Hi Bruce,
Thank you for this quick check. Your results confirm my suspicion: "aparc.a2005s+aseg.mgz" does not have this wm-label. So, after trying every segmentation file I found in /mri (aparc+aseg.mgz, wmparc.mgz, aparc.a2005s+aseg.mgz) without success, I asked me
Where can I find or how can I create the necessary segmentation volume to extract 3136 or 4181?
Somebody has an idea?
kind regards,
Markus
-----Ursprüngliche Nachricht----- Von: Bruce Fischl [mailto:fischl@nmr.mgh.harvard.edu] Gesendet: Mittwoch, 10. Dezember 2008 19:16 An: Markus Röhl Betreff: Re: AW: AW: [Freesurfer] mri_extract_label 4181
Hi Markus,
are you sure there is a 3136 in the data? That is a wm label, and is probably in the wmparc.mgz, not this one. The 1136 is the ctx label. I did:
matlab
[v,M,mr] = load_mgh('aparc.a2005s+aseg.mgz'); find(v(:) == 3136)
ans = Empty matrix: 0-by-1
find(v(:) == 1136)
ans = 4607151 4672429 4672430 4672687 4672944 4737453 4737708 4737709 4737963 4737965 4737966 4737967 . . .
On Wed, 10 Dec 2008, Markus Röhl wrote:
Hi Bruce,
the command line that fails is:
mri_extract_label ./aparc.a2005s+aseg.mgz 3136 ./test.nii.gz
This works fine:
mri_extract_label ./aparc.a2005s+aseg.mgz 1136 ./test2.nii.gz
Best regards,
Markus
-----Ursprüngliche Nachricht----- Von: Bruce Fischl [mailto:fischl@nmr.mgh.harvard.edu] Gesendet: Mittwoch, 10. Dezember 2008 18:38 An: Markus Röhl Betreff: Re: AW: [Freesurfer] mri_extract_label 4181
Hi Markus
can you email me the aparc.a2005s+aseg.mgz and the exact command line that fails? Bruce On Wed, 10 Dec 2008, Markus Röhl wrote:
Hi
for the purpose of illustration I change to
3136 wm-lh-G_temp_sup-Planum_tempolare
The corresponding gray matter part (1136) can seen in cortex.tiff. It is labeled green and can be found quite in the middle of the picture right
from
the red labeled S_temporal_transverse. The successful extraction of the
1136
ctx-lh-G_temp_sup-Planum_tempolare is shown in success.png.
The corresponding white matter of this landmark (3136 wm-lh-G_temp_sup-Planum_tempolare) can also be seen on the right side of WM.tiff as a green labeled area superior of the pink labeled lateral
aspect
of G_temp_sup. (the 3181 wm-lh-S_temporal_transverse is also present as a very very small red labeled area between the small blue and green areas
in
the middle of WM.tiff).
I am not able to extract 3136 and I think this seems to hold also for the other 3XXX or 4XXX wm-labels from aparc.a2005s+aseg.mgz.
I extracted 2034 ctx-rh-transversetemporal from "aparc+aseg.mgz" and
analog
4034 wm-lh-transversetemporal from "wmparc.mgz". I assumed that there
must
be a corresponding file of "aparc.a2005s+aseg.mgz" like "wmparc.a2005s+aseg.mgz" to extract white matter labels. "aparc.a2005s+aseg.mgz" seems to work only for ctx- but not for
wm-labels,
and I don't found something else. The problem is not subject-related.
Regards,
Markus
-----Ursprüngliche Nachricht----- Von: Christophe Destrieux [mailto:destrieux@med.univ-tours.fr] Gesendet: Mittwoch, 10. Dezember 2008 14:24 An: Bruce Fischl Cc: Markus Röhl; freesurfer@nmr.mgh.harvard.edu Betreff: Re: [Freesurfer] mri_extract_label 4181
Hi Bruce and Markus
S_temporal_transverse is constant and should be present on the surface ;
cheers
HI Markus,
that's one of Christophe's labels, so would only be in the aparc.a2005s+aseg.mgz. Have you cheched the aparc.a2005s.annot on the surface to verify that it is present there? There are also some labels that don't appear on every subject, Christophe would know if this is one of them.
cheers, Bruce
On Wed, 10 Dec 2008, Markus Röhl wrote:
Dear all,
I used mri_extract_label recently to get NIfTI -binary maps for certain anatomical landmarks, e.g. the Gyrus temporalis inferior of the left hemisphere (cortex) which index is listed in FreeSurferColotLUT.txt:
mri_extract_label $FREESURFER_HOME/subjects/VP1/mri/aparc+aseg.mgz
1009
$FREESURFER_HOME/subjects/ROI/VP1/CL_IT.nii.gz
This works fine. Now I want to extract the labels of anatomical landmarks like
4181 wm-rh-S_temporal_transverse
I tried every segmentation file I found (aparc+aseg.mgz, wmparc.mgz, aparc.a2005s+aseg.mgz), but only ended up with empty maps. Where I can find or how can I create the necessary segmentation volume?
Regards,
Markus