Generative Image Translation for Data Augmentation in Colorectal Histopathology Images

Jerry Wei, Arief Suriawinata, Louis Vaickus, Bing Ren, Xiaoying Liu, Jason Wei, Saeed Hassanpour
; Proceedings of the Machine Learning for Health NeurIPS Workshop, PMLR 116:10-24, 2020.

Abstract

We present an image translation approach to generate augmented data for mitigating data imbalances in a dataset of histopathology images of colorectal polyps, adenomatous tumors that can lead to colorectal cancer if left untreated. By applying cycle-consistent generative adversarial networks (CycleGANs) to a source domain of normal colonic mucosa images, we generate synthetic colorectal polyp images that belong to diagnostically less common polyp classes. Generated images maintain the general structure of their source image but exhibit adenomatous features that can be enhanced with our proposed filtration module, called Path-Rank-Filter. We evaluate the quality of generated images through Turing tests with four gastrointestinal pathologists, finding that at least two of the four pathologists could not identify generated images at a statistically significant level. Finally, we demonstrate that using CycleGAN-generated images to augment training data improves the AUC of a convolutional neural network for detecting sessile serrated adenomas by over 10%, suggesting that our approach might warrant further research for other histopathology image classification tasks.

Cite this Paper


BibTeX
@InProceedings{pmlr-v116-wei20a, title = {{Generative Image Translation for Data Augmentation in Colorectal Histopathology Images}}, author = {Wei, Jerry and Suriawinata, Arief and Vaickus, Louis and Ren, Bing and Liu, Xiaoying and Wei, Jason and Hassanpour, Saeed}, booktitle = {Proceedings of the Machine Learning for Health NeurIPS Workshop}, pages = {10--24}, year = {2020}, editor = {Adrian V. Dalca and Matthew B.A. McDermott and Emily Alsentzer and Samuel G. Finlayson and Michael Oberst and Fabian Falck and Brett Beaulieu-Jones}, volume = {116}, series = {Proceedings of Machine Learning Research}, address = {}, month = {13 Dec}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v116/wei20a/wei20a.pdf}, url = {http://proceedings.mlr.press/v116/wei20a.html}, abstract = {We present an image translation approach to generate augmented data for mitigating data imbalances in a dataset of histopathology images of colorectal polyps, adenomatous tumors that can lead to colorectal cancer if left untreated. By applying cycle-consistent generative adversarial networks (CycleGANs) to a source domain of normal colonic mucosa images, we generate synthetic colorectal polyp images that belong to diagnostically less common polyp classes. Generated images maintain the general structure of their source image but exhibit adenomatous features that can be enhanced with our proposed filtration module, called Path-Rank-Filter. We evaluate the quality of generated images through Turing tests with four gastrointestinal pathologists, finding that at least two of the four pathologists could not identify generated images at a statistically significant level. Finally, we demonstrate that using CycleGAN-generated images to augment training data improves the AUC of a convolutional neural network for detecting sessile serrated adenomas by over 10%, suggesting that our approach might warrant further research for other histopathology image classification tasks.} }
Endnote
%0 Conference Paper %T Generative Image Translation for Data Augmentation in Colorectal Histopathology Images %A Jerry Wei %A Arief Suriawinata %A Louis Vaickus %A Bing Ren %A Xiaoying Liu %A Jason Wei %A Saeed Hassanpour %B Proceedings of the Machine Learning for Health NeurIPS Workshop %C Proceedings of Machine Learning Research %D 2020 %E Adrian V. Dalca %E Matthew B.A. McDermott %E Emily Alsentzer %E Samuel G. Finlayson %E Michael Oberst %E Fabian Falck %E Brett Beaulieu-Jones %F pmlr-v116-wei20a %I PMLR %J Proceedings of Machine Learning Research %P 10--24 %U http://proceedings.mlr.press %V 116 %W PMLR %X We present an image translation approach to generate augmented data for mitigating data imbalances in a dataset of histopathology images of colorectal polyps, adenomatous tumors that can lead to colorectal cancer if left untreated. By applying cycle-consistent generative adversarial networks (CycleGANs) to a source domain of normal colonic mucosa images, we generate synthetic colorectal polyp images that belong to diagnostically less common polyp classes. Generated images maintain the general structure of their source image but exhibit adenomatous features that can be enhanced with our proposed filtration module, called Path-Rank-Filter. We evaluate the quality of generated images through Turing tests with four gastrointestinal pathologists, finding that at least two of the four pathologists could not identify generated images at a statistically significant level. Finally, we demonstrate that using CycleGAN-generated images to augment training data improves the AUC of a convolutional neural network for detecting sessile serrated adenomas by over 10%, suggesting that our approach might warrant further research for other histopathology image classification tasks.
APA
Wei, J., Suriawinata, A., Vaickus, L., Ren, B., Liu, X., Wei, J. & Hassanpour, S.. (2020). Generative Image Translation for Data Augmentation in Colorectal Histopathology Images. Proceedings of the Machine Learning for Health NeurIPS Workshop, in PMLR 116:10-24

Related Material