I'm not familiar with the MRC CBU fieldmap undistortion tool, so I can't
really comment. In general though, it is much safer to use the same flip all
the time. If it is changed around, then the results can be unpredictable.
Whatever you do, I would suggest keeping a close eye on your image
orientations.
I believe that at the CBU, your data is stored right-handed. If you have a
one-off dataset that is stored left-handed, then generally the safest thing
to do is to flip the dataset via the Display routine before doing anything
else.
Best regards,
-John
> I'm sorry if I missed one of the previous mails that might have covered
> some of this.
>
> I've just read the issues regarding SPM2 and orientation somewhat
> belatedly .. I have acquired a data set in radiological convention. The
> fieldmap undistortion tool (adapted from the MRC CBU one for use with
> siemens data) is normally used post reorientation and prenormalisation and
> states that it assumes radiological convention. Given this - do I set
> defaults.analyze.flip=1. right from the beginning? Or will this mess things
> up for the undistortion step? If so, should I use SPM99 normalisation and
> do the old style flip there? In the case of this particular dataset where
> I've processed through to the undistortion step without altering the
> defaults.analyze away from the CBU default of 0, am I best to go back to
> square one or can I alter anything at this point?
> many thanks,
|