I'm not sure exactly what AFNI's 3x4 matrix represents, but you probably just
need to change the header information for SPM to be happy with the files.
Note that I would suggest only applying a rigid-body transform to the headers
(using spm_get_space), so the extra stuff (zooms and shears) should be
factored out from the affine transform. If you apply a full affine
transform, then any volumetric measurements will be a bit off and it is
likely that the registration procedures in SPM will be sub-optimal (because
it "knows", a priori, roughly how much to zoom a typical head to match it to
the MNI data).
On Friday 11 July 2008 20:13, Michael Yassa wrote:
> Dear SPMers,
> I would like to use John's reorient.m script to do change the orientation
> of images. I acquire brains in a tilted oblique orientation and would like
> to use AFNI to automatically get them approximately in talairach space.
> AFNI outputs the parameters it uses as a 3x4 matrix of translations,
> rotations and scales (the same params that can be specified using Reorient
> images in the GUI).
> I would like to use reorient.m to apply these parameters and create a new
> mat file (or adjust the old one) without interpolating and writing new
> files. I'm not sure which of the params in reorient.m would allow me to do
> this. If it's not possible, is there a command line version or function for
> reorient images that I can modify for this purpose? The idea is to create a
> mat file that can then be combined with DARTEL transformations so I would
> only have to interpolate once.
> Thanks so much.
> Mike Yassa