Figure 3
From: Application of reinforcement learning for segmentation of transrectal ultrasound images

From left to right: original images, ground-truth image (manually segmented), and the results of the proposed approach for eight sample images.
From: Application of reinforcement learning for segmentation of transrectal ultrasound images
From left to right: original images, ground-truth image (manually segmented), and the results of the proposed approach for eight sample images.