SPM really shouldn't create any missing data. This would suggest to me that
something (I've no idea what exactly) has gone wrong. My best possible
explanation is that your images contained a few empty slices. This may cause
the segmentation of SPM to give these zero intensity voxels a small
probability of being cerebellum. Alternatively, the spatial normalisation
may have tried to stretch the brain to create something resembling
cerebellum.
Best regards,
-John
> I have structural MRI images where the most inferior part of the cerebellum
> are missing because the FOV during scanning was to small in -z direction.
> After normalisation and segmentation the missed part of the cerebellum were
> created by SPM. My question is:
> Whether the normalisation and segmentation procedure of the neocortex is
> influenced
> by the missing part of the cerebellum or by the voxels that were introduced
> by SPM
> so that the results of the VBM in the neocortex have to be interpreted with
> caution.
|