Dear SPM experts,

I am analyzing H2 15O PET measured rCBF data. I wanted to look at differences in thalamocortical responses between two different groups. For this analysis I used the subtracted images (before intervention-after intervention)and built an ancova model with the two groups and entered data from the thalamic seed ROI as a covariate with a group interaction. For my (limited) understanding this model now compares differences in slopes between the two groups, when using  "1 -1/-1 1" as filter for the thalamic seed ROI covariates. The results looked pretty promising and I was able to confirm that with post-hoc comparison of extracted data outside of SPM.

Is this a valid model to use? This is probably a very old-school approach, but I didn't find a better way to do it with this kind of data as most people use fMRI paradigms with specific toolboxes for similar questions.

I would be happy for any feedback

Best

Chris