Dual-domain Multi-path Self-supervised Diffusion Model for Accelerated MRI Reconstruction
Journal:
arXiv
Published Date:
Mar 24, 2025
Abstract
Magnetic resonance imaging (MRI) is a vital diagnostic tool, but its
inherently long acquisition times reduce clinical efficiency and patient
comfort. Recent advancements in deep learning, particularly diffusion models,
have improved accelerated MRI reconstruction. However, existing diffusion
models' training often relies on fully sampled data, models incur high
computational costs, and often lack uncertainty estimation, limiting their
clinical applicability. To overcome these challenges, we propose a novel
framework, called Dual-domain Multi-path Self-supervised Diffusion Model
(DMSM), that integrates a self-supervised dual-domain diffusion model training
scheme, a lightweight hybrid attention network for the reconstruction diffusion
model, and a multi-path inference strategy, to enhance reconstruction accuracy,
efficiency, and explainability. Unlike traditional diffusion-based models, DMSM
eliminates the dependency on training from fully sampled data, making it more
practical for real-world clinical settings. We evaluated DMSM on two human MRI
datasets, demonstrating that it achieves favorable performance over several
supervised and self-supervised baselines, particularly in preserving fine
anatomical structures and suppressing artifacts under high acceleration
factors. Additionally, our model generates uncertainty maps that correlate
reasonably well with reconstruction errors, offering valuable clinically
interpretable guidance and potentially enhancing diagnostic confidence.