MC-DC: An MLP-CNN Based Dual-path Complementary Network for Medical Image Segmentation

Xiaoben Jiang, Yu Zhu, Yatong Liu, Nan Wang, Lei Yi

Abstract: Highlights•A novel MLP-CNN based dual-path complementary network (MC-DC) for medical image segmentation.•The CNN encoder can focus on local spatial contextual information. The MLP encoder is employed to capture the long-range dependency, without using a complex self-attention mechanism.•A dual-path complementary (DPC) module is designed to effectively fuse multi-level features from MLP and CNN.•The cross-scale global feature fusion (CS-GF) module aims to rebuild global semantic information with the help of cross-scale attention, while the proposed cross-scale local feature fusion (CS-LF) module pays attention to reconstructing local spatial contextual information.•The comprehensive experiments on three typical medical image segmentation tasks demonstrate the effectiveness of the proposed MC-DC network for improving medical image segmentation.