Hi - We have been using a relatively standard (so far as I can tell)
processing stream for our DTI data, aligning all of the
diffusion-weighted images to the b=0 image, then normalizing the b=0
image to the T2 MNI template and applying the transform to the DWI
images as well. I've done this on a fairly large group of subjects,
and it's pretty clear that while this approach does a decent job of
normalizing the brains overall, there is a large degree of variability
in the resulting FA maps. For example, in some cases there is no
overlap between the high-FA regions of the corpus callosum between two
subjects. This amount of variability seems to preclude any kind of
group analysis of the FA values, which is what we are most interested
in.
I've also tried to directly coregister the FA maps across subjects
(using mcflirt with -dof 12) which did not seem to appreciably decrease
the variability in location of the white matter tracts. Can anyone on
the list provide suggestions as to how to deal with this problem? It's
not clear to me whether it's a limitation of the affine transform, a
limitation of the cost function (I've tried both normcorr and mutual
information), or something else.
cheers,
russ
---
Russell A. Poldrack, Ph.d.
Assistant Professor
UCLA Department of Psychology
Franz Hall, Box 951563
Los Angeles, CA 90095-1563
phone: 310-794-1224
fax: 310-206-5895
email: [log in to unmask]
web: www.poldracklab.org
|