zax130 / rdp

Recursive Deformable Pyramid Network for Unsupervised Medical Image Registration (TMI2024)

By Haiqiao Wang, Dong Ni, Yi Wang.

Paper link: [TMI]

Description

An unsupervised brain MR deformable registration method that achieves precise alignment through a pure convolutional pyramid structure and a semantics-infused progressive recursive inter-level looping strategy for modeling complex deformations, even without pre-alignment of brain MR images.

Dataset

The official access addresses of the public data sets are as follows：

LPBA [link]

Mindboggle [link]

IXI [link] [freesurfer link]

Note that we use the processed IXI dataset provided by freesurfer.

Instructions

For convenience, we are sharing the preprocessed LPBA dataset used in our experiments. Once uncompressed, simply modify the "LPBA_path" in train.py to the path name of the extracted data. Next, you can execute train.py to train the network, and after training, you can run infer.py to test the network performance.

Citation

If you use the code in your research, please cite:

@ARTICLE{10423043,
  author={Wang, Haiqiao and Ni, Dong and Wang, Yi},
  journal={IEEE Transactions on Medical Imaging}, 
  title={Recursive Deformable Pyramid Network for Unsupervised Medical Image Registration}, 
  year={2024},
  volume={},
  number={},
  pages={1-1},
  keywords={Deformation;Decoding;Feature extraction;Deformable models;Training;Image resolution;Image registration;Deformable image registration;convolutional neural networks;brain MRI},
  doi={10.1109/TMI.2024.3362968}}

The overall framework and some network components of the code are heavily based on TransMorph and VoxelMorph. We are very grateful for their contributions.

The file makePklDataset.py shows how to make a pkl dataset from the original LPBA dataset. If you have any other questions about the .pkl format, please refer to the github page of [TransMorph_on_IXI].

Baseline Methods

Several PyTorch implementations of some baseline methods can be found at [SmileCode].

rdp's Issues

LPBA数据集问题

尊敬的作者您好，非常感谢您能提供代码和预处理数据集供我们学习。
我在使用您提供的预处理数据集运行代码时出现了问题：

应该是使用LPAB数据集时产生了问题，请问您当时遇到这个问题了吗？

Mindboggle101数据集的训练问题

您好，我想请问一下Mindboggle101数据集的训练问题，我使用np.unique()函数获取了测试集的标签值并替换了seg_norm()函数中的值以及dice_val_VOI()函数中的索引值，如下图所示，但验证集的结果并不正确。请问我具体应该怎么做呢？期待您的回复，感谢！

桥哥神中神

文章好，代码好，不仅开源，还提供baseline代码，唯一真神

Mindboggle101数据集以及标签问题

尊敬的作者您好，打扰您了，感谢您提供了Mindboggle数据集，有两个问题想请教您
1.Mindboggle数据集我第一次使用，请问您是用的哪个压缩包作为图片，哪个作为标签？

2.关于dice_val_VOI中标签不匹配的问题
想请问原VOI_lbls中有一些标签值与数据集标签值中有些对不上，这样是否会影响到dice的平均值，我能否改成与seg_table中相匹配的标签值

Mindboggle数据集

您好，想请问一下Mindboggle数据集需要做哪些预处理工作？我拿到的Mindboggle是(182, 218, 182)，请问您是使用什么工具进行裁剪的？裁剪后需要做归一化吗？感谢🙇‍

datasets problem

Hello, can you provide the processed datasets