Are you matching a resliced version of your image to the template? If so,
then the algorithm will try to stretch the low intensity regions (that may be
set to zero in the resliced data) so that they match the air outside the head
in the template.
On Tuesday 15 April 2008 16:53, Randolph Andrews wrote:
> Hello experts,
> I'm currently limited to SPM2 (Matlab 6.5.0 ver 13.0.1) and I'm trying to
> finesse the spatial normalization parameters to correct the following
> issue. I'm normalizing to the MNI152 template and while the x- and y-
> outcomes look just fine in the central and lower slices, the resulting
> warped image is "stretched" in the z-direction (top of brain only).
> The best results have come from:
> (no weighting)
> (don't weight sources)
> (extra heavy regularization)
> (16 nonlinear iterations)
> I hope someone's encountered this before and can provide a solution (or at
> least point me in the right direction). I don't particularly want to skull-
> strip if there's a simpler fix.