted to extract an fsaverage TP mask using the mri_annotation2label, how would I convert that .label file to a subject's native space (for example to correspond with the labels I derive from each subject)?
Thank you, Kareem
Quoting Douglas N Greve greve@nmr.mgh.harvard.edu:
Don't use the temporal pole mask as the input to the registration. You need something with some anatomical contrast. The TP mask will basically be an empty image with a few 1s. Load a gray-scale image in fslview that matches the TP mask. Then bbregister to that gray scale image (with --12 for 12 dof). Check to make sure that the registration is ok. Then run mri_label2vol using the TP mask and the registration.
doug
kwa1@rice.edu wrote:
Hi Bruce,
Thank you for your reply. The labels are coming from me exporting the temporal pole label for a subject from fslview -- so that would mean it's in standard space? All I would need is that 4x4 affine matrix in order to calculate transformations but I'm not sure how to calculate that kind of matrix between the different spaces. However since I'm also trying to overlay the converted label to individual subjects (e.g. tksurfer visualization) do I need structural or native subject space conversion?
Specifically the label came from a t1 image and it should register quite well, the scans were fine and we have no reason to think they shouldn't register well (no abnormalities etc.).
Thank you very much, Kareem
Quoting Bruce Fischl fischl@nmr.mgh.harvard.edu:
> Hi Kareem, > > can you explain a bit more about where these atlas-derived > labeling comes from? Is it in MNI space? Is there any reason > to think it will register well to this subject? Or is it in > this individual's space? If the latter, you probably want to > give -t2 to bbregister since in your case as I understand it > the gray matter is bright (e.g. 1) and the white matter is > dark (0?). Is that correct? > > cheers > Bruce > > > On Mon, 30 May 2011 kwa1@rice.edu wrote: > >> Hi Doug, >> >> The bbregister didn't work out well, it didn't register properly but I >> must be putting in some erroneous argument: >> >> bbregister --s SUBJ1 --mov TP.nii.gz --reg register.dat >> --init-fsl --t1 >> >> The final check with tkregister shows the temporal pole mask several >> slices off bilaterally and rotated about 30-40 degrees CW and also >> rotated spatially from frontal to posterior brain. >> >> I also tried a flirt command in FSL but couldn't not get a proper 4x4 >> affine matrix to use. Any ideas on why the registration isn't working? >> >> Thank you, >> Kareem >> >> >> Quoting Douglas Greve greve@nmr.mgh.harvard.edu: >> >>> You use bbregister to compute the registration. You will need a >>> template for your atlas (eg, an average anatomy). >>> >>> doug >>> >>> On 5/27/11 7:09 PM, kwa1@rice.edu wrote: >>>> Do I feed the registration matrix into another FS command or does >>>> bbregister use the reg matrix to transform the atlas mask into FS >>>> anatomical space? I will try the bbregister with 12 DOF as you >>>> recommend too. >>>> >>>> Thank you, >>>> Kareem >>>> >>>> >>>> Quoting Douglas N Greve greve@nmr.mgh.harvard.edu: >>>> >>>>> Yes, there's no other way for FS to know where data should go -- >>>>> it has no idea how you derived the input. You can try using >>>>> bbregister with 12 DOF to register the atlas template to the FS >>>>> anatomical. This will give you a registration matrix. Check it >>>>> visually to make sure that it is reasonably accurate. >>>>> >>>>> doug >>>>> >>>>> kwa1@rice.edu wrote: >>>>>> Hmm, well I derived the volume from an FSL atlas which I >>>>>> exported, binarized, then fed into the commands I listed >>>>>> initially. Do I need to register the temporal pole volume to some >>>>>> FS space? And how would I so? >>>>>> >>>>>> Best, >>>>>> Kareem >>>>>> >>>>>> >>>>>> >>>>>> Quoting Douglas N Greve greve@nmr.mgh.harvard.edu: >>>>>> >>>>>>> Then it means that the registration is incorrect. How