Dear Rosalia,
What gray matter are you comparing? Outside of the lesion?
I have not worked with lesioned brains myself, so unfortunately I
cannot be much help with the process. I imagine that normalizing the
lesioned brains in DARTEL4 will be difficult because they will not
match in shape—I would look for examples of others that have used
DARTEL for this purpose, or try to describe the specific problem
directly. Using the lesion mask for segmentation is one thing (which
it sounds like has worked ok), but I'm not sure how that might be
incorporated during registration (i.e. normalization).
Also, instead of using DARTEL, you can have New Segment save warped
(i.e. modulated) versions of your images—do these look ok? (These are
produced by specifying "modulated", "unmodulated", or "modulated +
unmodulated" for the "Warped Tissue" option in New Segment). This
would also be acceptable if DARTEL is not working and may be the
easiest route to go.
Best regards,
Jonathan
--
Dr. Jonathan Peelle
Center for Cognitive Neuroscience and
Department of Neurology
University of Pennsylvania
3 West Gates
3400 Spruce Street
Philadelphia, PA 19104
USA
http://jonathanpeelle.net/
On Fri, Jul 27, 2012 at 10:30 AM, Rosalia Dacosta <[log in to unmask]> wrote:
> Dear Jonathan,
>
> Thank you a lot for your reply.
>
> What I am trying to do is search statistical differences between stroke
> group and control group and correlations of these differences with cognitive
> tests.
>
> I would like using DARTEL...but how?? If I try to sum (fslmaths) the zeroed
> mask with the modulated c1 image...it does not work. If I try to do it with
> only with the segmented native image (c1 and c2)...is ok...but then I can
> not make modulation for every subject independently...What should I do?
>
> Any suggestion would be very very appreciate.
>
> Best regards,
>
> Rosalia.
>
>
|