Dear SPM experts

 
 I have been working on SPM to analyze SPECT images and find abnormalities by comparing SPECT images of the patients with the normal control group images.
 I am completely new in SPM.

For preprocessing I perform Coregistration, Normalization and Smoothing.

After that for model specification I don't know which design (sample t-test, ANOVA.  ) to choose? In two sample t-test, for example, there exists two mandatory selections “Group 1 scans”and“Group 2 scans”. Now my question is this: What (patient or control group) should I select for each above-mentioned group?

Then for the results section I should define contrast (t contrast or f contrast). Again, I don't know how to define it?
 
I would greatly appreciate if kindly help me in this case. 
 
 
Best.