You are correct in assuming that S is a linearly spaced vector with the
spacing you described. However, I'm not sure of the reasons behind the
original choice (from SPM94 and SPM96) of using 8 slices (rather than
fewer or more) amd the 8mm from the top and bottom. Of course we could
have made these values modifiable by the user, but part of the
philosophy behind SPM is to keep things relatively simple, and eliminate
a lot of the choices that the users can make. Some value needed to be
used, so I guess that some pragmatic decision was made based on a little
empirical evidence.
I spent very little time modifying the realignment code in SPM96, so
the differences between the versions was minimal. However, I can't
remember how the "bb(1,3) + 6/V1(6)" bit crept in to the SPM96 version.
I'm pretty sure that it wasn't deliberate - but it should make very
little difference.
Regards,
-John
> I have a question about line 367 of spm_realign.m.
>
> 363 % define height of transverse slices (S) used in
> 364 % subsampling the volume
> 365 %-------------------------------------------------------------------
> 366 % transverse
> 367 S = linspace((bb(1,3) + 6/V1(6)),(bb(2,3) - 8/V1(6)),8);
>
> My understanding is that S is a linearly spaced vector with elements
> assigning voxels in z direction starting from a voxel corresponding to
> 6mm above zmin and ending in a voxel correspondig to 8mm below zmax with
> 8 slices subsampled. Is it right? And if so, what is the reason for
> choosing these particular numbers (6mm, 8mm, 8slices)?
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|