Assuming you are defining your ROI on the MNI template, here is what I would do:
-use flirt then fnirt to warp each patient's T1 to MNI space (check the results, especially if your brains have lesions you might have trouble with fnirt)
-invert each warp field (invwarp command I think)
-apply each inverted warp field (applywarp command) to the MNI ROI, therefore giving you the "same" ROI in each patients T1
-register the T1 to the nodif (non-diffusion weighted scan), therefore giving you a T1-2-nodif.mat
-your seed in probtrackx would be this ROI, using the seed space is not diffusion space option, then you can also supply the seed-2-diffusion matrix (which is T1-2-nodif.mat)
your results from tractography will then be in each subjects T1 space, which is where you will probably want them. Hope this helps.
Dustin
________________________________________
From: FSL - FMRIB's Software Library [[log in to unmask]] On Behalf Of Jean-Sebastien Provost [[log in to unmask]]
Sent: Tuesday, April 24, 2012 7:29 PM
To: [log in to unmask]
Subject: [FSL] Diffusion space vs structural space problem
Dear FSL experts,
I am currently performing a DTI analysis on lesioned brains and I encountered a small problem. Here is what I did so far
I have a T1 acquisition and a DTI acquisition for four lesioned brains. What I want to do is to see if the major white matter pathways are intact or not. I want to check for the same ROI for all of the patients so I registered the T1 acquisitions with each other using the MNI152-1mm template. Then, I identified the uncinate tract and I created a small mask (9 voxels) as a ROI. But then, I wanted to do the tractography with each patient to see how the tract is damaged. So I registered my mask/ROI back into the different patient' space to be able to perform the tractography. However, it seems that I'm having an hard time doing so since the T1 was not acquired in the same space as the DTI acquisition.
Originally, I registered my mask into the diffusion space using the FMRIB58-FA-1mm and the standard2diff.mat created previously in the registration step following the bedpostx, and registered the DTI scan using the FMRIB58-FA-1mm but without specifying a .mat (because I'm a bit confused). I realized that this approach didn't make sense because if I launch the probtractx step, it is asking me for the bedpostx folder which is not registered to any of these spaces. Or is it ? I am a bit confused on how I should proceed from this point ?
How can I put my mask in the same space as my DTI scan so I could perform the probtractx step smoothly ?
Thank you for your kindness and your expertise,
JS
|