Dear Michael,
> SPMers,
> I have figured out the re-alingment issue from my previous e-mail. The problem I am encountering is the following: We use a variable skip when acquiring our images. The "normal" voxel dimensions of our axial slices are 3.125x3.125x7. The problem arises when the "skip" is added into the equation. It seems that the value of the skip (usually between .5 and 2mm) is added to the Z dimension in the header file of our scans. SPM pre-processes all of the data without a problem but when we go to set-up our model SPM gives an error message b/c rarely are the voxel dimensions (the Z dimension) the same on any set of scans (bc/ the skip is usually different for each person). Does anyone out there use a variable skip and SPM? Can it be done or should I give up?
from the perspective of realignment/reslicing the relevant measure is inter-slice distance, which is slice-thickness + slice-gap (skip). The reason for the gap typically is to account for non-ideal slice profiles and slight movements, both of which would otherwise introduce spin-history (T1) effects.
In principle this is no obstacle to perform statistical calculations in SPM. Once the spatial normalisation has been done they will all be in the same format, which is what SPM wants.
However, it generally seems like a pretty bad idea to have different slice-gaps in different subject who are all part of the same project. In the future it would probably be good to standardise this parameter.
Good luck Jesper
|