It crashes because your images appear to be in different orientations
according to the matrices in the headers. There is a little flexibility
in deciding whether images are aligned (to allow for rounding errors),
and I figured that 1e-4 would be reasonable for this (it is greater than
eps(single(512)) - so should deal with the single-precision tolerance
for representing the elements of the affine transform matrix). How
similar are the matrices? Perhaps the images have been realigned - but
not resliced?
Best regards,
-John
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]]
On Behalf Of cyril pernet
Sent: Thursday, September 06, 2007 2:08 PM
To: [log in to unmask]
Subject: [SPM] error spm_orientation at 32
Hi there,
while loading a set of EPI images for a 1st level analysis I got a
'funny' error using ==> spm_check_orientations at 32
for sure my images are correct, using a loop across my images with
V.dim(1) == V.dim(i) I got my sum = all my images, ie all dimensions are
equals, but still spm_check_orientations gives me an error ; note they
are also all well oriented (spm_orientation ok).
the only remaining thing is reshape(cat(3,V.mat),[16,numel(V)]) is >
1e-4
and this is where spm_check_orientations crashes but I have no idea why
this limit ??
any help would be greatly appreciated
cyril
|