Dear SPM community,
I'm having a hard time understanding all the different options regarding image intensity scaling/normalization in SPM (8, if that makes a difference. But it seems like they're identical since SPM96 at least). I'm working with PET images and a two sample t-test. What I'd like to do is to scale every single scan in each group by its global mean. I'm not sure I understand all the options that SPM offers me. I feel like there's a lot of redundancy, and the manual doesn't help in understanding which one does what.
Indeed, in the factorial design specification, there's:
-grand mean scaling (yes/no) in the options of the design section
-Global calculation (which I set to mean, therefore I expect that it performs a grand mean scaling)
-Global normalisation (which is set to grand mean scaling) which has a subsection
-normalization, which again "scales the images so that they all have the same global value". which again sounds like a grand mean scaling.
What does exactly all of these 4 parameters do? Do I have to set them all to yes? What are the differences?
BTW, I've tried comparing two large groups of scans in this way and the two beta images I get in output have a different global mean (11.2 vs 11.7. Small but definitely not just rounding.). Shouldn't they be numerically identical by definition? (assuming no covariates...)
Thank you very much,
Luca Presotto
Rispetta l’ambiente: non stampare questa mail se non è necessario.
Respect the environment: if it's not necessary, don't print this mail.
|