Swin Transformer Assisted Prior Attention Network for Medical Image Segmentation

Transformer complements convolutional neural network (CNN) has achieved better performance than improved CNN-based methods. Specially, is utilized to be combined with U-shaped structure, skip-connections, encoder, and even them all together. However, the intermediate supervision based on coarse-to-fine strategy not been improve generalization of In this paper, we propose Swin-PANet, which applying a window-based self-attention mechanism by Swin in network, called prior attention network. A new enhanced block CCA also proposed aggregate features from skip-connections further refine details boundaries. Swin-PANet can address dilemma that traditional poor interpretability process calculation insert its predictions into for learning humanly interpretable controllable. Hence, assisted provides accurate automatic medical image segmentation. The experimental results evaluate effectiveness outperforms state-of-the-art methods some famous segmentation tasks including cell skin lesion