did you >>>>>>> create the reg? >>>>>>> >>>>>>> kwa1@rice.edu wrote: >>>>>>>> No it doesn't look right. It is supposed to map on to the >>>>>>>> temporal pole but doesn't do so. I don't think it's in the >>>>>>>> right space. Is there a command we are perhaps missing to bring >>>>>>>> it to native FS space? >>>>>>>> >>>>>>>> Best, >>>>>>>> Kareem >>>>>>>> >>>>>>>> >>>>>>>> Quoting Douglas N Greve greve@nmr.mgh.harvard.edu: >>>>>>>> >>>>>>>>> It looks like your commands are correct. When you load >>>>>>>>> mask.surf.nii.gz onto the surface, does it look ok? >>>>>>>>> >>>>>>>>> doug >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> kwa1@rice.edu wrote: >>>>>>>>>> Hello, >>>>>>>>>> >>>>>>>>>> I've derived a .nii.gz atlas-based volume for a specific part >>>>>>>>>> of the cortical surface. I've binarized the volume and now I >>>>>>>>>> want to convert the file to a .label file that I can use to >>>>>>>>>> identify particular vertices in native FS space for that >>>>>>>>>> same subject. >>>>>>>>>> >>>>>>>>>> This is what I have tried so far: >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> mri_vol2surf --src mask.nii.gz --srcreg /tmp/reg.dat --surf >>>>>>>>>> pial --projfrac 0.5 --hemi lh --out mask.surf.nii.gz >>>>>>>>>> >>>>>>>>>> mri_cor2label --i mask.surf.nii.gz --id 1 --l mask.label >>>>>>>>>> --surf SUBJ1 lh pial >>>>>>>>>> >>>>>>>>>> However when I overlayed the .label onto SUBJ1 the vertices >>>>>>>>>> did not match up and was completely off the region of >>>>>>>>>> interest on the cortical surface. >>>>>>>>>> >>>>>>>>>> Does anyone know how to fix this or how to proceed with >>>>>>>>>> conversion differently? >>>>>>>>>> >>>>>>>>>> Thank you very much, >>>>>>>>>> Kareem >>>>>>>>>> >>>>>>>>>> _______________________________________________ >>>>>>>>>> Freesurfer mailing list >>>>>>>>>> Freesurfer@nmr.mgh.harvard.edu >>>>>>>>>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >>>>>>>>> >>>>>>>>> -- >>>>>>>>> Douglas N. Greve, Ph.D. >>>>>>>>> MGH-NMR Center >>>>>>>>> greve@nmr.mgh.harvard.edu >>>>>>>>> Phone Number: 617-724-2358 Fax: 617-726-7422 >>>>>>>>> >>>>>>>>> Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting >>>>>>>>> FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> The information in this e-mail is intended only for the person >>>>>>>>> to whom it is >>>>>>>>> addressed. If you believe this e-mail was sent to you in error >>>>>>>>> and the e-mail >>>>>>>>> contains patient information, please contact the Partners >>>>>>>>> Compliance HelpLine at >>>>>>>>> http://www.partners.org/complianceline . If the e-mail was >>>>>>>>> sent to you in error >>>>>>>>> but does not contain patient information, please contact the >>>>>>>>> sender and properly >>>>>>>>> dispose of the e-mail. >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Douglas N. Greve, Ph.D. >>>>>>> MGH-NMR Center >>>>>>> greve@nmr.mgh.harvard.edu >>>>>>> Phone Number: 617-724-2358 Fax: 617-726-7422 >>>>>>> >>>>>>> Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting >>>>>>> FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html >>>>>>> >>>>>>> >>>>>>> >>>>>> >>>>>> >>>>>> >>>>>> >>>>> >>>>> -- >>>>> Douglas N. Greve, Ph.D. >>>>> MGH-NMR Center >>>>> greve@nmr.mgh.harvard.edu >>>>> Phone Number: 617-724-2358 Fax: 617-726-7422 >>>>> >>>>> Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting >>>>> FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html >>>>> >>>>> >>>>> >>>> >>>> >>>> >>>> >>> >>> >> >> >> >> _______________________________________________ >> Freesurfer mailing list >> Freesurfer@nmr.mgh.harvard.edu >> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer >> >> >> > >
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
-- Douglas N. Greve, Ph.D. MGH-NMR Center greve@nmr.mgh.harvard.edu Phone Number: 617-724-2358 Fax: 617-726-7422
Bugs: surfer.nmr.mgh.harvard.edu/fswiki/BugReporting FileDrop: www.nmr.mgh.harvard.edu/facility/filedrop/index.html
