Generative Image Translation for Data Augmentation of Bone Lesion Pathology

Anant Gupta, Srivas Venkatesh, Sumit Chopra, Christian Ledig
Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning, PMLR 102:225-235, 2019.

Abstract

Insufficient training data and severe class imbalance are often limiting factors when developing machine learning models for the classification of rare diseases. In this work, we address the problem of classifying bone lesions from X-ray images by increasing the small number of positive samples in the training set. We propose a generative data augmentation approach based on a cycle-consistent generative adversarial network that synthesizes bone lesions on images without pathology. We pose the generative task as an image-patch translation problem that we optimize specifically for distinct bones (humerus, tibia, femur). In experimental results, we confirm that the described method mitigates the class imbalance problem in the binary classification task of bone lesion detection. We show that the augmented training sets enable the training of superior classifiers achieving better performance on a held-out test set. Additionally, we demonstrate the feasibility of transfer learning and apply a generative model that was trained on one body part to another.

Cite this Paper

BibTeX
@InProceedings{pmlr-v102-gupta19b, title = {Generative Image Translation for Data Augmentation of Bone Lesion Pathology}, author = {Gupta, Anant and Venkatesh, Srivas and Chopra, Sumit and Ledig, Christian}, booktitle = {Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning}, pages = {225--235}, year = {2019}, editor = {Cardoso, M. Jorge and Feragen, Aasa and Glocker, Ben and Konukoglu, Ender and Oguz, Ipek and Unal, Gozde and Vercauteren, Tom}, volume = {102}, series = {Proceedings of Machine Learning Research}, month = {08--10 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v102/gupta19b/gupta19b.pdf}, url = {https://proceedings.mlr.press/v102/gupta19b.html}, abstract = {Insufficient training data and severe class imbalance are often limiting factors when developing machine learning models for the classification of rare diseases. In this work, we address the problem of classifying bone lesions from X-ray images by increasing the small number of positive samples in the training set. We propose a generative data augmentation approach based on a cycle-consistent generative adversarial network that synthesizes bone lesions on images without pathology. We pose the generative task as an image-patch translation problem that we optimize specifically for distinct bones (humerus, tibia, femur). In experimental results, we confirm that the described method mitigates the class imbalance problem in the binary classification task of bone lesion detection. We show that the augmented training sets enable the training of superior classifiers achieving better performance on a held-out test set. Additionally, we demonstrate the feasibility of transfer learning and apply a generative model that was trained on one body part to another.} }
Endnote
%0 Conference Paper %T Generative Image Translation for Data Augmentation of Bone Lesion Pathology %A Anant Gupta %A Srivas Venkatesh %A Sumit Chopra %A Christian Ledig %B Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2019 %E M. Jorge Cardoso %E Aasa Feragen %E Ben Glocker %E Ender Konukoglu %E Ipek Oguz %E Gozde Unal %E Tom Vercauteren %F pmlr-v102-gupta19b %I PMLR %P 225--235 %U https://proceedings.mlr.press/v102/gupta19b.html %V 102 %X Insufficient training data and severe class imbalance are often limiting factors when developing machine learning models for the classification of rare diseases. In this work, we address the problem of classifying bone lesions from X-ray images by increasing the small number of positive samples in the training set. We propose a generative data augmentation approach based on a cycle-consistent generative adversarial network that synthesizes bone lesions on images without pathology. We pose the generative task as an image-patch translation problem that we optimize specifically for distinct bones (humerus, tibia, femur). In experimental results, we confirm that the described method mitigates the class imbalance problem in the binary classification task of bone lesion detection. We show that the augmented training sets enable the training of superior classifiers achieving better performance on a held-out test set. Additionally, we demonstrate the feasibility of transfer learning and apply a generative model that was trained on one body part to another.
APA
Gupta, A., Venkatesh, S., Chopra, S. & Ledig, C.. (2019). Generative Image Translation for Data Augmentation of Bone Lesion Pathology. Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 102:225-235 Available from https://proceedings.mlr.press/v102/gupta19b.html.

Related Material