At the FIL, we have found that SPM99 spatial normalisation works better when
it is based on the mean of the functional images of the subject - rather than
on the subjects structural image (after coregistration). The reasons for
this are:
1) It is not possible to get a good rigid registration of fMRI data
to the subjects' structurals. This is because of the distortion
of EPI data. This is a pre-requisite when applying spatial
normalisation parameters derived from structural images to the
functionals.
2) The objective is to spatially normalise grey matter. Structural
MR images contain a lot of other information that can confound the
spatial normalisation. As only about 1000 parameters are used,
these confounding effects can make the spatial normalisation less
precise.
After coregistration and the spatial normalisation parameters have been
estimated, then the same parameters can also be applied to the structural
images.
Note that if you are using EPI images (because of the distortions), then the
spatially normalised structurals won't correspond exactly with the spatially
normalised functional images.
If you don't coregister your structural and functional images, then SPM won't
be able to apply the same normalisation to both the structural and functional
datasets.
Best regards,
-John
On Thursday 28 June 2001 11:28, Karel Deblaere wrote:
> > Dear colleagues,
> I am performing a group study with 12 subjects. We are only interested
> in the group results. Do I have to perform a coreg for each single
> subject with its anatomy image or can I just do a spatial normalisation
> after realignment. I want to overlay my results on a mean normalised
> anatomy image. What I mean is, is spatial normalisation reliable enough
> on its own ?
>
> Thanks in advance,
>
> Karel
----------------------------------------
Content-Type: text/x-vcard; charset="us-ascii"; name="karel.deblaere.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Karel Deblaere
----------------------------------------
--
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]
|