Since no one else has answered, I thought I'd jump in here.
If you've used CheckReg to ensure your unnormalized structural is
coregistered to your unnormalized realigned functional images, and
CheckReg again to check that the normalized structural matches the T1
template (including the orbital frontal regions you're concerned
about)—which it sounds like you have—I can't imagine it's a
coregistration or normalization problem.
I suspect you are just seeing the effects of signal dropout, which is
common in orbital frontal cortex. In this case the voxels in question
may even be masked in your first level analyses (you can check by
using CheckReg on the first level mask.img and normalized structural),
and therefore in your second level mask; it's probably worth checking
the mask.img file in the group analysis directory.
If you have fieldmaps, you may be able to recover some by undistorting
the images, or you might consider trying a different sequence.
Perhaps some orbital frontal experts on the list can offer more
Cusack R, Brett M, Osswald K (2003) An evaluation of the use of
magnetic field maps to undistort echo-planar images. NeuroImage 18,
127-142. http://dx.doi.org/10.1006/nimg.2002.1281 (and
http://imaging.mrc-cbu.cam.ac.uk/imaging/FmReference, although that
might be a little dated)
Deichmann R, Gottfried JA, Hutton C, Turner R (2003) Optimized EPI for
fMRI studies of the orbitofrontal cortex. NeuroImage 19, 430-441.
Weiskopf N, Hutton C, Josephs O, Deichmann R (2006) Optimal EPI
parameters for reduction of susceptibility-induced BOLD sensitivity
losses: A whole-brain analysis at 3 T and 1.5 T. NeuroImage 33,
On Wed, May 27, 2009 at 4:06 PM, Corey McMillan
<[log in to unmask]> wrote:
> Dear list users,
> I recently preprocessed and analyzed data from several experiments using
> SPM5. The preprocessing results appear to have worked relatively well when
> inspecting the images with CheckReg and Display. However, by chance, I have
> noticed that if I threshold some sample contrasts at a very low level
> (p<0.999) we never observe activation in orbital/anterior frontal
> regions. It appears that the brains are simply shifted a few millimeters
> from the front. Why might this be the case? Is this likely a coregistration
> or normalization problem?
> Many thanks for your help,
> Corey McMillan
> University of Pennsylvania
> Department of Neurology &
> Institute for Research in Cognitive Science
> email: [log in to unmask]
> web: http://www.seas.upenn.edu/~mcmillac