Dear Experts,
I am a Chinese graduate student major in psychology/cognitive neuroscience, who adopt SPM5/8 as the major tool for neuroimaging data analysis. Recently we conducted a simple block-design fMRI study. Four EPI runs were used, in which there was only one condition (namely one types of stimuli). There were two types of stimuli in total and each of them appeared third times in two runs respectively. For example, procedure for one subject was like:
Run1: rest(60s)+task(stimuli1:120s)+rest(60s)+task(stimuli1:120s)+rest(60s)+task(stimuli1:120s)+rest(60s)
Run2: rest(60s)+task(stimuli2:120s)+rest(60s)+task(stimuli2:120s)+rest(60s)+task(stimuli2:120s)+rest(60s)
Run3: rest(60s)+task(stimuli1:120s)+rest(60s)+task(stimuli1:120s)+rest(60s)+task(stimuli1:120s)+rest(60s)
Run4:rest(60s)+task(stimuli2:120s)+rest(60s)+task(stimuli2:120s)+rest(60s)+task(stimuli2:120s)+rest(60s)
Now I hope to calculate the psycho-physical interactive effect between region A and B in the two experimental conditions mentioned above. Given the experimental design, however, I could calulate only one of the condition in each run (e.g. con1-rest) and then compared two PPI-estimated contrasts with Paired T-test (e.g. PPI_(con1-rest)*regionA.hdr/img vs. PPI_(con2-rest)*regionA.hdr/img). So two key questions were listed as follows:
1. Was the method mentioned above a correct way to calculate the PPI_(con1-con2)*regionA ?
2. If not corrected, what is the right way? Also, could I concatenate all the four EPI runs together?
I am eagerly looking forward to your reply! Thanks a lot!
Best wishes,
Yang Hu M.S.candidate
MRI unit, Institute of Cognitive Neuroscience,
East China Normal University, Shanghai, China
|