The spatial normalisation warps actually map from normalised to un-normalised
space, so regions of enlargement in the mapping from normalised to
un-normalised means that a structure will shrink in that region during
spatial normalisation.
Relative volumes can be computed that discount the effects of the affine
component by messing around with the code for computing the Jacobians.
I have put a Deformations toolbox in:
ftp://ftp.fil.ion.ucl.ac.uk/spm/toolbox/Deformations/Deformations-010409.tar.gz
This allows you to compute an inverse deformation field by:
1) Writing out a warp field from a sn3d.mat file
2) Computing the inverse of this warp.
Best regards,
-John
> I am working on some questions regarding the effect of spatial
> normalization, and I am puzzled by some findings. I wanted to look at
> the effect of non-linear normalization by computing the Jacobian
> Determinants, which (I thought) only reflect the non-linear components
> and have values <1 for volume reduction and >1 for volume enlargement
> (Gaser, NeuroImage 1999).
>
> Now, since the JD can be used to reverse the complete normalization (see
> also Johns message on preserving the original pixel counts), it must
> also contain information on the linear scaling. Also, since this
> reversal of normalization is done by simply modulating the normalized
> image with the JD, does this mean the JD does not really show the
> transformation matrix (from un-normalized to normalized) but the
> "inverse" ? This would explain why (despite a positive linear scaling,
> i.e. an overall volume increase during normalization), the values are
> almost excusively <1.
>
> So, my question is, is there a way to compute a modified JD only
> reflecting the non-linear components ? And (if my "inverse" theory is
> correct), how do I compute a deformation field which leads from the
> un-normalized to the normalized image ?
>
> Any input is greatly appreciated - thanks in advance :)
> Marko
>
> --
> ===============================================
> Marko Wilke, MD
> Research Fellow, Imaging Research Center
> Children's Hospital Medical Center Cincinnati
> [log in to unmask]
--
Dr John Ashburner.
Wellcome Department of Cognitive Neurology.
12 Queen Square, London WC1N 3BG, UK.
tel: +44 (0)20 78337491 or +44 (0)20 78373611 x4381
fax: +44 (0)20 78131420
http://www.fil.ion.ucl.ac.uk/~john
mail: [log in to unmask]
|