okay, let's go back to the beginning then. How did you generate the
orig.mgz volume? What format was the input data in? If the orig.mgz has
the wrong orientation then everything will fail. In general we are pretty
good about reading the orientation info out of the image header, so
something is pretty wrong.
On Tue, 1 May 2007,
Jackie_rwth wrote:
Hi Jenni,
all the mgz. are all in the wrong orientaions both in tkregister2 and
tkmedit.
except the talairach volume which I fixed, they are in the correct
orientation.
Jackie
Jenni Pacheco schrieb:
Hi Jackie,
So that command will bring up your subject and the talairach subject and
allow you to try and register them - it is not something before and after
talairach.
Going back to the problem of things oriented the wrong way. All the
volumes (orig.mgz, brainmask.mgz, T1.mgz etc) should be oriented the same
as each other - is this true?
Are you seeing them oriented correctly in tkmedit but incorrectly in
tkregister2?
Jenni
On Tue, 1 May 2007, Jackie_rwth wrote:
Hi Jenni,
This is the command line I used for checking talairach transform.
tkregister2 --mgz --s xxxxx --fstal.
cheers,
Jackie
Jenni Pacheco schrieb:
Hi Jackie,
What command line are you using to open tkregister2 to check your
talairach
transform? I'm not sure that the moveable volume is really the one after
talairach transform - likely (depending on the command you used) the
moveable volume is the talairach subject, the target volume that you want
your subject to match.
Can you send your command line.
Thanks,
Jenni
-----Original Message-----
From: Jackie_rwth [mailto:[EMAIL PROTECTED] Sent: Tuesday,
May 01, 2007 3:02 AM
To: Bruce Fischl
Cc: Jenni Pacheco; freesurfer@nmr.mgh.harvard.edu
Subject: Re: [Freesurfer] Talairach transform
Hi Bruce,
Yes, I have renamed the T1.xfm to talairach.xfm and put it in
mri/transforms dir, and also when I ran autorecon1 and put the -
notalairach flag after fixing.
Yes, I mean after fixing the tal transform.
The orientation of the orig.mgz , T1.mgz, and brain.mgz are in another
orietation as the "movebel volume"(the one after tal transform).
eg. when I used tkregister2 the COMPARE button to check the tal
transfrom, the orietations of two images ( one is after tal transform,
the other is T1.mgz)
are not identical, (when I pressed "coronal" button, the movble voulme is
in coronal direction but the target volume is in horozontal direction,
...etc.)
And then I used tkmedit to check the orig-, T1-, and brain.mgz, they are
all in the same situation but the wrong one.(coronal view became
horozontal view, horozontal view become sagittal view , etc...)
Any sugguestion will be helpful!
cheers,
Jackie
Bruce Fischl schrieb:
Hi Jackie,
you need to rename the T1.xfm to be talairach.xfm and put it in the
mri/transforms dir.
When you say "after this step, the orientation of the target volume is
mixed up" do you mean after applying the tal transform? The orientation
of all the .mgz (e.g. orig.mgz, T1.mgz, brain.mgz) should be identical.
cheers,
Bruce
On Mon, 30 Apr 2007, Jackie_rwth wrote:
Hi Jenni,
The orig volume is in the correct orientation when I viewed it with
tikmedit. But after I fixed the Talairach transform by using
mri_convert T1.mgz T1.mnc , mritotal -protocol icbm T1.mnc T1.xfm.,(and
I viewed it in tkregister2, which looks quite fine), then I ran
recon-all -subjid xxxx -autorecon1 -notalairach .
But after this step, the orientation of the target volume is mixed up,
same as the orig.mgz volume.(before fixing Talairach the orig voulme
are correct)
p.s the orientation of the moveble volumes are correct after fixing
They are nifti files and acquired by Siemens Sonata.
(We already used the avwswapdim to reorient them before running, and we
also put mark on the brain so that we could tell L and R)
Thank you for your help!
Jackie
Jenni Pacheco schrieb:
Hi,
I'm a little confused, so please correct me if I've made the wrong
assumptions. It sounds like your problem is becuase your orig volume
is in the wrong orientation when you view it in tkmedit (ie., the
coronal view in tkmedit does not show a coronal view, but shows the
sagital view). It seems, also, that this was not always the case -
until you fixed the talairach transform. Can you give me the exact
mri_convert command that you used to fix the talaraich?
What type of scans are these? Siemens, GE? what format were they in
before they were in mgz? It may be the case that you need to
re-convert them to correct the orientation. To do this, though, you'd
need to be sure you could identify the right side from the left side.
You can use some switched in mri_convert - I'd be glad to help you
figure that command out if this ends up to be the problem.
Jenni
On Mon, 30 Apr 2007 [EMAIL PROTECTED] wrote:
Dear list,
After fixing the Talairach transform automatically, (by mri_convert ,
the final objective funciton value is fine, ca.0.06)
So I replaced the talairach.xfm with T1.xfm and named it as
talairach.xfm, and it looked pretty fine with tkregister2.
Then I runed recon-all -subjid xxxx -autorecon1 -notalairach, and
then viewed it with tkmedit and tkreigister2.
Under tkregister2 the talairach transform looks pretty good but not
the target.(the oreientations of the target are totally mixed up,
those of the moveble volume are correct).
I checked orig.mgz are also mixed up, coronal becomes sagittal,
sagittal becomes horizontal, etc.
Any suggestions about this ?
Thanks.
Jackie
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
