Dear FSL users,
i am fsl newbie and i am trying to register a set of functional scans to the
avg152 brain. i am using the 3 step fMRI to epi (6 dof), epi to highres (6
dof), and highres to standard (12 dof) registration method in flirt within
feat. (trilinear, corr cost)
the fMRI to epi, and epi to highres registrations look fine. However, in
almost all of my registrations (60 sets) the fMRI to standard registrations
are incorrect. the outlines of the brains match, however, the fMRI scans are
shifted in the AP axis relative to the standard brain. please see the
following figures:
http://web.mit.edu/paymon/www/example_func2standard.jpg
http://web.mit.edu/paymon/www/highres2standard.jpg
http://web.mit.edu/paymon/www/example_func2highres.jpg
the highres to standard registration has the same problem and is probably
the source of the registration mismatch.
in the example_func2standard figure you can see that the third ventricles of
the fMRI scan are sitting bellow where they should be in a correct
registration (green arrows). this problem is consistent across almost all of
my data sets.
i would really appreciate any ideas on how i can solve this problem.
i am using fsl 3.3.7 for redhat9 on CentOS.
thanks a lot,
paymon hosseini
|