Transformation of DSI Studio output to FSL space

[Divya Joshi]

Hello,

I am comparing tractography results from probabilistic methods (using FSL) and deterministic methods (using DSI Studio).

I am running into an issue where the output from DSI Studio is not oriented in the same way as the FSL. I looked through the documentation and saw that I need to negate the x and y coordinates to flip the orientation to the one used by FSL. While this flips the coordinates to the proper orientation, there is still a translation issue where the tracts are shifted from the proper place they should be in FSL.

Another issue I have is that I do not know the transformation matrix to change the coordinates of the tracts in DSI studio from image space (voxels) to anatomical space (anatomical coordinates). In FSL, I would normally use the sform or qform matrices, but in this situation, the DSI studio does not output a transformation matrix. I tried to save the DSI studio ROI as a NIFTI file and see what the sform and qform matrices are. In this case, it only gives me the voxel dimensions as the sform/qform matrices but not the transformation matrix to go from voxel space to anatomical space. Screenshots are attached for reference.

Thank you!

[Frank Yeh]

At step T3, insert the b0 imaged used by FSL (which has sform) using
[Slices][Insert T1W/T2W]
Then, save Tracts using [Tracts][Save Tracts][Save Tracts in Slice
(T1W/T2W) space]
Hope this works for you.
Frank

On Tue, Oct 5, 2021 at 1:16 PM Divya Joshi

> --
> You received this message because you are subscribed to the Google Groups "DSI Studio" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to dsi-studio+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/dd99e2c7-928d-4756-ac40-04b34aa21a00n%40googlegroups.com.

[Divya Joshi]

Hi Frank,

Unfortunately that did not work. It is still in a different space. Can you please tell me what kind of transformation I can use to get the tracts to the space the original segmentation is in FSL?

Thanks,

Divya

[Frank Yeh]

DSI Studio saves tracts in voxel coordinate (i,j,k) and the origin is
located at the corner (right-anterior-bottom).
Would this help?
Frank


On Tue, Oct 5, 2021 at 3:59 PM Divya Joshi

> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/15af6748-46db-46f4-af93-3efb718abca0n%40googlegroups.com.

[Divya Joshi]

Hi Frank,

The orientation of my images are origin at right-posterior-bottom. My current method is to:

DSI Studio voxels --> negate x and y --> multiply by the sform matrix given by DSI studio (these are my voxel dimensions):

[1.25, 0, 0, 0;

0, 1.25, 0, 0;

0, 0, 6.25, 0;

0, 0, 0, 1]

This seems to give me the right orientation but it is shifted from where it should be. I am guess this has something to do with where the origin is in DSI Studio and where it is in FSL, but I am not sure how to fix this problem.

Thanks,

Divya

[Frank Yeh]

FSL uses an origin somewhere in the center. You may find it from the
sform matrix of the dwi NIFTI files. The sform will tell where the
origin is.
Frank

On Tue, Oct 5, 2021 at 4:49 PM Divya Joshi

> To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/8f7a12c2-023a-43f4-9442-c783fb578426n%40googlegroups.com.