I want to ensure that the imaging data I have is in the correct orientation prior to placing the brain into standard space and segmenting out certain regions of the brain.
I have used the fslorient utilites (-getorient, -getsform and -getqform) and got the following information for my file con001_all.nii :
-getorient provided that the image was in the radiological convention
-getsform provided the following values:
-1.19922 0 0 0 0 1.19922 0 0 0 0 1.19767 0 0 0 0 1
-getqform provided:
1.19922 0 0 0 0 1.19922 0 0 0 0 1.19767 0 0 0 0 1
In fslview, I have the sagittal view placed in the manner I want; however, I would like to rotate the axial view and the coronal view (I have attached a screen shot to display their current orientation).
When I go to run the fslreorient2std utility, I get an error "orientation information not stored in con001_all! Cannot reslice without orientation information (ie. need valid labels in FSLview). The NIFTI image must contain a non-zero code for either the sform or the qform - check your reconstruction/conversion software to try and fix this"
Initially, I had .ima files and used MRIcro (dcm2niigui.exe) to convert the files to NIFTI format (.nii). I then used fslmerge -z <output> <file1 file2...file (N)> to merge the files into one file called that I called con001_all. This produced con001_all.nii.gz. I then used fslchfiletype <NIFITI> con001_all.nii.gz to remove the .gz extension, leading to con001_all.nii
I was wondering three things based on my error message; 1) if the fslorient utilities can tell me that the images are in the radiological convention, and then provide sform and qform numbers, how is it that "orientation information is not stored in con001_all" (also when I look at FSLview, in the bottom left I see "coordinate space: unknown") and 2) was it possible something went wrong with the conversion or merging of files? Finally 3) how would I fix this? I have read through the orientation-related utilities on the FSL-FMRIB site, but cannot locate an answer that had proven to be helpful. Also, re-scanning is not an option. This is a data set I inherited.
Any help with this would be greatly appreciated. Thank you!
Victor
|