GR-Diffusion: 3D Gaussian Representation Meets Diffusion in Whole-Body PET Reconstruction

2026-02-12

AI Summary

This work addresses noise amplification and detail loss in low-dose whole-body PET reconstruction caused by sparse sampling and the ill-posed nature of the inverse problem. It proposes the first framework that synergistically integrates 3D Gaussian representation (GR) with a diffusion model for PET image reconstruction. The method leverages GR to provide geometric priors, generating structurally coherent reference images, and introduces a dual-granularity, multi-scale hierarchical guidance mechanism to jointly enforce global consistency and recover fine local details during the diffusion process. By transcending the low-pass limitations inherent in conventional point-based or voxel-based approaches, the framework effectively reconstructs sub-voxel information. Experiments demonstrate that the model significantly outperforms state-of-the-art methods on both the UDPET and clinical datasets, consistently enhancing image quality across varying dose levels while preserving critical anatomical structures.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Reasoning under Uncertainty: Graphical Models

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Web Mining and Content Analysis: Content-based information diffusion

Abstract

Positron emission tomography (PET) reconstruction is a critical challenge in molecular imaging, often hampered by noise amplification, structural blurring, and detail loss due to sparse sampling and the ill-posed nature of inverse problems. The three-dimensional discrete Gaussian representation (GR), which efficiently encodes 3D scenes using parameterized discrete Gaussian distributions, has shown promise in computer vision. In this work, we pro-pose a novel GR-Diffusion framework that synergistically integrates the geometric priors of GR with the generative power of diffusion models for 3D low-dose whole-body PET reconstruction. GR-Diffusion employs GR to generate a reference 3D PET image from projection data, establishing a physically grounded and structurally explicit benchmark that overcomes the low-pass limitations of conventional point-based or voxel-based methods. This reference image serves as a dual guide during the diffusion process, ensuring both global consistency and local accuracy. Specifically, we employ a hierarchical guidance mechanism based on the GR reference. Fine-grained guidance leverages differences to refine local details, while coarse-grained guidance uses multi-scale difference maps to correct deviations. This strategy allows the diffusion model to sequentially integrate the strong geometric prior from GR and recover sub-voxel information. Experimental results on the UDPET and Clinical datasets with varying dose levels show that GR-Diffusion outperforms state-of-the-art methods in enhancing 3D whole-body PET image quality and preserving physiological details.

Problem

• PET reconstruction
• noise amplification
• structural blurring
• detail loss
• low-dose imaging

Innovation

• GR-Diffusion
• 3D Gaussian Representation
• Diffusion Model
• PET Reconstruction
• Hierarchical Guidance