Dear Ville,
If you want to identify common areas, the easiest thing to do is inclusive
masking. Continue to model A and B separately as combining them could
increase the noise in your analysis and decrease sensitivity. Then use the
contrasts A>C and B>C and mask one with the other. For instance, if these
are contrasts 1 and 2, then
avwmaths stats/zstat1 -thr 3.1 -bin -mul stats/zstat2 -thr 3.1 -bin
masked_thresh_zstat2
will take all voxels with z-values of 3.1 or greater in zstat1 and use
those as a mask for finding all voxels in the mask at Z>3.1 in zstat2. The
results will be a binary file showing regions of common activation in the
two contrasts at whatever threshold you chose (3.1 in this case).
Good luck,
Joe
--------------------
Joseph T. Devlin, Ph. D.
FMRIB Centre, Dept. of Clinical Neurology
University of Oxford
John Radcliffe Hospital
Headley Way, Headington
Oxford OX3 9DU
Phone: 01865 222 738
Email: [log in to unmask]
|