TY - CHAP
T1 - Large-scale mammography CAD with deformable conv-nets
AU - Morrell, Stephen
AU - Wojna, Zbigniew
AU - Khoo, Can Son
AU - Ourselin, Sebastien
AU - Iglesias, Juan Eugenio
PY - 2018/1/1
Y1 - 2018/1/1
N2 - State-of-the-art deep learning methods for image processing are evolving into increasingly complex meta-architectures with a growing number of modules. Among them, region-based fully convolutional networks (R-FCN) and deformable convolutional nets (DCN) can improve CAD for mammography: R-FCN optimizes for speed and low consumption of memory, which is crucial for processing the high resolutions of to 50 μ m used by radiologists. Deformable convolution and pooling can model a wide range of mammographic findings of different morphology and scales, thanks to their versatility. In this study, we present a neural net architecture based on R-FCN/DCN, that we have adapted from the natural image domain to suit mammograms—particularly their larger image size—without compromising resolution. We trained the network on a large, recently released dataset (Optimam) including 6,500 cancerous mammograms. By combining our modern architecture with such a rich dataset, we achieved an area under the ROC curve of 0.879 for breast-wise detection in the DREAMS challenge (130,000 withheld images), which surpassed all other submissions in the competitive phase.
AB - State-of-the-art deep learning methods for image processing are evolving into increasingly complex meta-architectures with a growing number of modules. Among them, region-based fully convolutional networks (R-FCN) and deformable convolutional nets (DCN) can improve CAD for mammography: R-FCN optimizes for speed and low consumption of memory, which is crucial for processing the high resolutions of to 50 μ m used by radiologists. Deformable convolution and pooling can model a wide range of mammographic findings of different morphology and scales, thanks to their versatility. In this study, we present a neural net architecture based on R-FCN/DCN, that we have adapted from the natural image domain to suit mammograms—particularly their larger image size—without compromising resolution. We trained the network on a large, recently released dataset (Optimam) including 6,500 cancerous mammograms. By combining our modern architecture with such a rich dataset, we achieved an area under the ROC curve of 0.879 for breast-wise detection in the DREAMS challenge (130,000 withheld images), which surpassed all other submissions in the competitive phase.
UR - http://www.scopus.com/inward/record.url?scp=85053884757&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-00946-5_7
DO - 10.1007/978-3-030-00946-5_7
M3 - Conference paper
AN - SCOPUS:85053884757
SN - 9783030009458
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 64
EP - 72
BT - Image Analysis for Moving Organ, Breast, and Thoracic Images - Third International Workshop, RAMBO 2018, Fourth International Workshop, BIA 2018, and First International Workshop, TIA 2018, Held in Conjunction with MICCAI 2018, Proceedings
A2 - Snead, David
A2 - Trucco, Emanuele
A2 - Stoyanov, Danail
A2 - Taylor, Zeike
A2 - Maier-Hein, Lena
A2 - Rajpoot, Nasir
A2 - Bogunovic, Hrvoje
A2 - Ciompi, Francesco
A2 - Veta, Mitko
A2 - Garvin, Mona K.
A2 - Chen, Xin Jan
A2 - Martel, Anne
A2 - van der Laak, Jeroen
A2 - Xu, Yanwu
A2 - McKenna, Stephen
PB - Springer Verlag
T2 - 3rd International Workshop on Reconstruction and Analysis of Moving Body Organs, RAMBO 2018, 4th International Workshop on Breast Image Analysis, BIA 2018, and 1st International Workshop on Thoracic Image Analysis, TIA 2018, held in conjunction with 21st International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2018
Y2 - 16 September 2018 through 20 September 2018
ER -