Multi-Phase Liver-Specific DCE-MRI Translation via a Registration-Guided GAN

This paper has been accepted by: The 8th Simulation and Synthesis in Medical Imaging (SASHIMI) workshop, MICCAI 2023.

Prepare dataset and meta information

-dataset
    - training_path
        - id_1.npy
        - id_2.npy
        ...
    - test_path
        - id_1.npy
        - id_2.npy
        ...

You can use ./data/split.py to get meta imformation of your dataset.

data
    - train.txt
    - val.txt
    - test.txt

Modify config file

Train and test

Pretrain Unet:

python train_UNet.py --config './Yaml/UNet.yaml'

Start visdom：

python -m visdom.server -p 6019

Train:

python train.py --config './Yaml/mrgan.yaml'

Test:

python train.py --config './Yaml/mrgan.yaml'