DC-cycleGAN: Bidirectional CT-to-MR Synthesis from Unpaired Data

DC-cycleGAN: Bidirectional CT-to-MR Synthesis from Unpaired Data

---

DC-CycleGAN

Installation

Please use Python 3.6

pip3 install -r requirements.txt

Usage

Train:

python DC-CycleGAN.py --lambdaG=5 --date='time1' --dataset='100dataset'

You can change "--date" to any name you want. We use time1 because we train each method five times. Then calculate the average as the final quantitative result.

Test:

Please follows "test_image.ipynb" annotation to run, you should change the path to yours.
When you test your model meet the problem about "alueError: Dimension 0 in both shapes must be equal, but are 3 and 128. Shapes are [3,3,128,64] and [128,64,3,3]."
Please check the training code and test code setting for opt['use_resize_convolution'], they should be kept the same and then will work. 

Then use "evaluate.ipynb" to calculate MAE, PSNR, and SSIM.

Note

I am really thanks for your interest in my work, if you have any problem with the code, please contact me or submit an issue, and I will reply to you if I can help you.

Citation:

If you use this code for your research, please give us a star and cite our paper:

@article{wang2023dc,
title={DC-cycleGAN: Bidirectional CT-to-MR synthesis from unpaired data},
author={Wang, Jiayuan and Wu, QM Jonathan and Pourpanah, Farhad},
journal={Computerized Medical Imaging and Graphics},
pages={102249},
year={2023},
publisher={Elsevier} }