# Adversarial Deep Structured Nets for Mass Segmentation from Mammograms

arXiv: https://arxiv.org/abs/1710.09288 · HTML: https://mlsanity.com/paper/1710.09288/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Oct 2017.

## Built on by (most cited first, 6 of 6)

- [Test-Time Training for Deformable Multi-Scale Image Registration](https://mlsanity.com/paper/2103.13578/) (2021)
- [Deep Learning in Breast Cancer Imaging: A Decade of Progress and Future Directions](https://mlsanity.com/paper/2304.06662/) (2023)
- [Breast Cancer Classification using Deep Learned Features Boosted with Handcrafted Features](https://mlsanity.com/paper/2206.12815/) (2022)
- [Artificial Intelligence For Breast Cancer Detection: Trends & Directions](https://mlsanity.com/paper/2110.00942/) (2021)
- [Deep Active Contours Using Locally Controlled Distance Vector Flow](https://mlsanity.com/paper/2105.08447/) (2021)
- [MLP-GAN for Brain Vessel Image Segmentation](https://mlsanity.com/paper/2207.08265/) (2022)

## Abstract (arXiv)

Mass segmentation provides effective morphological features which are important for mass diagnosis. In this work, we propose a novel end-to-end network for mammographic mass segmentation which employs a fully convolutional network (FCN) to model a potential function, followed by a CRF to perform structured learning. Because the mass distribution varies greatly with pixel position, the FCN is combined with a position priori. Further, we employ adversarial training to eliminate over-fitting due to the small sizes of mammogram datasets. Multi-scale FCN is employed to improve the segmentation performance. Experimental results on two public datasets, INbreast and DDSM-BCRP, demonstrate that our end-to-end network achieves better performance than state-of-the-art approaches. footnotehttps://github.com/wentaozhu/adversarial-deep-structural-networks.git