Dear Freesurfers,
I haven't been able to run recon-all in a set of subjects. All them stop at the point of the talairach transform verification. I uploaded one of the files here (5.3MB): http://dl.getdropbox.com/u/74925/20070404.nii.zip
For that volume, the transform.xfm file looked indeed odd (m22=7.2??):
1.39488649 0.08332049 0.01164643 3.51684260 -0.14383204 7.20834160 -2.21084833 11.16247082 -0.16614650 1.02134335 1.85208952 -45.74753189
Using flirt (flirt -in 20070404.nii -ref MNI152_T1_1mm -out 20070404trans.nii -omat 20070404trans.mat -bins 256 -cost corratio -searchrx -90 90 -searchry -90 90 -searchrz -90 90 -dof 12 -interp trilinear), the transformation was:
1.05366 -0.050378 0.0606504 4.05766 0.0446432 1.09476 0.1321 -28.652 -0.058117 -0.180136 1.09158 31.7066 0 0 0 1
The voxel size of the volume is 1.17x1.17x1.2, so it shouldn't be a problem of resolution.
When I converted the volume to mgz (mri_convert 20070404.nii mri/orig/001.mgz), the orientation was wrong; but changing it using the Volume Tools in tkmedit did not alter the outcome...
Finally, I also used the --notal-check switch in recon-all, but then the script failed before completing skull stripping...
thank you very much in advance for your help!
roberto
Roberto,
Your scan has a FOV greater than 256, which freesurfer cannot handle by default. So you must add this to your initial recon-all command line along with the other flags (-all, or -autorecon1):
-cw256
this crops the image to 256, and, in most cases, crops correctly (ie, keeps the whole brain). I've tested your subject data and it crops correctly.
fyi, beginning with v4.1.0, recon-all checks for fov>256 and exits with an error message indicating that people should add the -cw256 flag. it doesnt just do this by default because it doesnt always work (so its better to make people aware of the problem upfront, otherwise you get strange downstream problems like you were seeing).
Nick
On Fri, 2009-02-06 at 21:05 +0100, roberto toro wrote:
Dear Freesurfers,
I haven't been able to run recon-all in a set of subjects. All them stop at the point of the talairach transform verification. I uploaded one of the files here (5.3MB): http://dl.getdropbox.com/u/74925/20070404.nii.zip
For that volume, the transform.xfm file looked indeed odd (m22=7.2??):
1.39488649 0.08332049 0.01164643 3.51684260 -0.14383204 7.20834160 -2.21084833 11.16247082 -0.16614650 1.02134335 1.85208952 -45.74753189
Using flirt (flirt -in 20070404.nii -ref MNI152_T1_1mm -out 20070404trans.nii -omat 20070404trans.mat -bins 256 -cost corratio -searchrx -90 90 -searchry -90 90 -searchrz -90 90 -dof 12 -interp trilinear), the transformation was:
1.05366 -0.050378 0.0606504 4.05766 0.0446432 1.09476 0.1321 -28.652 -0.058117 -0.180136 1.09158 31.7066 0 0 0 1
The voxel size of the volume is 1.17x1.17x1.2, so it shouldn't be a problem of resolution.
When I converted the volume to mgz (mri_convert 20070404.nii mri/orig/001.mgz), the orientation was wrong; but changing it using the Volume Tools in tkmedit did not alter the outcome...
Finally, I also used the --notal-check switch in recon-all, but then the script failed before completing skull stripping...
thank you very much in advance for your help!
roberto
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Roberto,
when you bring up the volume does it appear correctly in tkmedit? That is, is the coronal view really coronal, etc....?
Bruce On Fri, 6 Feb 2009, roberto toro wrote:
Dear Freesurfers,
I haven't been able to run recon-all in a set of subjects. All them stop at the point of the talairach transform verification. I uploaded one of the files here (5.3MB): http://dl.getdropbox.com/u/74925/20070404.nii.zip
For that volume, the transform.xfm file looked indeed odd (m22=7.2??):
1.39488649 0.08332049 0.01164643 3.51684260 -0.14383204 7.20834160 -2.21084833 11.16247082 -0.16614650 1.02134335 1.85208952 -45.74753189
Using flirt (flirt -in 20070404.nii -ref MNI152_T1_1mm -out 20070404trans.nii -omat 20070404trans.mat -bins 256 -cost corratio -searchrx -90 90 -searchry -90 90 -searchrz -90 90 -dof 12 -interp trilinear), the transformation was:
1.05366 -0.050378 0.0606504 4.05766 0.0446432 1.09476 0.1321 -28.652 -0.058117 -0.180136 1.09158 31.7066 0 0 0 1
The voxel size of the volume is 1.17x1.17x1.2, so it shouldn't be a problem of resolution.
When I converted the volume to mgz (mri_convert 20070404.nii mri/orig/001.mgz), the orientation was wrong; but changing it using the Volume Tools in tkmedit did not alter the outcome...
Finally, I also used the --notal-check switch in recon-all, but then the script failed before completing skull stripping...
thank you very much in advance for your help!
roberto
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Hi Nick and Bruce,
thank you for your help. Now everything's working :)
first I tried just the -cw256 flag and it didn't work, but then I erased all the old transformation files and volumes created by freesurfer, and everything went fine...
and yes, the orientation is correct.
thanks again! roberto
On Sat, Feb 7, 2009 at 3:37 PM, Bruce Fischl fischl@nmr.mgh.harvard.edu wrote:
Hi Roberto,
when you bring up the volume does it appear correctly in tkmedit? That is, is the coronal view really coronal, etc....?
Bruce On Fri, 6 Feb 2009, roberto toro wrote:
Dear Freesurfers,
I haven't been able to run recon-all in a set of subjects. All them stop at the point of the talairach transform verification. I uploaded one of the files here (5.3MB): http://dl.getdropbox.com/u/74925/20070404.nii.zip
For that volume, the transform.xfm file looked indeed odd (m22=7.2??):
1.39488649 0.08332049 0.01164643 3.51684260 -0.14383204 7.20834160 -2.21084833 11.16247082 -0.16614650 1.02134335 1.85208952 -45.74753189
Using flirt (flirt -in 20070404.nii -ref MNI152_T1_1mm -out 20070404trans.nii -omat 20070404trans.mat -bins 256 -cost corratio -searchrx -90 90 -searchry -90 90 -searchrz -90 90 -dof 12 -interp trilinear), the transformation was:
1.05366 -0.050378 0.0606504 4.05766 0.0446432 1.09476 0.1321 -28.652 -0.058117 -0.180136 1.09158 31.7066 0 0 0 1
The voxel size of the volume is 1.17x1.17x1.2, so it shouldn't be a problem of resolution.
When I converted the volume to mgz (mri_convert 20070404.nii mri/orig/001.mgz), the orientation was wrong; but changing it using the Volume Tools in tkmedit did not alter the outcome...
Finally, I also used the --notal-check switch in recon-all, but then the script failed before completing skull stripping...
thank you very much in advance for your help!
roberto
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Nick Schmansky -
roberto toro