> - when I try to invoke spm_sn3d.m from the command line, like
> spm_sn3d('test.img','test_sn3d.mat','-78 78 -112 76 -50 85','2 2 2 8 8 8
> 10 8 4','T1.img','brainmask.img').
> I get all sorts of strange error messages very early on. Sometimes it
> tries to do an affine registration but aborts due to "Not enough overlap
> to obtain a solution", and the convergence is only about 30%. However,
> the images are in the correct orientation and if I do the normalization
> (same parameters) via the GUI, everything works just fine and the result
> is decent. Any ideas ?
I think this is probably because spm_defaults has not been run. When spm
starts up, it runs spm_defaults.m, which loads a bunch of default settings as
global variables. If you just call spm_sn3d from the command line, then you
may not have all these defaults loaded into memory.
>
> - also, when I normalize a large group of images via the GUI, it goes
> through the process all right and writes out the images. However, after
> it is finished with the last one, it starts over with "Writing subject
> #1", which takes exceptionally long and overwrites the "first version"
> of this image. Is it supposed to do this or am I missing something ?
It is supposed to do this. After it has estimated the spatial normalisation
parameters, it writes out a spatially normalised version of the image that
the parameters were estimated from. This image is used for quality control,
and it is displayed in the orthogonal viewer next to the template image.
best regards,
-John
--
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]
|