[MINC-users] different image origin

[Peter Neelin]

On 6/16/05, Oliver Winz <o.winz@fz-juelich.de> wrote:

> i have a problem with the tool "ecattominc".
> There are no problems at the conversion step but when i want to realign
> using register i can't see the overlay image because of the huge
> different in the  image origin.

Yes. If I recall correctly, the scanner puts the the origin around the
middle of the body, so the  head has a large offset.

If you are aligning volumes with register, then they do not need to be
overlaid in the third column. Simply select a number of equivalent
points on the PET and MRI images (make sure that they do not all lie
in or close to a plane). Once you have selected at least four, the MRI
and PET should align approximately and the merged images become
useful. For a good registration you should select about 15 points. If
you are using a tracer that does not distribute well throughout the
brain, then you may need to do the registration first with a
transmission scan.

Note that if you just translate the PET data so that both MRI and PET
appear overlaid, they will not be properly aligned. So you should not
worry about the large displacement of the PET - it just helps prevent
you from reading to much into the unregistered merged images. Do a
registration first (which does not require the merged images), and
then you can think about looking at the merged PET and MRI.

Peter
-- 
Peter Neelin
(peter.neelin@gmail.com)