Unified Noise-aware Network for Low-count PET Denoising

TL;DR

This paper introduces a unified noise-aware neural network for low-count PET denoising that adapts to varying noise levels, improving image quality and diagnostic reliability in low-dose scans.

Contribution

The proposed UNN combines multiple sub-networks to effectively handle different noise levels without needing multiple models, advancing PET image denoising methods.

Findings

Consistently produces high-quality denoised images across various noise levels.

Outperforms single-noise-level trained networks, especially on extremely low-count data.

Validated on large-scale multi-center datasets.

Abstract

As PET imaging is accompanied by substantial radiation exposure and cancer risk, reducing radiation dose in PET scans is an important topic. However, low-count PET scans often suffer from high image noise, which can negatively impact image quality and diagnostic performance. Recent advances in deep learning have shown great potential for recovering underlying signal from noisy counterparts. However, neural networks trained on a specific noise level cannot be easily generalized to other noise levels due to different noise amplitude and variances. To obtain optimal denoised results, we may need to train multiple networks using data with different noise levels. But this approach may be infeasible in reality due to limited data availability. Denoising dynamic PET images presents additional challenge due to tracer decay and continuously changing noise levels across dynamic frames. To address these issues, we propose a Unified Noise-aware Network (UNN) that combines multiple sub-networks with varying denoising power to generate optimal denoised results regardless of the input noise levels. Evaluated using large-scale data from two medical centers with different vendors, presented results showed that the UNN can consistently produce promising denoised results regardless of input noise levels, and demonstrate superior performance over networks trained on single noise level data, especially for extremely low-count data.