Coregistration before everything should be OK, although if you are
registering an anatomical image to a load of functionals, then you may be
best off if you use the mean created at the realignment stage.
How much trouble there is depends on what is coregistered with what. It
should be less trouble (if any) if the structural is coregistered to match
the functional images than if it is done the other way around. This is
because one image is assumed to remain stationary and the other is assumed to
be moved.
If one realigns the functional images so that they all match each other, and
then coregisters one of these images to match the structural image, then I
would expect problems. The coregistered image is no longer in register with
the other functional images.
Best regards,
-John
On Tuesday 24 April 2001 15:56, WONG Ka Kwun Kelvin wrote:
> I'd like to ask if coregistration before everything (realignment,
> normalization, smooothing, parameters estimation etc.) is ok or not. As
> it seems to me that coregistration is the most trouble part of the whole
> process, and the way realignment does would probably just to change the
> mat (or the transform matrix) file (create if it doesn't exist).
>
> Would John please comment on this?
--
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]
|