Hi Torsten,
I have zero experience with qMT images, but if I understand correctly, you do qMT-->T1-->b0. There are unfortunately some distortions between b0 and T1 that can't be accounted for by a 6dof flirt, so your qMT images processed this way (let's call them b0-qMT) won't be exactly in the same space than your b0... On top of this, there might also be some (other) distortions between your qMT and T1?
One way I can think of for you to check whether the TBSS processing is good enough for the qMT data is to back-project the entire skeleton to the "native" b0-qMT images, to see if you're happy with both registration and projection steps.
Cheers,
Gwenaelle
> De: Torsten Ruest <[log in to unmask]>
> Objet: [FSL] using qMT data with TBSS
> À: [log in to unmask]
> Date: Lundi 28 mars 2011, 14h53
> Hi there,
>
> I am processing our qMT data with TBSS, ie I first register
> the qMT maps and the bo image to the T1 weighted image using
> flirt (6 dof), then I invert the transformation matrix from
> b0-->T1 and apply it to the qMT data in T1 space.
>
> Our qMT data are quite low resolution in the z plane and
> fairly anisotropic (1.75 x 1.75 x 5 mm3) and maybe as a
> result the qMT in the MNI-FA space are not as neat as the FA
> maps with respect to the registration results and I was
> wondering with how much deviation between participants the
> TBSS software can deal with. I am happy to upload some
> images for you to look at, just let me know.
>
> Thanks for your help!
>
> Torsten
--------------------------------------------------------------------
Gwenaëlle Douaud, PhD
FMRIB Centre, University of Oxford
John Radcliffe Hospital, Headington OX3 9DU Oxford UK
Tel: +44 (0) 1865 222 523 Fax: +44 (0) 1865 222 717
www.fmrib.ox.ac.uk/~douaud
--------------------------------------------------------------------
|