YiyuHong/ck_virtual_staining_paper

transform H&E image into CK image using GAN (virtual staining)

Introduction

virtual staining

This repository is for virtual staining (transforming H&E image into CK (cytokeratin) image).

Requirement

• python >= 3.6
  • numpy >=1.17.4
  • matplotlib >=3.1.1
  • opencv-python >= 4.1.2.30
  • openslide-python >= 1.1.2
  • pandas >= 1.1.3
  • scikit-image >= 0.15.0
  • scikit-learn >= 0.23.2
  • tifffile >= 2020.7.4
  • torch >= 1.5.1 (https://pytorch.org/)
  • torchvision >= 0.6.1
  • tqdm >= 4.50.2
• openslide >= 3.4.1 (https://openslide.org/)

Usage

• -input_path: path for input H&E image (.svs). example images in data/input
• -output_dir: directory for output CK image
• -model_checkpoint: path for model weight (https://drive.google.com/file/d/1-1TqdXRCApmqjSXXiDNwsMbo4Uz8EITJ/view?usp=sharing)

python run_ck_virtual_staining.py -input_path data/input/example.svs -output_dir data/output -model_checkpoint data/checkpoint/model.pth

Example Image

Input	Output_CK	Output_Segmentation (Tumor(red), Stroma(green))

Reference

Hong, Y., Heo, Y.J., Kim, B. et al. Deep learning-based virtual cytokeratin staining of gastric carcinomas to measure tumor–stroma ratio. Sci Rep 11, 19255 (2021). https://doi.org/10.1038/s41598-021-98857-1