编码器-解码器架构的U型网络一经提出就迅速在医学图像领域广泛运用。但正如一中提到的基于CNN的编码器-解码器难以获取医学图像的全局信息处理复杂的医学图像分割任务。在近几年的工作中开始尝试将transform与U型网络相结合。TransUNet在U型网路的编码部分加入了transformer layerswin unet将12层swin transformer block设计成U型u-transfo

The U-shaped network architecture with encoder-decoder structure has been widely used in the field of medical image analysis since its inception. However, as mentioned earlier, CNN-based encoder-decoder models struggle to capture global information of medical images, making it challenging to handle complex medical image segmentation tasks. In recent years, there have been attempts to combine transformers with U-shaped networks. TransUNet incorporates transformer layers into the encoding part of the U-shaped network, Swin UNet designs a U-shaped network with 12 Swin Transformer blocks, and U-Transform adds MHSA and MHCA modules to the classic UNet architecture