PWD: Prior-Guided and Wavelet-Enhanced Diffusion Model for Limited-Angle CT

2025-06-30

AI Summary

To address detail degradation in limited-angle CT (LACT) reconstruction caused by reduced diffusion sampling steps, this paper proposes a fast diffusion-based reconstruction method integrating explicit prior guidance and wavelet-domain multi-scale features. Methodologically, the low-dose LACT reconstruction serves as an explicit prior embedded into the diffusion process to constrain the sampling trajectory; additionally, a wavelet-domain multi-scale feature fusion module and a distribution mapping learning mechanism are jointly designed to enhance high-frequency structural recovery. Experiments on clinical dental cone-beam CT data demonstrate that the method achieves superior performance over state-of-the-art approaches using only 50 sampling steps—yielding ≥1.7 dB PSNR gain and 10% SSIM improvement—effectively balancing reconstruction fidelity and inference efficiency. The core contributions lie in a prior-driven lightweight sampling path optimization strategy and a novel wavelet-diffusion joint modeling framework.

Technology Category

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

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Content-based information diffusion
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work

Abstract

Generative diffusion models have received increasing attention in medical imaging, particularly in limited-angle computed tomography (LACT). Standard diffusion models achieve high-quality image reconstruction but require a large number of sampling steps during inference, resulting in substantial computational overhead. Although skip-sampling strategies have been proposed to improve efficiency, they often lead to loss of fine structural details. To address this issue, we propose a prior information embedding and wavelet feature fusion fast sampling diffusion model for LACT reconstruction. The PWD enables efficient sampling while preserving reconstruction fidelity in LACT, and effectively mitigates the degradation typically introduced by skip-sampling. Specifically, during the training phase, PWD maps the distribution of LACT images to that of fully sampled target images, enabling the model to learn structural correspondences between them. During inference, the LACT image serves as an explicit prior to guide the sampling trajectory, allowing for high-quality reconstruction with significantly fewer steps. In addition, PWD performs multi-scale feature fusion in the wavelet domain, effectively enhancing the reconstruction of fine details by leveraging both low-frequency and high-frequency information. Quantitative and qualitative evaluations on clinical dental arch CBCT and periapical datasets demonstrate that PWD outperforms existing methods under the same sampling condition. Using only 50 sampling steps, PWD achieves at least 1.7 dB improvement in PSNR and 10% gain in SSIM.

Problem

• Improves limited-angle CT reconstruction quality
• Reduces computational steps in diffusion models
• Preserves fine details using wavelet fusion

Innovation

• Prior-guided sampling for efficient trajectory
• Wavelet domain multi-scale feature fusion
• Embedding structural correspondences in training