Hi Maxine -

While I would not profess to be an expert,  I can tell you that if I received this paper as a reviewer, my immediate reaction would be precisely your two concerns - first, that your two ROIs are nonindependent (even though they come from difference sources), and as such formal statistical comparisons would be seriously confounding.  I suspect the BOLD time series of these two regions is nearly identical, which you can test in CONN by viewing the time series data, or by comparing plots in another program.

Second, I believe your contrast is comparing one fraction of the frequency band, vs. the entire frequency band?  That would seem to be an  incorrect comparison, simply because of the variance associated with one, i.e., narrow, band vs. a much larger broad frequency band. Perhaps Alfonso or someone else can weigh in on that portion.

HTH,

Noah