Medical image segmentation system and method fusing MLP branch and CNN branch

The invention belongs to the technical field of medical image segmentation, and discloses a medical image segmentation system and method fusing an MLP branch and a CNN branch, and the method comprises the following steps: 1, extracting global features and local features of a medical image through employing the MLP branch and the CNN branch, and fusing the global features and the local features through employing a fusion branch; 2, calculating a final fusion feature of each hierarchical fusion module; step 3, decoding all the final fusion features by using a decoder, and then performing up-sampling to generate a segmentation image; the method can extract and fully fuse local features and global features, and has better medical image segmentation performance. 本发明属于医学图像分割技术领域，公开一种融合MLP分支和CNN分支的医学图像分割系统及方法，医学图像分割方法包括以下步骤：步骤1，使用MLP分支和CNN分支提取医学图像的全局特征和局部特征，使用融合分支融合全局特征和局部特征；步骤2，计算每个层级融合模块的最终融合特征；步骤3，使用解码器对所有最终融合特征进行解码，然后进行上采样生成分割图；本发明能够提取并充分融合局部特征和全局特征，具有更好的医学图像分割性能。