Hi, I processed about 80 subjects with -prep stage and then checked the bbregister registrations with
tkregister2 --mov /study5/aa-scratch/TEENEMO/freesurfer_local/toot/057/dmri/dwi.nii.gz --reg /study5/aa-scratch/TEENEMO/freesurfer_local/toot/057/dmri/xfms/anatorig2diff.bbr.dat --surf
found 17 subjects with bad registrations, as always because the fsl registration initialization was bad.
so now i manually fix the registeration and use the manual fix for the initialization and rerun bbregister and get a good registration.
so now i need to rerun -prep phase but have it use the good registration instead of computing a new bbregister registration.
will trac-all -prep -nointra -c configfile accomplish this or do i need to get in and tinker with the scripts to make it run everything including computing new fsl and inverse registration matricies but using my bbregister good registration.
thanks
greg
Hi Greg,
You can do the following to skip eddy current correction and intra subject registration:
trac-all -inter -masks -tensor -prior -c /path/to/configurationfile/config_file
If you want to get tracula based files for the bbregister you've run with manual fixes, can you please let me know the command line you used for running bbregister?
This is what is typically used in tracula.
1. Set SUBJECTS_DIR variable
setenv SUBJECTS_DIR /path/to/recons/
2. Register low-b to Anatomical with bbregister
bbregister --s <SubjectID> --init-fsl --dti --mov <SubjectID>/dmri/dwi.nii.gz --reg <SubjectID>/dmri/xfms/anatorig2diff.bbr.dat --fslmat <SubjectID>/dmri/xfms/diff2anatorig.bbr.mat
After the bbregister step tracula expects, to find <SubjectID>/dmri/xfms/anatorig2diff.bbr.dat and <SubjectID>/dmri/xfms/diff2anatorig.bbr.mat in the tracula subject directory. Since you've already finished this step, you can probably rename your .dat and .mat files as above and proceed with the following steps[3-6].
3. Compute inverse
convert_xfm -omat <SubjectID>/dmri/xfms/anatorig2diff.bbr.mat -inverse <SubjectID>/dmri/xfms/diff2anatorig.bbr.mat
4. Compue Transformation beteween FSL's Flipped anatomical to low-b
convert_xfm -omat <SubjectID>/dmri/xfms/anat2diff.bbr.mat -concat <SubjectID>/dmri/xfms/anatorig2diff.bbr.mat <SubjectID>/dmri/xfms/anat2anatorig.mat
5. Compute inverse between low-b and fsl's flipped anatomical
convert_xfm -omat <SubjectID>/dmri/xfms/diff2anat.bbr.mat -inverse <SubjectID>/dmri/xfms/anat2diff.bbr.mat
6. Skip Eddy current and Intra Subject registration and run the rest of the steps in tracula as you've got all the registration files from the intra subject registration
trac-all -inter -masks -tensor -prior -c /path/to/configurationfile/config_file
To do it for all 17 subjects you can just create a shell script to run the steps 3-5 for all the subjects and then run trac-all to do the rest of the processing.
If you have any questions about the outputs from tracula or the files it creates/expects after every stage, please refer to the following page:
http://surfer.nmr.mgh.harvard.edu/fswiki/trac-all
Let us know if you have any questions.
-Priti
Hi, I processed about 80 subjects with -prep stage and then checked the
bbregister registrations with
tkregister2 --mov /study5/aa-scratch/TEENEMO/freesurfer_local/toot/057/dmri/dwi.nii.gz
--reg
/study5/aa-scratch/TEENEMO/freesurfer_local/toot/057/dmri/xfms/anatorig2diff.bbr.dat
--surf
found 17 subjects with bad registrations, as always because the fsl
registration initialization was bad.
so now i manually fix the registeration and use the manual fix for the
initialization and rerun bbregister and get a good registration.
so now i need to rerun -prep phase but have it use the good registration
instead of computing a new
bbregister registration.
will trac-all -prep -nointra -c configfile accomplish this or do i need to get in and tinker with the scripts to make it run
everything including computing new fsl and inverse registration matricies
but using my bbregister good registration.
thanks
greg _______________________________________________ Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu