When you say "cut off", do you mean set to zero, or do you mean that the
spatially normalised images don't cover this region? If it is the latter,
then the problem should be fixed by extending the bounding box. If it is the
former, then this may indicate something wrong with the initial affine
registration. One possible reason for this would be intensity
non-uniformity. My first though was that you may be using "customised
priors", generated with a too small bounding box - but this does not appear
to be the case (as you still get the problem when using the default tissue
probability maps as priors).
Best regards,
-John
> I'm conducting a VBM study using the optimized protocol. During
> segmentation parts of the lower cerebellum get cut off in some subjects in
> the GM, WM, and CSF-partitions.
>
> Spatial normalisation works fine when using a slightly extended bounding
> box. Using my own template with my own priors instead of the default ones
> doesn't seem to make a difference in the cut-off issue.
>
> What can I do? I'd be happy about any hints!
|