 |
 |
Thanks, Paul! Yes, I thought about that. It would mean tracking outputs in individual diffusion space, and I was hoping for standard space output. But the more I think about it, the more I realize that there is no such thing as a whole-brain standard space corresponding to my "tbss space" anyway. Thanks! Anna On Wed, May 22, 2013 at 11:33 PM, Chou Paul <[log in to unmask]> wrote: > Dear Anna > > I think you could following the following manual of how to perform fiber > tractogaphy using TBSS based ROI > > > http://fsl.fmrib.ox.ac.uk/fsl/fslwiki/TBSS/UserGuide#Transforming_TBSS_results_back_to_native_space > > Best > > Paul > > > Date: Wed, 22 May 2013 22:57:24 +0100 > > From: [log in to unmask] > > Subject: [FSL] tract-based seeds for probtrackx2? > > To: [log in to unmask] > > > > > Hi all, > > > > I'm getting set up for running probtrackx2 on a number of subjects, and > I would like to define my seeds and targets in "tbss space", for lack of a > better word. What I mean by that is: I'd like to pick an ROI on the > mean_FA_skeleton, and when that gets projected into diffusion space, I'd > like it to be projected to the corresponding tract center of the individual > subject (i.e. where the values for the all_FA_skeletonised 4D dataset were > drawn from). > > > > From the way I understand tbss, this is not the same place in diffusion > space where I would end up if I simply registered my bedpostx diffusion > data with the standard FMRIB_FA brain. Even though I warped everybody's FA > to that brain in tbss_2_reg, I then used the -S option for tbss_3, which in > itself would make my study-specific mean_FA image differ from the FMRIB_FA, > and the tract-center search step in tbss_4 makes me end up in yet a > different place. Yet that latter place seems to be the one I want to track > from, because only in that case I can be sure that I track from each > individual's tract center. > > > > Is this possible? Theoretically, all necessary transformations should be > stored somewhere after the tbss, correct? And if it's not, what is the > closest I can get with the available options? > > > > Thanks so much! > > Anna >