ROI-associated tracts

AFNI Message Board
1

Hi,

I would like to propagate tracts from an individual ROI to examine DTI metrics (FA, MD, etc.) on all those tracts that are associated with that particular ROI. I don’t care about the endpoints of these tracts. Can I do this with 3dTrackID? How should I do this? Is there something else I should use?

I’m not interested in tracts between pairs of ROIs.

Thanks,
Karin

2

Hi, Karin-

So, if I understand correctly, you want to make one white matter region of interest, which is: all the voxels hit by tracts that also pass through one special region?

For all tracking in AFNI, 3dTrackID is the function to use.

If you just want to use either deterministic or mini-probabilistic tracking, then you can use “3dTrackID -logic OR -mode {DET | MINIP} …” and the other normal inputs: mask “-mask …”, DTI parameters “-dti_in …”, ROI of interest with “-netrois”, etc.

If you want to do this with probabilistic tracking, which is generally a preferable way of going, then you could think of your question in an “AND-logic” sense: you have one ROI of interest, say called X, and you want to get regions hit by tracts between it and everything else in the WB mask, which could just be a region called Y. Therefore, you could make your “-netrois …” input be a whole brain mask with value 1 everywhere except for your special region X, which would have the value 2 in each voxel. That way, you have a 2-region network, and the off-diagonals of the connectivity-property matrices would give you stats on what you want.

… All of the above is predicated that I did, in fact, interpret your question correctly, so please let me know if I didn’t or if my rhapsodical waxing above is unclear at all.

–pt

3

Paul,

I’m not quite sure if you interpreted my question correctly. So I have a gray matter region X, and I would like to gather DTI metrics on the white matter tracts that connect with this region X.

Thank you,
Karin

4

Hi, Karin-

I’m not quite sure if you interpreted my question correctly. So I have a gray matter region X, and I would like to gather DTI metrics on the white matter tracts that connect with this region X.
OK. Does this mean that you have other regions already defined that you want to find connections to, or do you just want to find tracts that go from X to anywhere else in the brain?

–pt

5

I would like to find tracts that go from X to anywhere in the brain.

6

Hi-

OK, that actually, that is how I interpreted the question.

If map_ROI_X.nii.gz is a binary mask of your ROI X, and assuming your DT parameters like FA, MD, etc. are called DTI/DT_FA+orig, DTI/DT_MD+orig, etc., and your whole brain mask is mask_WB.nii.gz, then the following would give you a deterministic tracking result of what you want (and you could verify it by looking at the volumetric output “white matter ROI” map, or the tracts themselves in SUMA:


3dTrackID -mode DET                   \
    -dti_in DTI/DT                    \
    -mask mask_WB.nii.gz         \
    -netrois map_ROI_X.nii.gz            \
    -logic OR                         \
    -nifti                           \
    -dump_rois AFNI_MAP       \
    -prefix DTI/o.tracts_thru_X

–pt