Thanks, that was really helpful. The ROI's are definitely being created based on the literature and not by first doing the whole-head analyses. One additional question that I had was what threshold should I be using for creating such a mask? I can see the advantage of using a lower threshold such as 5% so that I can account for inter-subject variations but also can think of using the 25% or the 50% threshold so that the tract that I am choosing is more representative. Any input on this would be much appreciated. Thanks again, -Ruchika.