MAP-Diff: Multi-Anchor Guided Diffusion for Progressive 3D Whole-Body Low-Dose PET Denoising

2026-03-02

AI Summary

This study addresses the challenge of severe noise and quantitative distortion in low-dose PET imaging due to reduced radiation exposure, a problem inadequately resolved by existing methods that struggle to balance reconstruction quality with dose scalability. To overcome this, the authors propose MAP-Diff, a multi-anchor-guided diffusion framework that uniquely incorporates real clinical intermediate-dose scans as anchor points along the diffusion trajectory. By calibrating their timesteps through degradation matching and introducing a timestep-weighted loss, MAP-Diff enables stable and progressive 3D whole-body denoising from ultra-low-dose inputs alone. The method significantly outperforms CNN-, Transformer-, GAN-, and diffusion-based baselines on both internal and cross-scanner datasets, achieving an internal PSNR of 43.71 dB (+1.23 dB), SSIM of 0.986, and NMAE of 0.103, and an external PSNR of 34.42 dB with NMAE of 0.141.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Intelligent Robots: Localization, Mapping, and Navigation

Application Category

• Web Mining and Content Analysis: Content-based information diffusion
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Low-dose Positron Emission Tomography (PET) reduces radiation exposure but suffers from severe noise and quantitative degradation. Diffusion-based denoising models achieve strong final reconstructions, yet their reverse trajectories are typically unconstrained and not aligned with the progressive nature of PET dose formation. We propose MAP-Diff, a multi-anchor guided diffusion framework for progressive 3D whole-body PET denoising. MAP-Diff introduces clinically observed intermediate-dose scans as trajectory anchors and enforces timestep-dependent supervision to regularize the reverse process toward dose-aligned intermediate states. Anchor timesteps are calibrated via degradation matching between simulated diffusion corruption and real multi-dose PET pairs, and a timestep-weighted anchor loss stabilizes stage-wise learning. At inference, the model requires only ultra-low-dose input while enabling progressive, dose-consistent intermediate restoration. Experiments on internal (Siemens Biograph Vision Quadra) and cross-scanner (United Imaging uEXPLORER) datasets show consistent improvements over strong CNN-, Transformer-, GAN-, and diffusion-based baselines. On the internal dataset, MAP-Diff improves PSNR from 42.48 dB to 43.71 dB (+1.23 dB), increases SSIM to 0.986, and reduces NMAE from 0.115 to 0.103 (-0.012) compared to 3D DDPM. Performance gains generalize across scanners, achieving 34.42 dB PSNR and 0.141 NMAE on the external cohort, outperforming all competing methods.

Problem

• Low-dose PET
• Denoising
• Diffusion model
• Progressive reconstruction
• Quantitative degradation

Innovation

• multi-anchor guidance
• progressive denoising
• diffusion model
• low-dose PET
• dose-aligned trajectory