> 1. what is the actual purpose of the wrap function?
It is a different boundary condition (toroidal) for image interpolation. It
assumes that the image is repeated in a way similar to how a Fourier
transform would assume that the image would be repeated.
> does the wrap function improve normalisation?
When used properly, then it could. It is an option that is there for advanded
users to try if they so wish.
> is it recommended to use it in preprocessing for VBM?
I would suggest not using it here.
>
> 2. does the preserve total default has the same function as the
> Jacobian modulation?
Yes. It is the same as writing a spatially normalised image and "modulating"
it. It probably isn't much use for VBM in SPM2 though, as segmentation is
usually done on the spatially normalised images. If these are modulated,
then the segmentation will try to unmodulate them as part of the bias
correction process.
Best regards,
-John
|