Nuclear image synthesis and segmentation code

This repository contains source code for the following paper:
Robust Histopathology Image Analysis: to Label or to Synthesize? Hou, Le and Agarwal, Ayush and Samaras, Dimitris and Kurc, Tahsin M and Gupta, Rajarsi R and Saltz, Joel H. Computer Vision and Pattern Recognition (CVPR), 2019

The codes have been used to generate the datasets published in the following paper:

Le Hou, Rajarsi Gupta, John S. Van Arnam, Yuwei Zhang, Kaustubh Sivalenka, Dimitris Samaras, Tahsin M. Kurc, and Joel H. Saltz. Dataset of segmented nuclei in hematoxylin and eosin stained histopathology images of ten cancer types. Sci Data 7, 185 (2020). https://doi.org/10.1038/s41597-020-0528-1

You need the following python2 libraries to run this software:

tensorflow  
scipy  
numpy  
json  
pickle
cv2 

If you want to obtain initial synthetic images with nuclear masks, please read training-data-synthesis/README.md
If you want to train your nuclei segmentation model, or use our trained nuclei segmentation model, please read segmentation-of-nuclei/README.md

This software uses several open source repositories, especially this one

Trained model and released segmentation results

We also include:

1. A Dockerfile with trained model here
2. Segmentation results we've already generated here with readme.