Hi there,
> I am doing "Realign: Estimate & Reslice". I notice the realigned and
> resliced 3-D image sometimes has the outer most couple of slices missing
> (i.e. coded all as zeros) and/or has some partial 2-D images. Could someone
> tell me why? I notice when I use different interpolation options the
> extents of image missing are different. But I can't quite figure out a
> trend or reason yet.
Well this is explained by the use of a mask which corespond to the intersection
of all images and- this mandatory in order to have a full time series even at
the edges ... you should get different results using different interpolation
methods as the way you reconstruct the volumes differs and this is particularly
visible at the edges .. you can notice that this effect varies from subject to
subject due to their mvt.
Note that if you intend to normalize, there's no need to write your data at this
stage: estimate and write the mean image then normalize and write (with a
coresgister step in between if you want); in SPM2 it will write down .mats for
each image which will be use for the normalization, in SPM5, transformations
are stored in the header. Doing that you'll have only one interpolation :-)
Hope this helps
Cyril
------------------------------------------------------------------
Message sent via Psychology Dept. WEB-Mail Gateway.
http://www.psy.gla.ac.uk/
------------------------------------------------------------------
|