In medical image processing most of the published works try to reduce false positive rate (FPR) while in reality false negative is more dangerous than false positive? can anyone explain this for me?