Moving ROI to nearest GmWm interface

Dear Mrtrix users and experts,

I am trying to filtering tractograms based on a ROI which is located in gray matter area. My question is if I need to move this ROI to GMWM interface or not and if yes how?

Thanks for your help in advance.

1 Like

Hi Ehssan,

Welcome to the MRtrix community! Regarding your question, can you elaborate further? The pipeline that you are currently using, the data description, and the ROI you are interested in.

If I have understood your question correctly, by using Anatomically-Constrained Tractography and defining gmwm boundary with 5tt2gmwm, you can do it.

Let me know if it was helpful or you require further info.


Hi AmirHussein,

Thank you for your kind response.
Let me provide additional details to clarify the situation.

In line with the BATMAN tutorial, I successfully generated a whole-brain tractogram for each participant in our study. I achieved this by employing Anatomically-Constrained Tractography and defining gmwm boundaries.

Furthermore, we have also conducted task fMRI scans on the same participants. From this data, we identified a functionally relevant Region of Interest. The ROI is a binary mask, defined as the peak level activation in the task fMRI group-level analysis, with a radius of 5 mm around its center.

Now, the next step in our analysis is to filter the entire brain’s tractogram based on this ROI. However, to ensure the accuracy of our results, I need to move this ROI to the nearest gray matter-white matter interface.

Could you kindly advise me on how to proceed with moving the ROI from its current location to the nearest GMWM interface?

Thank you once again for your support. I look forward to your guidance.

Best regards, Ehssan

Cool idea and approach! My naive idea is that you would just simply dilate the ROI mask a little, then multiply it with the GM-WM boundary. Another naive thought is to use recon-all of FreeSurfer, then use 5ttgen hsvs flag, transform the generated 5tt file to the dMRI space, do tractography with GM-WM boundary. If you have analyzed your fMRI data using FreeSurfer surfaces (fmriprep or FS-FAST pipelines), you can simply map that ROI on the white matter surface which is the GM-WM boundary using FreeSurfer commands. Hope these help.