Pathologic liver tumor detection using feature aligned multi-scale convolutional network

作者：

Highlights：

• A Feature Aligned Multi-Scale Convolutional Network (FA-MSCN) for liver tumor detection on whole slide images is proposed.

• Two parallel convolutional networks in FA-MSCN would extract high-resolution and low-resolution features, respectively.

• Experiments demonstrated that the proposed FA-MSCN improves the HCC detection performance compared to both SSCN and MSCN.

摘要

•A Feature Aligned Multi-Scale Convolutional Network (FA-MSCN) for liver tumor detection on whole slide images is proposed.•Two parallel convolutional networks in FA-MSCN would extract high-resolution and low-resolution features, respectively.•Experiments demonstrated that the proposed FA-MSCN improves the HCC detection performance compared to both SSCN and MSCN.

论文关键词：Convolutional neural network,Liver tumor detection,Multi-scale,Whole slide image,Hepatocellular carcinoma

论文评审过程：Received 14 January 2021, Revised 3 November 2021, Accepted 3 January 2022, Available online 10 January 2022, Version of Record 29 January 2022.

论文官网地址：https://doi.org/10.1016/j.artmed.2022.102244