Paper

Cycle-consistent adversarial denoising network for multiphase coronary CT angiography
- Authors: Eunhee Kang, Hyun Jung Koo, Dong Hyun Yang, Joon Bum Seo, and Jong Chul Ye
- published in Medical Physics (2018): [https://doi.org/10.1002/mp.13284]

Implementation

A PyTorch implementation of cycleGAN for multiphase coronary CT angiography based on original cycleGAN code. [https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix] (*Thanks for Jun-Yan Zhu and Taesung Park, and Tongzhou Wang.)

Requirements
- OS: The package development version is tested on Windows operating systems with Anaconda.
- Python 3.5.5
- PyTorch 0.3.1.post2

The whole data used in the paper are private data from ASAN medical center, so only 3 test samples are uploaded.
CT image files are formated in *.mat.

Training: train.py which is handled by scripts/train_for_cardiac.sh
Test: test_for_cardiac.py which is handled by scripts/test_for_cardiac.sh
- Learned network for Paper is uploaded.

Name		Name	Last commit message	Last commit date
Latest commit History 11 Commits
checkpoints/ldnet_res_t1_t8		checkpoints/ldnet_res_t1_t8
data		data
datasets		datasets
models		models
options		options
pretrained_models		pretrained_models
scripts		scripts
test_samples/test/example01/1		test_samples/test/example01/1
util		util
.gitignore		.gitignore
README.md		README.md
test_for_cardiac.py		test_for_cardiac.py
train.py		train.py