Adversarial Deep Structural Networks for Mammographic Mass Segmentation

Abstract

Mass segmentation is an important task in mammogram analysis, providing effective morphological features and regions of interest (ROI) for mass detection and classification. Inspired by the success of using deep convolutional features for natural image analysis and conditional random fields (CRF) for structural learning, we propose an end-to-end network for mammographic mass segmentation. The network employs a fully convolutional network (FCN) to model potential function, followed by a CRF to perform structural learning. Because the mass distribution varies greatly with pixel position, the FCN is combined with position priori for the task. Due to the small size of mammogram datasets, we use adversarial training to control over-fitting. Four models with different convolutional kernels are further fused to improve the segmentation results. Experimental results on two public datasets, INbreast and DDSM-BCRP, show that our end-to-end network combined with adversarial training achieves the-state-of-the-art results.

First version on arXiv 2016, MICCAI 2017 Deep Learning in Medical Image Analysis (DLMIA) workshop

References in corpus (4)

• Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs
• Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials
• Distributional Smoothing with Virtual Adversarial Training
• Adversarial Training Methods for Semi-Supervised Text Classification

Cited by in corpus (11)

• Generative Adversarial Networks in Computer Vision: A Survey and Taxonomy
• Conditional Adversarial Network for Semantic Segmentation of Brain Tumor
• Deep Multi-instance Networks with Sparse Label Assignment for Whole Mammogram Classification
• Generative Adversarial Networks: A Survey Towards Private and Secure Applications
• DAWSON: A Domain Adaptive Few Shot Generation Framework
• Conditional Generative Adversarial and Convolutional Networks for X-ray Breast Mass Segmentation and Shape Classification
• DeepLung: 3D Deep Convolutional Nets for Automated Pulmonary Nodule Detection and Classification
• Deep Learning for Automated Medical Image Analysis
• Leak Event Identification in Water Systems Using High Order CRF
• Progressively Growing Generative Adversarial Networks for High Resolution Semantic Segmentation of Satellite Images
• A Multi-Scale CNN and Curriculum Learning Strategy for Mammogram Classification