Scale-Cascaded Diffusion Models for Super-Resolution in Medical Imaging

2026-01-30

AI Summary

Existing medical image super-resolution methods predominantly rely on single-scale diffusion priors, which struggle to effectively capture the multi-scale structural characteristics of images, thereby limiting both reconstruction quality and computational efficiency. To address this limitation, this work proposes a Laplacian pyramid-based multi-scale diffusion model that, for the first time, integrates pyramid decomposition with diffusion priors. The approach learns distinct diffusion priors at each frequency band and employs a cascaded strategy to enable progressive, cross-scale super-resolution reconstruction. Evaluated on brain, knee, and prostate MRI datasets, the method achieves substantial improvements in perceptual quality while reducing computational overhead through the use of compact networks at coarse scales, thus balancing performance and efficiency.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Diffusion models have been increasingly used as strong generative priors for solving inverse problems such as super-resolution in medical imaging. However, these approaches typically utilize a diffusion prior trained at a single scale, ignoring the hierarchical scale structure of image data. In this work, we propose to decompose images into Laplacian pyramid scales and train separate diffusion priors for each frequency band. We then develop an algorithm to perform super-resolution that utilizes these priors to progressively refine reconstructions across different scales. Evaluated on brain, knee, and prostate MRI data, our approach both improves perceptual quality over baselines and reduces inference time through smaller coarse-scale networks. Our framework unifies multiscale reconstruction and diffusion priors for medical image super-resolution.

Problem

• super-resolution
• diffusion models
• medical imaging
• multiscale
• Laplacian pyramid

Innovation

• scale-cascaded diffusion
• Laplacian pyramid
• multiscale reconstruction
• medical image super-resolution
• diffusion priors