부산대학교 도서관

Relying on data-driven, medical image segmentation based on deep learning can accurately and quickly locate the lesion area, reduce all kinds of interference information and human subjective errors, provide reliable quantitative information for clinical diagnosis, and has become the main application technology for building a healthy China and promoting the construction of a smart medical service system. However, due to the diversity and complexity of medical images, the robustness and generalisation ability of most medical image segmentation algorithms have yet to be improved, and there is a certain gap with practical applications. Improving the performance of intelligent segmentation of clinical medical images is of extraordinary significance for the future construction of smart medical services. In this paper, we propose a medical image segmentation algorithm based on two-branch context-aware Transformer network based on deep learning technology, which aims to improve the accuracy and generalisation ability of medical image segmentation and contribute to the construction of intelligent medical service system by refining and enriching the target boundaries and regions layer by layer and calculating the attention of cross-dimensional interaction.