Deep learning-based instance segmentation for the precise automated quantification of digital breast cancer immunohistochemistry images

作者：

Highlights：

• Feasibility in the quantification of Ki-67, ER, PR and HER2 biomarkers is proven.

• Models are effective regardless of variations in cell shape, color and texture.

• Via a web-based framework, researchers and pathologists collaborate seamlessly.

• The implemented framework is available online and used as a CAD tool by clinicians.

摘要

•Feasibility in the quantification of Ki-67, ER, PR and HER2 biomarkers is proven.•Models are effective regardless of variations in cell shape, color and texture.•Via a web-based framework, researchers and pathologists collaborate seamlessly.•The implemented framework is available online and used as a CAD tool by clinicians.

论文关键词：Breast cancer,IHC quantification,Instance segmentation,Deep learning,Biomarkers

论文评审过程：Received 2 June 2021, Revised 25 December 2021, Accepted 25 December 2021, Available online 14 January 2022, Version of Record 18 January 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116471