 |
 |
Dear Chiara, Sounds like an interesting project! There’s no simple rule on this, you just need to make sure your sphere contains some effect for the DCM to explain (e.g. as evidenced by a GLM analysis). If it’s a task-based experiment, the main source of variance in your sphere should be related to your task. Regarding your specific question, you can of course choose a bigger VOI than your smoothing kernel. Be guided either by the activations you get in your GLM analysis – if you do a second level analysis, you can see how far your effects extend spatially. And be guided by anatomy – if you want to make inferences about STS and IFG, you will need to try to exclude neighbouring gyri and sulci, although in practice you won’t fully be able to do this given the spatial smoothness. Best, Peter Peter Zeidman, PhD Methods Group Wellcome Trust Centre for Neuroimaging 12 Queen Square London WC1N 3BG [log in to unmask]<mailto:[log in to unmask]> From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of [log in to unmask] Sent: 04 December 2015 17:49 To: [log in to unmask] Subject: [SPM] how to define the VOI size for DCM analysis Dear SPM users, I am trying to perform DCM analysis on fMRI infants data. Images are quite noisy (even if the babies were sleeping during the acquisition), and I set the smooth in the preprocessing as 7mm. For the model I'd like to test, I just need to extract 2 VOI (STS and IFG) and I was wondering which is the size that I have to select about the VOI sphere. I've read in a DCM blog that there's supposed to be a rational between the smoothing applied in the preproccesing and the size of the VOI, or that at least I can't pick a bigger VOI in size that the smoothing (thus in my case, it would be 7 mm -- or less??) I could not find a paper or lessons slides on line that could exact define a rule for it, and there's nothing about it also in the SPM manual. All your suggestions are more than welcome!! Chiara ---------------- Chiara Bulgarelli PhD Student, Centre for Brain and Cognitive Development Birkbeck College, University of London [log in to unmask]<mailto:[log in to unmask]> // [log in to unmask]<mailto:[log in to unmask]>