LeqMod: Adaptable Lesion-Quantification-Consistent Modulation for Deep Learning Low-Count PET Image Denoising

doi:10.1109/TMI.2025.3618247

LeqMod: Adaptable Lesion-Quantification-Consistent Modulation for Deep Learning Low-Count PET Image Denoising

2025-10-06

0

PRE

AI

DOI：10.1109/TMI.2025.3618247

Original

Original request for help

Save

Abstract

En 中文

Deep learning-based positron emission tomography (PET) image denoising offers the potential to reduce radiation exposure and scanning time by transforming low-count images into high-count equivalents. However, existing methods typically blur crucial details, leading to inaccurate lesion quantification. This paper proposes a lesion-perceived and quantification-consistent modulation (LeqMod) strategy for enhanced PET image denoising, via employing downstream lesion quantification analysis as auxiliary tools. The LeqMod is a plug-and-play design adaptable to a wide range of model architectures, modulating the sampling and optimization procedures of model training without adding any computational burden to the inference phase. Specifically, the LeqMod consists of two components, the lesion-perceived modulation (LeMod) and the multiscale quantification-consistent modulation (QuMod). The LeMod enhances lesion contrast and visibility by allocating higher sampling weights and stricter loss criteria to lesion-present samples determined by an auxiliary segmentation network than lesion-absent ones. The QuMod further emphasizes quantification accuracy for both the mean and maximum standardized uptake value (<inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${\mathrm {SUV}}_{{\textit {mean}}}$ </tex-math></inline-formula> and <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${\mathrm {SUV}}_{{\textit {max}}}$ </tex-math></inline-formula>) across multiscale sub-regions throughout the entire image, thereby reducing biases of denoised results relative to high-count references. Experiments conducted on large PET datasets from multiple centers and vendors, and varying noise levels demonstrated the LeqMod efficacy across various denoising frameworks. Compared to frameworks without LeqMod, the integration of LeqMod reduces the lesion <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${\mathrm {SUV}}_{{\textit {max}}}$ </tex-math></inline-formula> bias by 5.92% on average and increases the peak signal-to-noise ratio (PSNR) by 0.36 on average, when denoising images across participating sites. (Code is available at <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/mhxiaaa/LeqMod_PET_denoising</uri>)

Keywords:

Low-count PET image denoising

deep learning

segmentation-assisted denoising

lesion quantification

LeqMod: Adaptable Lesion-Quantification-Consistent Modulation for Deep Learning Low-Count PET Image Denoising

Abstract

Journal

Organization