Parameter-Transferred Wasserstein Generative Adversarial Network (PT-WGAN) for Low-Dose PET Image Denoising

TL;DR

This paper introduces PT-WGAN, a novel transfer learning-based GAN model that effectively denoises low-dose PET images, reducing noise while maintaining image quality, thus enhancing clinical diagnostic performance.

Contribution

The paper proposes a new PT-WGAN framework with task-specific transfer learning for efficient low-dose PET image denoising, outperforming existing methods.

Findings

Superior noise suppression compared to state-of-the-art methods

Preserves structural details and image fidelity

Improves training efficiency through transfer learning

Abstract

Due to the widespread use of positron emission tomography (PET) in clinical practice, the potential risk of PET-associated radiation dose to patients needs to be minimized. However, with the reduction in the radiation dose, the resultant images may suffer from noise and artifacts that compromise diagnostic performance. In this paper, we propose a parameter-transferred Wasserstein generative adversarial network (PT-WGAN) for low-dose PET image denoising. The contributions of this paper are twofold: i) a PT-WGAN framework is designed to denoise low-dose PET images without compromising structural details, and ii) a task-specific initialization based on transfer learning is developed to train PT-WGAN using trainable parameters transferred from a pretrained model, which significantly improves the training efficiency of PT-WGAN. The experimental results on clinical data show that the proposed network can suppress image noise more effectively while preserving better image fidelity than recently published state-of-the-art methods. We make our code available at https://github.com/90n9-yu/PT-WGAN.