LDM-Morph: Latent Diffusion Model Guided Deformable Image Registration

Authors

Jiong Wu
Tinsu Pan
Kuang Gong

Language

English

Publication Date

6-1-2026

Journal

Pattern Recognition

DOI

10.1016/j.patcog.2025.112925

PMID

41736822

PMCID

PMC12928757

PubMedCentral® Posted Date

2-28-2026

PubMedCentral® Full Text Version

Author MSS

Abstract

Deformable image registration plays an essential role in various medical image tasks. Existing deep learning-based deformable registration frameworks primarily utilize convolutional neural networks (CNNs) or Transformers to learn features to predict the deformations. However, the lack of semantic information in the learned features limits the registration performance. Furthermore, the similarity metric of the loss function is often evaluated only in the pixel space, which ignores the matching of high-level anatomical features and can lead to deformation folding. To address these issues, in this work, we proposed LDM-Morph, an unsupervised deformable registration algorithm for medical image registration. LDM-Morph integrated features extracted from the latent diffusion model (LDM) to enrich the semantic information. Additionally, a latent and global feature-based cross-attention module (LGCA) was designed to enhance the interaction of semantic information from LDM and global information from multi-head self-attention operations. Finally, a hierarchical metric was proposed to evaluate the similarity of image pairs in both the original pixel space and latent-feature space, enhancing topology preservation while improving registration accuracy. Extensive experiments on four public 2D cardiac image datasets, two 3D image datasets, show that the proposed LDM-Morph framework outperformed existing state-of-the-art CNNs-and Transformers-based registration methods regarding accuracy with comparable topology preservation and computational efficiency. Our code is publicly available at: https://github.com/wujiong-hub/LDM-Morph.

Keywords

Deformable registration, Latent diffusion model, Dual-stream cross learning, Latent feature, Unsupervised learning

Published Open-Access

yes

Recommended Citation

Jiong Wu, Tinsu Pan, and Kuang Gong, "LDM-Morph: Latent Diffusion Model Guided Deformable Image Registration" (2026). Faculty, Staff and Student Publications. 6386.
https://digitalcommons.library.tmc.edu/uthgsbs_docs/6386