Deformable registration is essential in medical image analysis, especially for handling various multi- and mono-modal registration tasks in neuroimaging. Existing studies lack exploration of brain MR-CT registration, and face challenges in both accuracy and efficiency improvements of learning-based methods. To enlarge the practice of multi-modal registration in brain, we present SR-Reg, a new benchmark dataset comprising 180 volumetric paired MR-CT images and annotated anatomical regions. Building on this foundation, we introduce MambaMorph, a novel deformable registration network based on an efficient state space model Mamba for global feature learning, with a fine-grained feature extractor for low-level embedding. Experimental results demonstrate that MambaMorph surpasses advanced ConvNet-based and Transformer-based networks across several multi- and mono-modal tasks, showcasing impressive enhancements of efficacy and efficiency.
Install Mamba via https://github.com/state-spaces/mamba
python ./scripts/torch/train_cross.py --gpu 1 --epochs 1 --batch-size 1 --model-dir output/train_debug --model mm-feat
python ./scripts/torch/test_cross.py --gpu 0 --model mm-feat --load-model "/home/guotao/code/voxelmorph-dev/output/train_s46/min_train.pt"
We implement MambaMorph on our brain MR-CT data SR-Reg which is developed from SynthRAD 2023 (https://synthrad2023.grand-challenge.org/). Please go Here to access the SR-Reg dataset.
https://pan.baidu.com/s/1TlxqZHl6T17on_f3BDixmA (Extract code：hwwd)
https://drive.google.com/file/d/1idT6rCKXry-Yc8GF-DBxEn6e32-pBiFF/view?usp=sharing
@article{2025mambamorph,
title={MambaMorph: a Mamba-based Framework for Medical MR-CT Deformable Registration},
author={Yinuo Wang, Tao Guo, Weimin Yuan, Shihao Shu, Cai Meng, Xiangzhi Bai,},
journal={Computerized Medical Imaging and Graphics},
year={2025}
}