 |
 |
Hiya - I made a bit of a change, so bedpost no longer has the option of taking a T1-weighted image. You now have to compute the registrations using the registrations tab in FDT. just enter the bedpost directory, and the structural image - this will create and run all the flirt calls that you need. It will make a subdirectory within the .bedpost called xfms which contains all the relevant transformations. You can then just enter these in the "transformation matrix" slot when you run probtrack. Probtrack will then always know where it is in both seed (in your case structural) and diffusion images. Hope this helps... T On Thu, 8 Dec 2005, Kristen Lindgren wrote: > Thanks for your input. How would I enter my T1 image when I run bedpost? > Does it just need to have a specific name and be in the directory where I > run bedpost? It appears that the only input for bedpost is the directory. > > > > Since I've already run bedpost without entering the T1 image, I'll have to > use FLIRT to generate the transforms. If I enter my T1 as the reference > image and the B0 image for the input image, do I just need to select a > secondary image to apply the transform to? Would I enter the T1 image in > that field? > > > > Thanks! > > > > Kristen > > > > _____ > > From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf > Of Heidi Johansen-Berg > Sent: Thursday, December 08, 2005 2:27 PM > To: [log in to unmask] > Subject: Re: [FSL] Using FLIRT to align DTI and T1 images > > > > Dear Kristen, > > > > Yes those steps sounds fine. If you entered your T1 image when you ran > bedpost then it will have made the transforms for you. If you didn't then > you can generate them by running registration from the FDT GUI. You can > check to see if the transform is there by looking in the xfms subdirectory > within your .bedpost directory. > > (note that it's sensible to check the quality of this registration step > carefully to make sure that your masks are ending up where you think they > should in dti space) > > > > Heidi > > > > On 8 Dec 2005, at 19:14, Kristen Lindgren wrote: > > > > > > Hello. I'm trying to align DTI data (put through bedpost) with a T1 with > the hopes of creating two masks on the T1 to use as seeds for probabilistic > tractography. However, I want to see if I'm performing the correct initial > steps and I'm also confused on how to go forward from there. > > > > 1) Run Bedpost on DTI data > > 2) Run BET on nodif (B0) image and T1 > > 3) FLIRT - Reference image is T1, Input image is nodif; I named the output > image nodif_aligned > > 4) Create masks while viewing T1 and trace ROIs > > 5) Use ProbTrack with the "Symmetric - Two Masks" option, enter two created > masks > > > > For Prob Track, should I select "Seed space is not diffusion" and enter a > transform (str2diff.mat???)? If so, will FLIRT make the necessary transform > or do I have to do this separately? > > > > Thanks for your help! > > > > Kristen > > > > > > > > Heidi Johansen-Berg > > FMRIB Centre > > John Radcliffe Hospital > > Headington, Oxford OX3 9DU, UK > > Tel: +44 (0) 1865 222782 > > Fax: +44 (0) 1865 222717 > > > > [log in to unmask] <mailto:[log in to unmask]> > > http://www.fmrib.ox.ac.uk/~heidi <http://www.fmrib.ox.ac.uk/~heidi> > > > > > > > > -- ------------------------------------------------------------------------------- Tim Behrens Centre for Functional MRI of the Brain The John Radcliffe Hospital Headley Way Oxford OX3 9DU Oxford University Work 01865 222782 Mobile 07980 884537 -------------------------------------------------------------------------------