Hi Rick,
Thanks again, I really appreciate your insights. I guess the question about 'how can 3dvolreg detect such small motion?' implied the part about it being reliable. Perhaps the original question should have been 'how can 3dvolreg
reliably detect small motions?' Regardless, I was assuming that vibrations from the scanner itself could cause these motions in our phantom, but now I see what you mean about shifts in voxel intensity from random noise, fluctuating field inhomogeneities, etc. as a potential cause of 'psuedo-motion' to be detected.
~michael