Hello,
ok, let us not put the finger on the question why the native data are not
available :-)
I think (other opinions welcome) you could rather address this
empirically, use the same EPI time series and process it with SPM2 and
SPM5 with the respective settings you had selected before, and
subtract/divide the results.
This will give you a feeling of how much incongruency is introduced by
this situation, given that the same EPI raw data were used for both data
pools. The null hypothesis is that there are 'some' differences.
(If possible, set some landmarks in the native image and check, how they
come to lie after the normalisation). Also just subtract the used
templates, e. g. by imcalc, to know if they were identical.
br,
Philipp
|