UniPET: a universal network for high-quality PET image denoising across varied dose reduction factors

2026-06-09

AI Summary

Existing PET image denoising methods suffer significant performance degradation under varying dose reduction factors (DRFs) due to limited cross-DRF generalization capability. To address this, this work proposes UniPET, a universal denoising network that introduces domain generalization into PET denoising for the first time. UniPET employs a Style Alignment Network (SAN) to mitigate style collapse in low-dose images and incorporates a Region-Aware Learning Strategy (RALS) to perform adversarial training specifically in texture-critical regions. The method operates within a unified end-to-end framework that eliminates the need for retraining at each specific DRF. Experimental results demonstrate that UniPET achieves state-of-the-art performance in universal denoising, matching or surpassing specialized models across quantitative metrics, visual quality, and clinical evaluations.

Abstract

Most existing deep learning-based PET image denoising methods assume a fixed and known dose reduction factor (DRF) for low-dose PET images. However, these methods encounter significant performance degradation when the DRF varies beyond the assumed one in practical applications. To address the challenge posed by varied DRFs, several preliminary studies focus on the task of universal PET image denoising, aiming to train a universal model over low-dose data across DRFs. Nonetheless, these vanilla universal models often struggle with misaligned styles present in different DRF data, leading to the \textit{style elimination issue} with a significant over-smoothing effect. To deal with this issue, we innovatively introduce domain generalization to PET image denoising and propose a universal PET image denoising network (UniPET) to achieve high-quality PET image denoising across diverse DRFs. UniPET comprises two primary innovations: a style alignment network (SAN) and a region-aware learning strategy (RALS). Specifically, SAN utilizes style alignment techniques derived from domain generalization to align and recover styles across different DRFs, ensuring the model's generalizability across various DRFs while effectively preserving styles. Furthermore, to enhance style recovery, RALS distinguishes between flat and stylized regions, exclusively conducting adversarial learning on the latter, thereby more effectively guiding the model's focus towards learning stylized regions. It is demonstrated that our proposed UniPET can adaptively recover different DRF styles and achieve high-quality PET image denoising across DRFs. Comprehensive experiments show that UniPET exhibits comparable performance to individual DRF-specific models at specific DRFs and realizes state-of-the-art performance in universal PET image denoising quantitatively, perceptually, and clinically.

Problem

• PET image denoising
• dose reduction factor
• universal denoising
• style elimination
• domain generalization

Innovation

• domain generalization
• style alignment
• region-aware learning
• universal denoising
• PET image reconstruction