Contour-aware network with class-wise convolutions for 3D abdominal multi-organ segmentation

Hongjian Gao, Mengyao Lyu, Xinyue Zhao, Fan Yang, Xiangzhi Bai

Abstract: Highlights•Proposed a novel contour-aware network with class-wise operations.•Increased the certainty along delicate boundaries with a contour localization task.•Designed MuSeCWA module and CWASPP module to deal with class variability.•Built a dataset including 24528 CT slices with manual labels of 14 abdominal organs.