# Efficient Large-Deformation Medical Image Registration via Recurrent Dynamic Correlation

arXiv: https://arxiv.org/abs/2510.22380 · HTML: https://mlsanity.com/paper/2510.22380/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 20 of its 46 references that are in the map.
- First posted on arXiv 25 Oct 2025.

## Builds on (most cited first, 20 of 20)

- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [Swin Transformer: Hierarchical Vision Transformer using Shifted Windows](https://mlsanity.com/paper/2103.14030/) (2021)
- [Rethinking the Inception Architecture for Computer Vision](https://mlsanity.com/paper/1512.00567/) (2015)
- [RAFT: Recurrent All-Pairs Field Transforms for Optical Flow](https://mlsanity.com/paper/2003.12039/) (2020)
- [VoxelMorph: A Learning Framework for Deformable Medical Image Registration](https://mlsanity.com/paper/1809.05231/) (2018)
- [Gate-Variants of Gated Recurrent Unit (GRU) Neural Networks](https://mlsanity.com/paper/1701.05923/) (2017)
- [A Deep Learning Framework for Unsupervised Affine and Deformable Image Registration](https://mlsanity.com/paper/1809.06130/) (2018)
- [Large Deformation Diffeomorphic Image Registration with Laplacian Pyramid Networks](https://mlsanity.com/paper/2006.16148/) (2020)
- [CycleMorph: Cycle Consistent Unsupervised Deformable Image Registration](https://mlsanity.com/paper/2008.05772/) (2020)
- [VoCo: A Simple-yet-Effective Volume Contrastive Learning Framework for 3D Medical Image Analysis](https://mlsanity.com/paper/2402.17300/) (2024)
- [Unsupervised 3D End-to-End Medical Image Registration with Volume Tweening Network](https://mlsanity.com/paper/1902.05020/) (2019)
- [A survey on deep learning in medical image registration: new technologies, uncertainty, evaluation metrics, and beyond](https://mlsanity.com/paper/2307.15615/) (2023)
- [Dual-Stream Pyramid Registration Network](https://mlsanity.com/paper/1909.11966/) (2019)
- [XMorpher: Full Transformer for Deformable Medical Image Registration via Cross Attention](https://mlsanity.com/paper/2206.07349/) (2022)
- [Closing the Gap between Deep and Conventional Image Registration using Probabilistic Dense Displacement Networks](https://mlsanity.com/paper/1907.10931/) (2019)
- [Correlation-aware Coarse-to-fine MLPs for Deformable Medical Image Registration](https://mlsanity.com/paper/2406.00123/) (2024)
- [mlVIRNET: Multilevel Variational Image Registration Network](https://mlsanity.com/paper/1909.10084/) (2019)
- [Robust Image Registration with Absent Correspondences in Pre-operative and Follow-up Brain MRI Scans of Diffuse Glioma Patients](https://mlsanity.com/paper/2210.11045/) (2022)
- [One shot PACS: Patient specific Anatomic Context and Shape prior aware recurrent registration-segmentation of longitudinal thoracic cone beam CTs](https://mlsanity.com/paper/2201.11000/) (2022)
- [Decoder-Only Image Registration](https://mlsanity.com/paper/2402.03585/) (2024)

## Abstract (arXiv)

Deformable image registration estimates voxel-wise correspondences between images through spatial transformations, and plays a key role in medical imaging. While deep learning methods have significantly reduced runtime, efficiently handling large deformations remains a challenging task. Convolutional networks aggregate local features but lack direct modeling of voxel correspondences, promoting recent works to explore explicit feature matching. Among them, voxel-to-region matching is more efficient for direct correspondence modeling by computing local correlation features whithin neighbourhoods, while region-to-region matching incurs higher redundancy due to excessive correlation pairs across large regions. However, the inherent locality of voxel-to-region matching hinders the capture of long-range correspondences required for large deformations. To address this, we propose a Recurrent Correlation-based framework that dynamically relocates the matching region toward more promising positions. At each step, local matching is performed with low cost, and the estimated offset guides the next search region, supporting efficient convergence toward large deformations. In addition, we uses a lightweight recurrent update module with memory capacity and decouples motion-related and texture features to suppress semantic redundancy. We conduct extensive experiments on brain MRI and abdominal CT datasets under two settings: with and without affine pre-registration. Results show that our method exibits a strong accuracy-computation trade-off, surpassing or matching the state-of-the-art performance. For example, it achieves comparable performance on the non-affine OASIS dataset, while using only 9.5% of the FLOPs and running 96% faster than RDP, a representative high-performing method.