Tracts and ROIs are not aligned in MNI space

[Tong Wu]

Hi Frank,

Thank you for addressing the issue that I have asked previously.

However, I have run into an issue while checking the tracts saved in MNI space. I tried to verify the coordinates by loading several tract bundles to see if they correctly pass through specific ROIs. I also looked at the ROIs in DSI Studio to visually confirm the alignment. It seems that the tracts and ROIs are not aligning as expected, neither visually nor mathematically. Could you please take another look at this?

Thank you so much!

Best,

Tong

[Tong Wu]

Hi Frank, 

Like to follow up on this. Also, after reviewing the tracts again yesterday, I noticed that post-DSI Studio upgrade, the alignment has been completely off. Previously, saving and loading the tracts in MNI space showed proper alignment, but now they seem to persist in the native space. Would you like to fix this? Thank you!

Best,

Tong

[Frank Yeh]

I am sorry for the issue.

Could you provide a step-by-step instruction to reproduce the problem? If I can reproduce on my side, I can fix it right away.

Best regards,

Frank

--
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/afa47430-745e-492f-b907-0e7140759665n%40googlegroups.com.

[Tong Wu]

Thank you for getting back to me. Here's the procedure I follow to inspect the tracts of various subjects in MNI space:

I start by opening the subject's .src file to generate tracts in the subject's native space.
Next, I right-click on the generated tract and select "save tracts" followed by "save current tract in MNI coordinates."
Then, I open the .src file of a template brain already in MNI space and attempt to load the previously saved .tt files by navigating to "tracts" and selecting "open MNI-space tracts."
This process worked before the DSI Studio upgrade. Which I can load the tracts in MNI space with perfect alignments. However, after installing the latest version, it no longer functions as expected. These tracts are totally off the MNI space.

[Frank Yeh]

Save tracts using [save Current Tract in Template-Space ] and see if it works.

The [load MNI-space tracts] should be renamed as [load Template-Space Tracts] (I am correcting it now).

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/30a90c9c-2b49-485d-9b92-46db583d37b6n%40googlegroups.com.

[Tong Wu]

Thank you for the guidance! Following your instructions, the tracts now align correctly with the template brain, without any discrepancies. However, I've encountered a new issue as I mentioned previously. 

To determine the coordinates, I've been loading the tracts in the template space and saving them as .mat files. I then try to identify the tracts' coordinates and proceed to load the atlas by opening the regions/template atlas. When I load the ROIs that supposedly share the same coordinates as the tracts, I expect the tracts to pass through these ROIs. Unfortunately, they do not align as anticipated. Could you shed some light on what might be causing this misalignment?

[Frank Yeh]

Could you provide the steps you did?

I am not sure how you loaded the ROIs.

Frank

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/762ebc8d-5545-45b5-bb71-aae891e75b24n%40googlegroups.com.

[Tong Wu]

Certainly! Here are the steps I've been following to pinpoint the issue with the coordinates:

I start by loading the tracts in the template space and saving them as .mat files to capture their coordinates.

Next, I open the .mat files in Matlab to examine the coordinates of the tracts. For instance, I select a bundle of tracts that intersects the point with coordinates (66,77,57).

Following this, I load the AAL626 atlas in Matlab to identify which Region of Interest (ROI) corresponds to these coordinates, finding that ROI should be numbered 85.

Finally, I switch back to DSI Studio, load a template brain .src file, and navigate through Regions > Open region > Load AAL626, and specifically select AAL_626_85. The atlas and the template brain aligns well.

Despite these steps, I find that the tracts and the ROI do not align as they theoretically should.

[Frank Yeh]

Next, I open the .mat files in Matlab to examine the coordinates of the tracts. For instance, I select a bundle of tracts that intersects the point with coordinates (66,77,57).

This coordinate is the template-space voxel coordinate.

 

Following this, I load the AAL626 atlas in Matlab to identify which Region of Interest (ROI) corresponds to these coordinates, finding that ROI should be numbered 85.

AAL626's template-space voxel coordinate may not be the same as the one used in DSI Studio, which is HCP1065 2mm 

The step here is likely having a problem.

 

Finally, I switch back to DSI Studio, load a template brain .src file, and navigate through Regions > Open region > Load AAL626, and specifically select AAL_626_85. The atlas and the template brain aligns well.

DSI Studio converts AAL's voxel coordinates to HCP1065's

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/a480d813-9920-4390-aa94-2943d4f8df86n%40googlegroups.com.

[Tong Wu]

Glad to know that! But how to know the transformation matrix that converts AAL's coordinates to HCP1065's?

[Tong Wu]

Hi Frank, 

Would you like to let me know how to get the transformation matrix that converts AAL's coordinates to HCP1065's in dsi-studio?

Thanks!

[Frank Yeh]

You may check out how NIFTI header define ijk and xyz and how srow matrix work.

It is a matrix-vector multiplication.

To view this discussion on the web visit https://groups.google.com/d/msgid/dsi-studio/295abc5e-fa60-4d72-8b11-40ab5f2eb3c6n%40googlegroups.com.

[Frank Yeh]

This video also tells about this topic: https://www.youtube.com/watch?v=IVZIQRUBepw