Artificial Intelligence networks are structures that attempt to simulate the neural networks that comprise human intelligence and are increasingly finding applications in modern medicine, as in all technological fields. In medical fields, especially diagnostic imaging, there is a very high development potential for artificial intelligence. Like various medical departments, orthopedics and traumatology also benefit from segmentation tools. Segmentation of the musculoskeletal system is a crucial key to evaluating these structures and planning possible medical interventions. Various artifacts present in images make segmentation difficult. Segmentation, which means identifying and separating the desired region from an image, can be performed with great success by artificial intelligence networks. These networks make segmentation free from the influence of human constraints and provide higher reliability and speed because they are automated. In the current literature, there are numerous works demonstrating the success of these networks. The use of these networks requires a proper understanding of their working principles. In this review, the segmentation applications commonly used in orthopedics and traumatology will be discussed, and the general characteristics of the networks, their working principles, the stages of operation, and the evaluation of the results will be presented.