Hi,

thank you Saad for your quick response and clarification.

I am actually calling probtrackx with --target3, and I am running into another issue: when the seed and target are single ROIs, everything runs fine.  However, when I supply an ASCII file containing a list of ROIS (built from freesurfer segmentations and transformed to diffusion space), I get:

CSV::load_rois:Unknown file type: CON_1_aparc_aseg_seed_1000.nii.gz

Note that the same .nii.gz file works fine as a seed and/or target when supplied directly (i.e. not listed in a separate ASCII file) via the -x and -target3 options (see below for full output with error message and probtrackx call).

Thanks again,

Nicola
_____________________________________________

*call to probtrackx*

/usr/local/fsl/bin/probtrackx2  \
-x $SUBJECTS_DIR/${seeddir}/${sub}_aparc.a2009s+aseg_seeds.txt \
-V 1 \
-l \
-f \
--onewaycondition \
-c 0.2 -S 2000 \
--steplength=0.5 \
-P ${p} \
--fibthresh=0.01 \
--distthresh=0.0 \
--sampvox=0.0 \
--forcedir \
--opd \
-s $SUBJECTS_DIR/${subbpx}/merged \
-m $SUBJECTS_DIR/${subbpx}/nodif_brain_mask \
--dir=$SUBJECTS_DIR/${trackdir}/${p} \
--omatrix3 \
--target3=$SUBJECTS_DIR/${seeddir}/${sub}_aparc.a2009s+aseg_seeds.txt

*output*

verbose    1
nparticles 2
nsteps     2000
usef       1
rseed      12345
randfib    0
fibst      1
Running in seedmask mode
load seeds
done.
Load bedpostx samples
1_1
1_2
1_3
2_1
2_2
2_3

nfibres  : 2
nsamples : 50

Done loading samples.
Initialise pathdist
....done
CSV::load_rois:Unknown file type: CON_1_aparc_aseg_seed_1000.nii.gz

On 3/11/2013 10:10 PM, Saad Jbabdi wrote:

[log in to unmask]" type="cite">Hi

Hello list,

a few quick questions about the format of coords_for_fdt_matrix* and fdt_matrix3.dot:

1) The file looks something like

57  69  27  0  1
57  70  27  0  2....

Are these the X/Y/Z coords of voxel number n, while n is in the 5th column? Is the 4th column always 0?

first 3 numbers are coordinates in SEED space. 4th is your ROI id (if you only have one roi then it's always 0 - this is useful when your roi, in this case target3, is composed of several combinations of surface and volume rois). the 5th column is the index of the different locations, i.e. vertices or voxels, within each roi).

2) is n the same number/reference which is in the 1st and 2nd column of "fdt_matrix3.dot" (i.e. a unique voxel identifier)?

yes

3) What are the origin and space of this coordinate system? Is it diffusion space (same space as fdt_paths.nii.gz), and is the origin (0,0,0) or (1,1,1)?

see (1). if your roi is a volume roi, then the coordinates are the same are those shown in fslview (voxels).

Cheers

Saad

Thanks a lot,

Nicola

--

Saad Jbabdi

University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK

(+44)1865-222466  (fax 717)

www.ndcn.ox.ac.uk/team/researchers/saad-jbabdi