Hi Guang,
even in the longitudinal stream the asegs (and all images) are in the TP native space. They still need to be registered (either to each other or to the base space). The base space is recommended, because that way both images are treated the same (both are resampled). The registrations are available in the base directory baseid/mri/transforms.
Here is the corresponding passage from the WIKI page http://surfer.nmr.mgh.harvard.edu/fswiki/LongitudinalProcessing
--------------------------- It is possible to check the registrations and template creation either using the norm.mgz or orig.mgz of all time points. These images are first mapped to and resampled in the space of the base with:
mri_convert -rl <baseid>/mri/norm.mgz -at <baseid>/mri/transforms/<tpNid>_to_<baseid>.lta <tpNid>/mri/norm.mgz <tpNid>_to_<baseid>.norm.mgz
The resampled files <tpNid>_to_<baseid>.norm.mgz can than be compared with each other or with the base's norm_template.mgz using e.g. tkmedit -f file1 -aux file2. If you want to compare aseg.mgz files, make sure, you use nearest for the resample type:
mri_convert -rl <baseid>/mri/aseg.mgz -rt nearest -at <baseid>/mri/transforms/<tpNid>_to_<baseid>.lta <tpNid>/mri/aseg.mgz <tpNid>_to_<baseid>.aseg.mgz
----------------------------
good luck, Martin
On Wed, 2009-09-23 at 13:54 -0500, Guang Zeng wrote:
I think I still need register the second tp aseg.mgz to the first tp aseg.mgz, even they are longitudinal-stream analyzed. The dice coefficients are higher after registration.
From: freesurfer_zg@hotmail.com To: nicks@nmr.mgh.harvard.edu Date: Wed, 23 Sep 2009 13:04:59 -0500 CC: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] how to use mri_compute_seg_overlap
Hello, Nick,
Yes, I can get the results now, I set the wrong path.
Just want to double check, I have send the two subjects (from different time points) to the FreeSurfer longitudinal stream, so the resulted two aseg.mgz files have been registered to the base-template, right?
Thanks a lot! Guang
From: nicks@nmr.mgh.harvard.edu To: freesurfer_zg@hotmail.com Date: Wed, 23 Sep 2009 13:01:52 -0400 CC: freesurfer@nmr.mgh.harvard.edu Subject: Re: [Freesurfer] how to use mri_compute_seg_overlap
Guang,
So you did not get output that looked like this?
logging overall Dice to text Jaccard Coefficients: correct ratio for label 2 = 1 correct ratio for label 41 = 1 . . correct ratio for label 44 = 1 mean +/- std = 1.0000 +/- 0.0000 Dice Coefficients: label 2 = 1 label 41 = 1 . . . label 44 = 1 mean +/- std = 1.0000 +/- 0.0000 Overall subcortical dice = 1.0000
(in this example, i compared the same aseg, so dice will be 1).
also, make sure your second time point aseg is registered to the
first.
if it is not, the your dices will be low, because its just comparing voxel to voxel.
Nick
On Wed, 2009-09-23 at 11:30 -0500, Guang Zeng wrote:
Hi, there,
I have two subjects from two time points, I'd like to compare
their
overlap ratio of FreeSurfer segmented hippocampus. I try to use the command mri_compute_seg_overlap to do it,
My command line is: mri_compute_seg_overlap -olog text tp1/mri/aseg.mgz
tp2/mri/aseg.mgz
Is it correct? I ran it, but nothing happen.
Thanks!
Guang
Insert movie times and more without leaving Hotmail®. See how. _______________________________________________ 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
Insert movie times and more without leaving Hotmail®. See how.
Insert movie times and more without leaving Hotmail®. See how. _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer