I've got a functional mask dataset which is nonzero at all places where some (interaction contrast) t-stat is sufficiently high and are overlapping the amygdala. I want to extract a single number representing the average value of some beta value across this mask for all subjects. This will be repeated for many betas, some of which are part of the interaction contrast, and some of which are not.
Anyway, I think this can be accomplished with a bunch of 3dmerge commands (one per subject to average across the mask, then one to average across all of those outputs), but this may not be the simplest way of doing it (especially the latter command, which is painful).
Is there a clean way, or is the analysis I described unusual?
Thanks!
Aditya