CARE: Class Attention to Regions of Lesion for Classification on Imbalanced Data

Jiaxin Zhuang, Jiabin Cai, Ruixuan Wang, Jianguo Zhang, Weishi Zheng
Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning, PMLR 102:588-597, 2019.

Abstract

To date, it is still an open and challenging problem for intelligent diagnosis systems to effectively learn from imbalanced data, especially with large samples of common diseases and much smaller samples of rare ones. Inspired by the process of human learning, this paper proposes a novel and effective way to embed attention into the machine learning process, particularly for learning characteristics of rare diseases. This approach does not change architectures of the original CNN classifiers and therefore can directly plug and play for any existing CNN architecture. Comprehensive experiments on a skin lesion dataset and a pneumonia chest X-ray dataset showed that paying attention to lesion regions of rare diseases during learning not only improved the classification performance on rare diseases, but also on the mean class accuracy.

Cite this Paper

BibTeX
@InProceedings{pmlr-v102-zhuang19a, title = {CARE: Class Attention to Regions of Lesion for Classification on Imbalanced Data}, author = {Zhuang, Jiaxin and Cai, Jiabin and Wang, Ruixuan and Zhang, Jianguo and Zheng, Weishi}, booktitle = {Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning}, pages = {588--597}, 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/zhuang19a/zhuang19a.pdf}, url = {https://proceedings.mlr.press/v102/zhuang19a.html}, abstract = {To date, it is still an open and challenging problem for intelligent diagnosis systems to effectively learn from imbalanced data, especially with large samples of common diseases and much smaller samples of rare ones. Inspired by the process of human learning, this paper proposes a novel and effective way to embed attention into the machine learning process, particularly for learning characteristics of rare diseases. This approach does not change architectures of the original CNN classifiers and therefore can directly plug and play for any existing CNN architecture. Comprehensive experiments on a skin lesion dataset and a pneumonia chest X-ray dataset showed that paying attention to lesion regions of rare diseases during learning not only improved the classification performance on rare diseases, but also on the mean class accuracy.} }
Endnote
%0 Conference Paper %T CARE: Class Attention to Regions of Lesion for Classification on Imbalanced Data %A Jiaxin Zhuang %A Jiabin Cai %A Ruixuan Wang %A Jianguo Zhang %A Weishi Zheng %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-zhuang19a %I PMLR %P 588--597 %U https://proceedings.mlr.press/v102/zhuang19a.html %V 102 %X To date, it is still an open and challenging problem for intelligent diagnosis systems to effectively learn from imbalanced data, especially with large samples of common diseases and much smaller samples of rare ones. Inspired by the process of human learning, this paper proposes a novel and effective way to embed attention into the machine learning process, particularly for learning characteristics of rare diseases. This approach does not change architectures of the original CNN classifiers and therefore can directly plug and play for any existing CNN architecture. Comprehensive experiments on a skin lesion dataset and a pneumonia chest X-ray dataset showed that paying attention to lesion regions of rare diseases during learning not only improved the classification performance on rare diseases, but also on the mean class accuracy.
APA
Zhuang, J., Cai, J., Wang, R., Zhang, J. & Zheng, W.. (2019). CARE: Class Attention to Regions of Lesion for Classification on Imbalanced Data. Proceedings of The 2nd International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 102:588-597 Available from https://proceedings.mlr.press/v102/zhuang19a.html.

Related Material