Texture Matching GAN for CT Image Enhancement

TL;DR

This paper introduces TMGAN, a novel GAN-based method that enhances CT images by matching their texture to target textures, improving clinical image quality without compromising anatomical details.

Contribution

The paper presents a texture matching GAN with parallel generators that separate anatomical features from texture, enabling targeted texture enhancement in CT images.

Findings

TMGAN produces higher quality CT images with desirable textures.

The method effectively separates anatomy from texture during training.

Enhanced images are suitable for clinical applications.

Abstract

Deep neural networks (DNN) are commonly used to denoise and sharpen X-ray computed tomography (CT) images with the goal of reducing patient X-ray dosage while maintaining reconstruction quality. However, naive application of DNN-based methods can result in image texture that is undesirable in clinical applications. Alternatively, generative adversarial network (GAN) based methods can produce appropriate texture, but naive application of GANs can introduce inaccurate or even unreal image detail. In this paper, we propose a texture matching generative adversarial network (TMGAN) that enhances CT images while generating an image texture that can be matched to a target texture. We use parallel generators to separate anatomical features from the generated texture, which allows the GAN to be trained to match the desired texture without directly affecting the underlying CT image. We demonstrate that TMGAN generates enhanced image quality while also producing image texture that is desirable for clinical application.