Fig. 2From: Reinforcement learning using Deep \(Q\) networks and \(Q\) learning accurately localizes brain tumors on MRI with very small training setsTwo possible testing/deployment results. a shows a case of an accurate prediction, a true positive. After the 20 steps of forward inference on a presumed testing set image, the agent overlies the lesion. b shows a testing set miss, a false positive, where the agent does not overlap the lesion. In this particular case, there is no way for the lesion to get back to the lesion, since only the three actions of stay in place, move down and move to the right are defined in our formulation, although a more general formulation with 5 directions is possible in future workBack to article page