Generative Image Translation for Data Augmentation in Colorectal Histopathology Images
arXiv:1910.05827
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.
NeurIPS 2019 Machine Learning for Health Workshop Full Paper (19/111 accepted papers = 17% acceptance rate)
References in corpus (8)
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks
- Deep Residual Learning for Image Recognition
- The Effectiveness of Data Augmentation in Image Classification using Deep Learning
- Neural Stain-Style Transfer Learning using GAN for Histopathological Images
- Computational Histological Staining and Destaining of Prostate Core Biopsy RGB Images with Generative Adversarial Neural Networks
- Unsupervised Histopathology Image Synthesis
- Deep-Learning for Classification of Colorectal Polyps on Whole-Slide Images
- Cross-modality image synthesis from unpaired data using CycleGAN: Effects of gradient consistency loss and training data size