Synthesis of diagnostic quality cancer pathology images

This is the implementation of the Synthesis of diagnostic quality cancer pathology images Citation information:

@article{doi:10.1002/path.5509,
author = {Levine, Adrian B. and Peng, Jason and Farnell, David and Nursey, Mitchell and Wang, Yiping and Naso, Julia R. and Ren, Hezhen and Farahani, Hossein and Chen, Colin and Chiu, Derek and Talhouk, Aline and Sheffield, Brandon and Riazy, Maziar and Ip, Philip P. and Parra-Herran, Carlos and Mills, Anne and Singh, Naveena and Tessier-Cloutier, Basile and Salisbury, Taylor and Lee, Jonathan and Salcudean, Tim and Jones, Steven J.M. and Huntsman, David G. and Gilks, C. Blake and Yip, Stephen and Bashashati, Ali},
title = {Synthesis of diagnostic quality cancer pathology images by generative adversarial networks},
journal = {The Journal of Pathology},
volume = {n/a},
number = {n/a},
pages = {},
doi = {10.1002/path.5509},
url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/path.5509},
eprint = {https://onlinelibrary.wiley.com/doi/pdf/10.1002/path.5509},
}

Repo structure

• generative_models contains the implementation of the custom progressive GAN used in this study.

  • Based on the Official TensorFlow implementation of the ICLR 2018 paper Progressive Growing of GANs for Improved Quality, Stability, and Variation

• classification contains the implementation of our evaluation experiments for both OVCARE and TCGA dataset.