Dear Alexandre,
You should make sure whether it is only a retrospective rigid-body motion correction or whether it is the prospective correction PACE. In PACE, motion has to be inferred by comparing previous volumes, see http://cbs.fas.harvard.edu/science/core-facilities/neuroimaging/information-investigators/MRphysicsfaq#PACE for details. In short, in case the subject moves rapidly this motion correction fails and it wouldn't be too surprising to see large motion remaining.
Best regards
Helmut
|