FSDiffReg: Feature-wise and Score-wise Diffusion-guided Unsupervised Deformable Image Registration for Cardiac Images

TL;DR

This paper introduces FSDiffReg, a novel unsupervised deformable image registration method for cardiac images that leverages diffusion models to guide deformation field generation and topology preservation, achieving refined results.

Contribution

The paper proposes two diffusion-guided modules, FDG and SDG, to improve deformation field quality and topology preservation in unsupervised cardiac image registration.

Findings

Effective deformation field refinement demonstrated

Topology preservation achieved with minimal additional computation

Validated on 3D cardiac image registration tasks

Abstract

Unsupervised deformable image registration is one of the challenging tasks in medical imaging. Obtaining a high-quality deformation field while preserving deformation topology remains demanding amid a series of deep-learning-based solutions. Meanwhile, the diffusion model's latent feature space shows potential in modeling the deformation semantics. To fully exploit the diffusion model's ability to guide the registration task, we present two modules: Feature-wise Diffusion-Guided Module (FDG) and Score-wise Diffusion-Guided Module (SDG). Specifically, FDG uses the diffusion model's multi-scale semantic features to guide the generation of the deformation field. SDG uses the diffusion score to guide the optimization process for preserving deformation topology with barely any additional computation. Experiment results on the 3D medical cardiac image registration task validate our model's ability to provide refined deformation fields with preserved topology effectively. Code is available at: https://github.com/xmed-lab/FSDiffReg.git.