I tried registering DTI data to T1 of the same subject, but some structures
(such as the hippocampal stem) are consistently off. I had given up on that.
Is anyone having good results registering EPI to T1 images?
Stefano Marenco, MD
GCAP, Clinical Brain Disorders Branch,
NIMH
10 Center Drive, room 4S235
Bethesda, MD 20892
tel. (301) 435-8964
fax. (301) 480-7795
email: [log in to unmask]
-----Original Message-----
From: Stephen Smith [mailto:[log in to unmask]]
Sent: Tuesday, January 11, 2005 5:35 AM
To: [log in to unmask]
Subject: Re: [FSL] normalizing DTI data
Thanks Kate, that makes sense.
Can I add a couple of thoughts. One minor one is that I presume Russ meant
FLIRT not MCFLIRT? Next, some of the same points apply here as for FMRI.
In particular, I agree with Kate - we would generally recommend also
taking a T1, and getting the best intra-subject registration from the DTI
data to that, and then using the T1 to register to standard space. For the
former, PRELUDE&FUGUE may help, as well as Kate's suggestions, though
newer scanners, at least at 1.5T, don't necessarily generate a lot of EPI
distortion. In my experience, however, there are currently no really
excellent solutions to the [old VBM chestnut] problem that structural
differences across subjects confound voxelwise stats on (eg) "aligned" FA
maps.
Cheers, Steve.
On Tue, 11 Jan 2005, Kate Watkins wrote:
> Russ
> I too was unhappy with the lack of overlap in anatomy resulting from
affine
> transformation - and yes it is a limitation of the affine transformation.
I
> also tried the spatial normalization in SPM2 using all the defaults and
was
> still unhappy. It seems you need a high-dimensional nonlinear
registration
> so I used MRITOTAL part of the MNI_ANIMAL package. This produced
sufficient
> overlap to run a group voxel-wise analysis, which I think is what you
want.
> Also, you might find you need an extra step in your processing stream and
> that is to work out an affine transform (probably only need 7 dof) between
> your b=0 image and a high-res structural in the same subject and
concatenate
> that with a nonlinear transformation between the high-res structural and
the
> template image.
> Hope that helps.
> Kate
>
>
>
--
Stephen M. Smith DPhil
Associate Director, FMRIB and Analysis Research Coordinator
Oxford University Centre for Functional MRI of the Brain
John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK
+44 (0) 1865 222726 (fax 222717)
[log in to unmask] http://www.fmrib.ox.ac.uk/~steve
|