In an fMRI project I'm working on I ran multiple psychophysiological interactions from 8 different seed regions. Since I'm interested in showing the presence of a functional connectivity based network I was wondering how "valid" it is if I were to add the data originating from different seed regions into the same second level ANOVA.
Would it be incorrect to treat the different seed regions as different conditions?
You treat the seed regions as different seed regions, not as different experimental conditions.
Anyway, it is not clear what you want to do in the ANOVA. Using PPI, you have a connectivity map for each seed and for each subject. Do you want to compare each map (each voxel)? A second possibility is that you may want to compare connectivity from each seed with specific regions of interest; is that the case? A third possibility is that you may want to compare the topography of maps from each seed at the group level, after applying a threshold.
Depending on which is your aim, the answer to your question may slightly vary. But I'd say that generally it is not incorrect to compare functional connectivity from different seeds. It just depends on how you do that, and which is the purpose.
Simone
My intend was to show the combined network of which the connectivity differs between the conditions (in either direction) using a main effect F-contrast (identity matrix). This map would include the connectivity maps of all seeds and subjects, especially highlighting regions which show up on multiple of those maps.
I don't really understand the meaning of combined networks. You may want to say patterns of brain regions instead?
In any case, my opinion is in the answer above.
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling
- Analyses of Variance
- ANOVA