Title: Semantic Feature Attention Network for Liver Tumor Segmentation in Large-scale CT database Show Chinese title

Title: 大规模CT数据库中肝脏肿瘤分割的语义特征注意力网络

Abstract: Liver tumor segmentation plays an important role in hepatocellular carcinoma diagnosis and surgical planning. In this paper, we propose a novel Semantic Feature Attention Network (SFAN) for liver tumor segmentation from Computed Tomography (CT) volumes, which exploits the impact of both low-level and high-level features. In the SFAN, a Semantic Attention Transmission (SAT) module is designed to select discriminative low-level localization details with the guidance of neighboring high-level semantic information. Furthermore, a Global Context Attention (GCA) module is proposed to effectively fuse the multi-level features with the guidance of global context. Our experiments are based on 2 challenging databases, the public Liver Tumor Segmentation (LiTS) Challenge database and a large-scale in-house clinical database with 912 CT volumes. Experimental results show that our proposed framework can not only achieve the state-of-the-art performance with the Dice per case on liver tumor segmentation in LiTS database, but also outperform some widely used segmentation algorithms in the large-scale clinical database. Show Chinese abstract

Abstract: 肝肿瘤分割在肝细胞癌诊断和手术规划中起着重要作用。 在本文中，我们提出了一种新颖的语义特征注意力网络（SFAN），用于从计算机断层扫描（CT）体积中进行肝肿瘤分割，该网络利用了低级和高级特征的影响。 在SFAN中，设计了一个语义注意力传输（SAT）模块，该模块在相邻高级语义信息的指导下选择具有区分性的低级定位细节。 此外，提出了一种全局上下文注意力（GCA）模块，在全局上下文的指导下有效地融合多级特征。 我们的实验基于两个具有挑战性的数据库，公共的肝肿瘤分割（LiTS）挑战数据库和一个包含912个CT体积的大规模内部临床数据库。 实验结果表明，我们提出的框架不仅可以实现在LiTS数据库中肝肿瘤分割的病例Dice分数的最先进性能，而且在大规模临床数据库中也优于一些广泛使用的分割算法。