Radiotherapy is a highly complex medical treatment that involves several steps. To ensure proper quality assurance, each step of the radiotherapy process is subjected to rigorous testing and verification. For each patient receiving radiotherapy, a treatment plan is prepared using treatment planning system (TPS), and the patient is then irradiated based on the plan. The purpose of implementing artificial intelligence (AI) in radiotherapy is to verify the quality assurance of dose calculations and treatment plans suggested by TPS, as well as to optimize the treatment plan. In the present review, we have undertaken a comprehensive search using pertinent key terms, such as \\\"artificial intelligence,\\\" \\\"dose prediction,\\\" and \\\"quality assurance,\\\" to provide an overview of the role of AI in radiotherapy. The search was conducted from 2020 to the present and has yielded a total of 169 relevant articles, which have been meticulously reviewed and analyzed. The insights gleaned from this review shed light on the current state of AI research in radiotherapy. Selection process identified articles that use CT images, RT structure, RT plan, dose prescription, and beam shaping features to predict 3D dose, DVH, and specific dose points in tissue. Various AI networks have been used to achieve the study\\\'s goals, with deep learning networks receiving more attention. These networks have more hidden layers, resulting in highly accurate outputs. Convolutional networks are more suitable for processing computed tomography images. Among these networks, U-Net, Res-Net, and Dense-Net are commonly used because of their ability to quickly process a large amount of data. The U-Net networks take input data and reduce its volume while retaining the basic features. They then deliver the output in the same size as the input to the user. This allows them to process large amounts of data quickly and with minimal errors.

Keywords