Global context and boundary structure-guided network for cross-modal organ segmentation
作者：
Highlights：
• We firstly propose to utilize global context to guide deformable convolution, which can obtain reasonable receptive fields with a global perspective.
• We introduce the class-wise global context to handle intensity non-uniformities in cross-modal organ segmentation.
• A novel loss which focuses on the areas near the boundary is proposed here and can deal with the border blurs well.
摘要
•We firstly propose to utilize global context to guide deformable convolution, which can obtain reasonable receptive fields with a global perspective.•We introduce the class-wise global context to handle intensity non-uniformities in cross-modal organ segmentation.•A novel loss which focuses on the areas near the boundary is proposed here and can deal with the border blurs well.
论文关键词：Cross-modal,Organ segmentation,Global context,Boundary structure,Loss function
论文评审过程：Received 15 January 2020, Revised 20 March 2020, Accepted 20 March 2020, Available online 4 April 2020, Version of Record 4 April 2020.
论文官网地址：https://doi.org/10.1016/j.ipm.2020.102252
