Hi,
I am attempting to morph deterministic tractography results (.trk file) from the individual diffusion space into the average brain cvs_avg35_inMNI152. However, something goes wrong, as the output trk file is greatly misplaced / strongly distorted / wrongly sized.
The input file (dsi60.trk, a TrackVis trk file with an angle threshold of 60 degrees) loads fine in Trackvis, as do the inref and outref volumes (in .nii format). The rigid body (.mat) and nonlinear (.m3z) registrations also appear ok. The mat file (or more precisely, the corresponding register.dat file) looked fine with tkregister2 and the latter was successfully used to morph the corresponding probabilistic tractography results from the individual diffusion space to cvs_avg35_inMNI152 (= the same transformation I am trying to do here).
Any suggestions what I am doing wrong? For example, could this be related to inref/outref voxel sizes or coordinate systems?
Just in case, I already tried using the inverse rigid registration (anat2diff.mat) and the FDT-generated files inside DSI.bedpostX/xfms (diff2str.mat and its inverse) but the results were in all cases very much wrong, just in different ways.
FS 5.3 (nmr-std-env) on machine avml cd $SUBJECTS_DIR
dmri_trk2trk \ --in $SUBJECT/DSI/dsi60.trk \ --out TMSDSIGA_19Jul14/${SUBJECT}inCVS35MNI152_dsi60.trk \ --inref $SUBJECT/mri/norm.mgz \ --outref $FREESURFER_HOME/subjects/cvs_avg35_inMNI152/mri/norm.mgz \ --reg $SUBJECT/DSI/${SUBJECT}_diff2anat.mat \ --regnl $SUBJECT/cvs_CVS35MNI152/final_CVSmorph_tocvs_avg35_inMNI152.m3z \
reg (.mat) was computed as follows: bbregister --mov dsi_b0.nii --frame 0 --bold --s $SUBJECT --init-fsl --reg DSIregister.dat --fslmat ${SUBJECT}_anat2diff.mat
I would be happy to copy the data to a location of your choosing if you would like to replicate the error. Thanks!
Tommi
--- Tommi Raij, MD, PhD MGH/MIT/HMS Athinoula A. Martinos Center for Biomedical Imaging Bldg 149, 13th St Charlestown, MA 02129 U.S.A.