If the EPI and anatomical images are registered OK, and the spatially
normalised anatomical is well aligned with the MNI data, then I can't figure
out why the spatially normalised EPI should not be in alignment with the MNI
data. Is it possible that the anatomical image was moved (reoriented in any
way by the matrix in its header) between segmentation and writing the
spatially normalised EPI? That is all that I can think of.
> Following SPM5 normalization of the functional data sets after
> segmentation, we are getting a distortion on functional data in only
> two subject from our study. We have a full brain EPI acquisition, not
> partial. We wonder which may be the reason and if going back to EPI
> normalization after anatomical one (e.g. SPM2, without segmentation) may
> solve this problem or, are we introduccing any change because of the
> difference in the normalization procedure between subjects?
> We have checked corregistration between both T1 and EPI images and it is
> OK. Thanks in advance.