Learning-a-Model-Driven-Variational-Network-for-Deformable-Image-Registration

This is the code for the paper Learning a Model-Driven Variational Network for Deformable Image Registration

If you find the code is somehow useful, please consider cite this paper.

@ARTICLE{9525092,
author={Jia, Xi and Thorley, Alexander and Chen, Wei and Qiu, Huaqi and Shen, Linlin and Styles, Iain B. and Chang, Hyung Jin and Leonardis, Ales and de Marvao, Antonio and O’Regan, Declan P. and Rueckert, Daniel and Duan, Jinming},
journal={IEEE Transactions on Medical Imaging},
title={Learning a Model-Driven Variational Network for Deformable Image Registration},
year={2022},
volume={41},
number={1},
pages={199-212},
doi={10.1109/TMI.2021.3108881}}

The current version is kind of messy and only contains the architectures.

If you have any problems or notice any bugs related to this version, please fell free to contact me (x.jia.1@cs.bham.ac.uk).

We will update the code (such as including the training and testing code) in the future (hopefully soon). The training and testing code can be found in our LKU-Net and Fourier-Net. They are bascially the same.

Others

The 3D U-Net architecture used in this work is from ICNet and SYMNet.

You may want to have a look at Nesterov Accelerated ADMM for Fast Diffeomorphic Image Registration as well.