MRI super-resolution reconstruction using efficient diffusion probabilistic model with residual shifting

2025-03-03

AI Summary

To address the excessive sampling steps and structural detail loss in MRI super-resolution (SR) reconstruction using diffusion models, this paper proposes Res-SRDiff. The method introduces a residual error offset mechanism during the forward diffusion process to explicitly align the distributions of degraded high-resolution and low-resolution images, enabling ultra-efficient sampling. Integrated with a Vision Transformer backbone and multi-scale perceptual loss, Res-SRDiff preserves anatomical fidelity. It achieves high-fidelity reconstruction in only four sampling steps. Quantitative evaluation on brain T1 MP2RAGE and prostate T2-weighted MRI datasets demonstrates statistically significant improvements (p ≪ 0.05) in PSNR, SSIM, and GMSD over baselines including TM-DDPM. Each slice inference takes under one second, achieving a 20× speedup over conventional diffusion-based SR methods—thereby overcoming the computational redundancy bottleneck of diffusion models in medical image SR.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Multimodal Learning
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Objective:This study introduces a residual error-shifting mechanism that drastically reduces sampling steps while preserving critical anatomical details, thus accelerating MRI reconstruction. Approach:We propose a novel diffusion-based SR framework called Res-SRDiff, which integrates residual error shifting into the forward diffusion process. This enables efficient HR image reconstruction by aligning the degraded HR and LR distributions.We evaluated Res-SRDiff on ultra-high-field brain T1 MP2RAGE maps and T2-weighted prostate images, comparing it with Bicubic, Pix2pix, CycleGAN, and a conventional denoising diffusion probabilistic model with vision transformer backbone (TM-DDPM), using quantitative metrics such as peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), gradient magnitude similarity deviation (GMSD), and learned perceptual image patch similarity (LPIPS). Main results: Res-SRDiff significantly outperformed all comparative methods in terms of PSNR, SSIM, and GMSD across both datasets, with statistically significant improvements (p-values<<0.05). The model achieved high-fidelity image restoration with only four sampling steps, drastically reducing computational time to under one second per slice, which is substantially faster than conventional TM-DDPM with around 20 seconds per slice. Qualitative analyses further demonstrated that Res-SRDiff effectively preserved fine anatomical details and lesion morphology in both brain and pelvic MRI images. Significance: Our findings show that Res-SRDiff is an efficient and accurate MRI SR method, markedly improving computational efficiency and image quality. Integrating residual error shifting into the diffusion process allows for rapid and robust HR image reconstruction, enhancing clinical MRI workflows and advancing medical imaging research. The source at:https://github.com/mosaf/Res-SRDiff

Problem

• Accelerates MRI reconstruction with fewer sampling steps.
• Preserves anatomical details in high-resolution MRI images.
• Improves computational efficiency and image quality significantly.

Innovation

• Res-SRDiff integrates residual error shifting
• Efficient HR image reconstruction in four steps
• Significantly reduces computational time to under one second