The amalgamated tool of two renowned tools deep learning and reinforcement learning is a powerful representation for deep neural networks that improves the basic reinforcement learning framework. With the use of deep learning’s representational power, it learns from the agent’s activities on how to increase the expected reward. Recent work has shown a lot of success of deep reinforcement learning in different domains such as video games, robotics, finance, medical and computer vision. In this project, various DRL models and methods for planning medical image analysis are discussed. This study covers the fundamentals of reinforcement learning. DRL algorithms can address the problems with limited and inconsistent annotated medical imaging data, which has been a major obstacle to the implementation of deep learning models in clinical settings. DRL algorithms support these models for the reward function, interactions between agents, and the environment. There has been an extensive amount of research being done in this area, and it has the potential to enhance the utilisation of deep learning in medical imaging.