help
help > RE: Transform individual ROIs to MNI space
Nov 11, 2015 08:11 AM | Jalmar Teeuw
RE: Transform individual ROIs to MNI space
Thanks again! That is really
useful.
Originally posted by Alfonso Nieto-Castanon:
Originally posted by Alfonso Nieto-Castanon:
Hi
Jalmar,
Just for reference, I am attaching a patch (this patch is for release 15g, simply copy the file to the conn distribution folder overwriting the file with the same name there) that will create an additional "centering_*.mat" file containing the translation affine-matrix transformation applied to your structural/functional data when using any of the "set origin to (0,0,0)" or "manual reorient" options. To apply the same transformation to a different file you can then use something like:
load('centering_mystructuralfile.mat','R');
M = spm_get_space('myotherfile.nii');
spm_get_space('myotherfile.nii',R*M);
This option will work even if myotherfile.nii and mystructuralfile.nii do not have the same voxel dimensions (as long as they were originally coregistered before centering)
Hope this helps
Alfonso
Originally posted by Jalmar Teeuw:
Just for reference, I am attaching a patch (this patch is for release 15g, simply copy the file to the conn distribution folder overwriting the file with the same name there) that will create an additional "centering_*.mat" file containing the translation affine-matrix transformation applied to your structural/functional data when using any of the "set origin to (0,0,0)" or "manual reorient" options. To apply the same transformation to a different file you can then use something like:
load('centering_mystructuralfile.mat','R');
M = spm_get_space('myotherfile.nii');
spm_get_space('myotherfile.nii',R*M);
This option will work even if myotherfile.nii and mystructuralfile.nii do not have the same voxel dimensions (as long as they were originally coregistered before centering)
Hope this helps
Alfonso
Originally posted by Jalmar Teeuw:
Hi Alfonso,
Thanks for the quick response. We tried option 2, but we do have the problem that we require the 'set origin to (0,0,0)' option during preprocessing, otherwise SPM complains about to little overlap between scans during structural normalisation. How is this 'set origin' step performed? Is there a way to get the translation parameters? Because if we can apply it to the structural scan and then the same translation to our tissue maps prior to using CONN it should work with option 2.
Regards,
Jalmar
Thanks for the quick response. We tried option 2, but we do have the problem that we require the 'set origin to (0,0,0)' option during preprocessing, otherwise SPM complains about to little overlap between scans during structural normalisation. How is this 'set origin' step performed? Is there a way to get the translation parameters? Because if we can apply it to the structural scan and then the same translation to our tissue maps prior to using CONN it should work with option 2.
Regards,
Jalmar
Threaded View
| Title | Author | Date |
|---|---|---|
| Matthew Defenderfer | Oct 27, 2015 | |
| Alfonso Nieto-Castanon | Nov 10, 2015 | |
| Jalmar Teeuw | Nov 10, 2015 | |
| Alfonso Nieto-Castanon | Nov 10, 2015 | |
| Jalmar Teeuw | Nov 11, 2015 | |
| Jalmar Teeuw | Nov 10, 2015 | |
| Alfonso Nieto-Castanon | Nov 10, 2015 | |
| Jalmar Teeuw | Nov 9, 2015 | |