Adversarial Deep Structured Nets for Mass Segmentation from Mammograms

October 24, 2017 · Entered Twilight · 🏛 IEEE International Symposium on Biomedical Imaging

"Last commit was 6.0 years ago (≥5 year threshold)"

Evidence collected by the PWNC Scanner

Repo contents: .gitignore, CalPvalue.m, LICENSE, README.md, ballenhance.m, ballimenhance.m, binarizelabel.m, cal_pvalue.py, convertdbnpred2im.m, convertgmmpred2im.m, convertpred2im.m, crfcnn.py, crfcnn_combine.py, crfcnn_combine_ddsm.py, crfcnn_visbias.py, dice.m, fetchmassim.m, inbreastcrfcomprocess.m, isbi18_poster.pptx, loadcrfdata.m, loadssvmdata.m, miccai15.m, mirror.m, prepmaskim.m, prioresti.m, readxml.m, readxmlmasswise.m, rectbox.m, roiextract.m, roiresize.m, test_ssvm.m, testcrf.m, utils.py, utils_combine.py

Authors Wentao Zhu, Xiang Xiang, Trac D. Tran, Gregory D. Hager, Xiaohui Xie arXiv ID 1710.09288 Category cs.CV: Computer Vision Cross-listed cs.LG, cs.NE Citations 122 Venue IEEE International Symposium on Biomedical Imaging Repository https://github.com/wentaozhu/adversarial-deep-structural-networks.git} ⭐ 51 Last Checked 1 month ago

Abstract

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. \footnote{https://github.com/wentaozhu/adversarial-deep-structural-networks.git}