A Fair Comparison in Deep Learning-Based QSM Reconstruction

This repository contains the source code for training/inference of QSMnet used in the paper "A Fair Comparison in Deep Learning-Based QSM Reconstruction".

Conda environment

Create the conda environment using the provided YAML file:

conda env create -n qsmnet2 -f environment.yml

QSMnet training

1. Generate patches for deep learning

python patch.py

2. Training the network

To modify hyperparameters, update the default values in the argument parser within "train.py".

python train.py -g 0 -s ../save/qsmnet111

3. Inference the network

To modify hyperparameters, update the default values in the argument parser within "test.py".

python test.py -g 0 -s ../save/qsmnet111/result -e 25

4. Calculate evaluation metrics (NRMSE, SSIM, PSNR, HFEN)

python testmetric.py -p ../network/qsmnet111/result