DDPET-3D: Dose-aware Diffusion Model for 3D Ultra Low-dose PET Imaging

TL;DR

This paper introduces DDPET-3D, a novel diffusion model designed for 3D low-dose PET image reconstruction, effectively reducing radiation exposure while maintaining image quality across various noise levels.

Contribution

The paper presents a new dose-aware 3D diffusion model that addresses the challenges of PET image denoising at multiple noise levels, outperforming previous methods.

Findings

Superior performance over previous diffusion models in 3D PET denoising

Effective handling of multiple noise levels in clinical PET images

Validated on 600 patient studies with various dose levels

Abstract

As PET imaging is accompanied by substantial radiation exposure and cancer risk, reducing radiation dose in PET scans is an important topic. Recently, diffusion models have emerged as the new state-of-the-art generative model to generate high-quality samples and have demonstrated strong potential for various tasks in medical imaging. However, it is difficult to extend diffusion models for 3D image reconstructions due to the memory burden. Directly stacking 2D slices together to create 3D image volumes would results in severe inconsistencies between slices. Previous works tried to either apply a penalty term along the z-axis to remove inconsistencies or reconstruct the 3D image volumes with 2 pre-trained perpendicular 2D diffusion models. Nonetheless, these previous methods failed to produce satisfactory results in challenging cases for PET image denoising. In addition to administered dose, the noise levels in PET images are affected by several other factors in clinical settings, e.g. scan time, medical history, patient size, and weight, etc. Therefore, a method to simultaneously denoise PET images with different noise-levels is needed. Here, we proposed a Dose-aware Diffusion model for 3D low-dose PET imaging (DDPET-3D) to address these challenges. We extensively evaluated DDPET-3D on 100 patients with 6 different low-dose levels (a total of 600 testing studies), and demonstrated superior performance over previous diffusion models for 3D imaging problems as well as previous noise-aware medical image denoising models. The code is available at: https://github.com/xxx/xxx.