Breast tumor segmentation and shape classification in mammograms using generative adversarial and convolutional neural network

作者：

Highlights：

• A conditional generative adversarial network (cGAN) is proposed to segment the breast tumor.

• A convolutional neural network (CNN) based shape classification descriptor is proposed.

• The segmented regions are classified into four different shapes and correlated with four molecular subtypes.

摘要

•A conditional generative adversarial network (cGAN) is proposed to segment the breast tumor.•A convolutional neural network (CNN) based shape classification descriptor is proposed.•The segmented regions are classified into four different shapes and correlated with four molecular subtypes.

论文关键词：Mammograms,Conditional generative adversarial network,Convolutional neural network,Tumor segmentation and shape classification

论文评审过程：Received 24 October 2018, Revised 4 July 2019, Accepted 29 July 2019, Available online 29 July 2019, Version of Record 3 August 2019.

论文官网地址：https://doi.org/10.1016/j.eswa.2019.112855