3D Segmentation Guided Style-based Generative Adversarial Networks for PET Synthesis

TL;DR

This paper introduces SGSGAN, a novel style-based GAN that uses segmentation guidance to improve low-dose PET image synthesis into full-dose images, capturing fine details and regions of interest.

Contribution

The paper proposes a style modulation generator and a task-driven segmentation-guided framework to enhance PET image translation quality.

Findings

Outperforms existing methods in PET synthesis quality

Better preservation of regions of interest

More realistic textures in generated images

Abstract

Potential radioactive hazards in full-dose positron emission tomography (PET) imaging remain a concern, whereas the quality of low-dose images is never desirable for clinical use. So it is of great interest to translate low-dose PET images into full-dose. Previous studies based on deep learning methods usually directly extract hierarchical features for reconstruction. We notice that the importance of each feature is different and they should be weighted dissimilarly so that tiny information can be captured by the neural network. Furthermore, the synthesis on some regions of interest is important in some applications. Here we propose a novel segmentation guided style-based generative adversarial network (SGSGAN) for PET synthesis. (1) We put forward a style-based generator employing style modulation, which specifically controls the hierarchical features in the translation process, to generate images with more realistic textures. (2) We adopt a task-driven strategy that couples a segmentation task with a generative adversarial network (GAN) framework to improve the translation performance. Extensive experiments show the superiority of our overall framework in PET synthesis, especially on those regions of interest.