MDPG: Multi-domain Diffusion Prior Guidance for MRI Reconstruction

2025-06-30

AI Summary

MRI undersampled reconstruction faces challenges of low fidelity and poor data consistency in generative models. To address these, we propose a multi-domain diffusion prior-guided framework: leveraging a pre-trained latent diffusion model to jointly provide strong conditional priors in both latent and image spaces; introducing a Latent-Guided Attention (LGA) mechanism and a Dual-domain Fusion Branch (DFB) for adaptive cross-domain feature alignment; and incorporating a non-self-calibrating signal-driven k-space regularization to enforce frequency-domain consistency. Additionally, we integrate a visual Mamba backbone to effectively model long-range spatial dependencies. Evaluated on two public MRI datasets, our method achieves significant improvements in image quality (PSNR/SSIM) and k-space consistency over state-of-the-art approaches. The source code is publicly available.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Magnetic Resonance Imaging (MRI) reconstruction is essential in medical diagnostics. As the latest generative models, diffusion models (DMs) have struggled to produce high-fidelity images due to their stochastic nature in image domains. Latent diffusion models (LDMs) yield both compact and detailed prior knowledge in latent domains, which could effectively guide the model towards more effective learning of the original data distribution. Inspired by this, we propose Multi-domain Diffusion Prior Guidance (MDPG) provided by pre-trained LDMs to enhance data consistency in MRI reconstruction tasks. Specifically, we first construct a Visual-Mamba-based backbone, which enables efficient encoding and reconstruction of under-sampled images. Then pre-trained LDMs are integrated to provide conditional priors in both latent and image domains. A novel Latent Guided Attention (LGA) is proposed for efficient fusion in multi-level latent domains. Simultaneously, to effectively utilize a prior in both the k-space and image domain, under-sampled images are fused with generated full-sampled images by the Dual-domain Fusion Branch (DFB) for self-adaption guidance. Lastly, to further enhance the data consistency, we propose a k-space regularization strategy based on the non-auto-calibration signal (NACS) set. Extensive experiments on two public MRI datasets fully demonstrate the effectiveness of the proposed methodology. The code is available at https://github.com/Zolento/MDPG.

Problem

• Enhance MRI reconstruction fidelity using diffusion models
• Integrate latent and image domain priors for better guidance
• Improve data consistency with k-space regularization strategy

Innovation

• Visual-Mamba-based backbone for efficient encoding
• Latent Guided Attention for multi-level fusion
• Dual-domain Fusion Branch for k-space guidance