---

Noise-generating and imaging mechanism inspired implicit regularization learning network for low dose ct reconstrution

(IEEE Transactions on Medical Imaging, 2024)

Overview

This repository provides the official implementation of the paper:

Noise-generating and imaging mechanism inspired implicit regularization learning network for low dose ct reconstrution
IEEE Transactions on Medical Imaging (TMI), 2024

If you find this code useful for your research, please consider citing our paper.

---

Citation

@article{li2023noise,
  title={Noise-generating and imaging mechanism inspired implicit regularization learning network for low dose ct reconstrution},
  author={Li, Xing and Jing, Kaili and Yang, Yan and Wang, Yongbo and Ma, Jianhua and Zheng, Hairong and Xu, Zongben},
  journal={IEEE Transactions on Medical Imaging},
  volume={43},
  number={5},
  pages={1677--1689},
  year={2024},
  publisher={IEEE}
}

---

Installation

This guide shows how to set up the environment using conda.

1. Clone the repository

git clone https://github.com/lixing0810/NGIM-IRL.git
cd NGIM-IRL

2. Create and activate the conda environment

conda create --name NGIM-IRL python=3.9
conda activate NGIM-IRL

3. Install PyTorch

pip install torch torchvision

⚠️ You may need to install the CUDA-specific version of PyTorch depending on your GPU setup.

4. Install required dependencies

pip install -r requirements.txt

---

Training

To start training the model, simply run:

python train.py