I want to compare 2 populations and I have only 1 condition but 3 PET scans
per subject (in this condition).
Should I dump all the scans into a "Compare Populations: 1scan/subject" and
pretend they're all different subjects or create a mean image for each
subject and compare them?
I've done both but I'm not sure which is the most accurate. Any help (and/or
explanation) will be appreciated.
Regrads,
Joel
BTW - the former is a lot more sensitive but I suspect specificity is poor.
|