Print

Print


Hi
In order to get the mni2dti warpfield, you can run invwarp after putting together your two transforms using convertwarp. Something like:

convertwarp -o dti2mni_warp -r <MNI ref> -m dti2str.mat -w nonlinear_transf
invwarp -w dti2mni_warp -o mni2dti_warp -r nodif_brain


Note: probtrackx does not allow you to use a nonlinear warpfield yet, but the next release (4.1.5 - coming very soon) will.
With the current release (4.1.4) you can only use a FLIRT transform matrix.

Cheers,
Saad.


On 30 Nov 2009, at 09:14, cheng andy wrote:

Hi, all
 
I want to do probtrackx from a seed region in MNI space, but I wondered how to get the inverse transformation from MNI space to native diffusion space.
From diffusion to MNI space, I implemented following steps:
 
flirt -ref b3D -in nodif_brain -dof 7 -omat dti2str.mat
flirt -ref ${FSLDIR}/data/standard/MNI152_T1_2mm_brain -in b3D -omat affine_transf.mat
fnirt --in=b3D --aff=affine_transf.mat --cout=nonlinear_transf --config=T1_2_MNI152_2mm
applywarp --ref=${FSLDIR}/data/standard/MNI152_T1_2mm_brain --in=nodif_brain --warp=nonlinear_transf --premat=dti2str.mat --out=dti_MNI
 
PS: b3D is the betted T1 image, and nodif_brain is the betted b0 image.
 
However, when do probtrackx, it need to select a Seed to Diffusion Transform.
How to get a inverse transform of the above tranformations steps?
Thanks in advance!
 
Andy
 
 
 

--
Saad Jbabdi
University of Oxford, FMRIB Centre

JR Hospital, Headington, OX3 9DU, UK
+44 (0) 1865 222466  (fax 717)