Once you have your images in the appropriate format, you can see them using
the <Display> button.  This will allow you to reorient one of them manually
so that it is roughly in line with the other.  You can check the results of
this using <Check Reg>.

When they are roughly aligned, then try coregistering them.  Providing you
tell the coregistration routine that both images are of the same modality
(you can lie if you want), then it will register the images by minimising the
sum of squares difference between them.

Images can be resampled in a different orientation via the coregister button,
then you can use <ImCalc> to create an average.

ImCalc allows basic algebraic calculations to be performed on selected images
and the result written as an image. The user must specify the expression to
be evaluated in standard matlab form with the selected images referred to in
order as i1, i2, i3 ?etc. Logical, relational and trigonometric operators can
be used as well as basic arithmetic operators. For information on using these
operators in matlab, type 'help arith', 'help relop', 'help slash' and 'help
elfun' at the matlab prompt. The user must enter a name for the resulting
output image. This will be written to the current working directory. (This
paragraph © Dick & Chloe, 2001).

Best regards,
-John

> I am interested in co-registering two volumetric CT
> image datasets with an angle of around 90 degrees
> between the scan axes of the two datasets.  The
> objective is to circumvent the limited spatial
> resolution associated with the (relatively) large
> slice spacing of 1 mm versus a FOV of 100 mm (512x512
> matrix).
>
> The image datasets are of a human pelvis.  We hope
> that by co-registering the two datasets and then
> resampling the registered images that we can obtain a
> higher resolution dataset for subsequent segmentation.
>
> From reading the documentation and manuals, I
> understand that SPM can coregister fMRI and PET
> images.  Will the algorithms in the program
> work for any type of imaging modality?  What about the
> large initial misalignment between these two datasets?
>
>
> If this is possible, what is the general process to
> arrive at this co-registration starting with original
> CT data? Finally, does SPM have any built-in
> capability to average and resample the resulting
> co-registered images?

--
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]