MRI Synthesis Using Deep Learning Models

This repository provides the official PyTorch implementation of the following paper: A Deep Learning Model for Multi-Domain MRI Synthesis Using Generative Adversarial Networks

Dependencies

• Python 3.5+
• PyTorch 0.4.0+
• TensorFlow 1.3+ (optional for tensorboard)

Installaiton

python -m venv venv
. venv/bin/activate
pip install -r requirements.txt

Training networks

To train model on both BraTS2020 and IXI:

# Train Customed ResUnet using both BraTS2020 and IXI datasets
python main.py --mode=train --dataset Both --image_size 256 --c_dim 4 --c2_dim 4 \
               --sample_dir resunet_custom_new_loss_both/samples --log_dir resunet_custom_new_loss_both/logs \
               --model_save_dir resunet_custom_new_loss_both/models --result_dir resunet_custom_new_loss_both/results \
               --batch_size 4

# Test Customed ResUnet using both BraTS2020 and IXI datasets
python main.py --mode test --dataset Both --image_size 256 --c_dim 4 --c2_dim 4 \
               --sample_dir resunet_custom_new_loss_both/samples --log_dir resunet_custom_new_loss_both/logs \
               --model_save_dir resunet_custom_new_loss_both/models --result_dir resunet_custom_new_loss_both/results \
               --batch_size 4

Citation

If you find this work useful for your research, please cite paper:

@article{han2024,
  author = {Le Hoang Ngoc Han and Ngo Le Huy Hien and Luu Van Huy and Nguyen Van Hieu},
  title = {A Deep Learning Model for Multi-Domain MRI Synthesis Using Generative Adversarial Networks},
  journal = {Informatica},
  volume = {35},
  number = {2},
  pages = {283-309},
  year = {2024},
  doi = {10.15388/24-INFOR556}
}