Sounds like a coregistration problem. It is not horribly uncommon to
have coregistration be a little bit off if there is a lot of non-brain
tissue in the T1 scan. Checking to see that it worked properly for each
subject would be a good place to start.
-Neil
Jens Foell wrote:
> Hello everybody,
>
> I have a problem using SPM 8: In some, but not all, subjects of a study with
> block design, the activity shown in the contrast is tilted backwards (so
> that activation that should be parietal is now located in the cerebellum
> etc.) or forwards. This problem did not occur when the same data set was
> analyzed with SPM 2. In SPM 8, we used a batch to do the analysis. For
> pre-processing, we did a re-alignment, slice timing, co-registration,
> segmentation, normalization and smoothing.
>
> Can anyone help me with this?
>
>
> Thank you very much in advance,
>
> Jens
>
|