> we would like to normalize intensity of each slice independently to avoid
> some acquisition artifacts on a fmri study. We tried to assign the
> correction factor for each plane 'i' to V.info(1,i) but when we did that
> spm_slice_vol seemed to fail: the first slice was repeated as times as
> slices we had with different intensities and the other slices disappeared.
>
> Is there any limitation to use spm_slice_vol when having different values
> of V.pinfo for each slice? Are there any alternative way to correct image
> intensity of each slice separately, apart from direct product of each voxel
> value by the correction factor?
If there are multiple entries in the V.pinfo, then you also need to include an
appropriate offset into the file for each slice in V.pinfo(3,:). The first
entry is typically zero, the second is the number of bytes in a slice, the
third is 2 times this number etc.
Best regards,
-John
|