The blobs are thresholded to account for multiple dependent
comparisons, based on random field theory. In order for this not to
be too conservative, the spatial smoothness of your data (estimated
FWHM) should be more than about 2 or 3 voxels. Your original data are
quite low resolution and have not been spatially smoothed, which will
give you more conservative results. The spatially normalised data
have smaller voxels relative to the estimated smoothness, so the
random field theory corrections work better.
Best regards,
-John
On 11 January 2012 13:39, marina stosic <[log in to unmask]> wrote:
>
> Dear SPM users,
>
>
>
> I have a question concerning the steps of preprocessing and how they affect
> subsequent analysis. When I make a contrast with the images that are
> only corrected for slice timing and realigned, there is a certain amount
> activation, but the thing is that this activation gets significantly
> increased when I make the same contrast with the images that have also been
> normalized. I am sending the contrast images attached. Can anyone tell me
> why this happens?
>
> Thank you in advance!
>
> Marina
>
|