Dear SPMer,
I have scanned 16 subjects, belonging to 2 different groups (9 and 7
subjects, respectively).
During an activation task the first ("normal") group activates mainly in
one hemisphere, the second ("atypical") group in the other hemisphere.
I want to adress the following question:
"Do the subjects from the second group activate homologous areas, i.e.
is their activation pattern mirror-reversed?"
What is the best approach to answer the question?
I could do the following:
1. Normalize all subjects, the first group with neurological convention,
the other with the radiological convention
Now, all subjects activate on the same hemisphere. (Is it a alternative
to use "resize x -1" under the DISPLAY option? )
2. Do a fixed effect analysis (my group size is too small for a random
effect approach) for both groups.
3a. Present the activations of both groups e.g. as rendering on the
3D-brain. Now I can see, if the blobs are in coverage or not.
3b. Alternatively I could use masking: inclusive masking for the areas
that are activated in both groups, exclusive masking to get the areas
that are common to both groups.
Is there a better approach than a fixed effect analysis? What about a
conjunction analysis?
Comments are appreciated,
Andreas
|