Dear SPMers,
i use SPM99 routine for analysing FDG-PET data from patients with alzheimer
dementia. I’d like to find brain regions, whose CMRglc positively correlate
with MMSE score independently of global CMRglc reduction and age.
Proportional scaling or ANCOVA would be more appropriate to use to account
for global CMRglc changes?
Would be 1 0 -1 -1 (0 cond + 1 cov (MMSE) + 2 nuisance (age,globalCMRglc) a
proper contrast in case of applying ANCOVA?
Thank you in advance.
With respect,
Igor
|