Unsupervised Deformable Image Registration with Local-Global Attention and Image Decomposition

2026-01-20

AI Summary

This work addresses the challenge of insufficient accuracy in unsupervised deformable medical image registration caused by the high anatomical variability across subjects. To overcome this limitation, the authors propose LGANet++, a novel framework that integrates a local-global attention mechanism with an image decomposition strategy to enhance feature interaction and multi-scale representation. This design significantly improves the robustness and generalization capability of the registration process. Extensive experiments on five public datasets demonstrate that LGANet++ consistently outperforms existing methods, achieving accuracy gains of 1.39%, 0.71%, and 6.12% in cross-patient, longitudinal (cross-time), and cross-modality (CT-MR) registration tasks, respectively.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments

Abstract

Deformable image registration is a critical technology in medical image analysis, with broad applications in clinical practice such as disease diagnosis, multi-modal fusion, and surgical navigation. Traditional methods often rely on iterative optimization, which is computationally intensive and lacks generalizability. Recent advances in deep learning have introduced attention-based mechanisms that improve feature alignment, yet accurately registering regions with high anatomical variability remains challenging. In this study, we proposed a novel unsupervised deformable image registration framework, LGANet++, which employs a novel local-global attention mechanism integrated with a unique technique for feature interaction and fusion to enhance registration accuracy, robustness, and generalizability. We evaluated our approach using five publicly available datasets, representing three distinct registration scenarios: cross-patient, cross-time, and cross-modal CT-MR registration. The results demonstrated that our approach consistently outperforms several state-of-the-art registration methods, improving registration accuracy by 1.39% in cross-patient registration, 0.71% in cross-time registration, and 6.12% in cross-modal CT-MR registration tasks. These results underscore the potential of LGANet++ to support clinical workflows requiring reliable and efficient image registration. The source code is available at https://github.com/huangzyong/LGANet-Registration.

Problem

• unsupervised deformable image registration
• medical image analysis
• anatomical variability
• cross-modal registration
• image alignment

Innovation

• unsupervised deformable registration
• local-global attention
• image decomposition
• feature fusion
• cross-modal registration