Multi-Scale Texture Loss for CT denoising with GANs

Francesco Di Feola,Lorenzo Tronchin,Valerio Guarrasi,Paolo Soda

2024-03-25

Abstract:Generative Adversarial Networks (GANs) have proved as a powerful framework for denoising applications in medical imaging. However, GAN-based denoising algorithms still suffer from limitations in capturing complex relationships within the images. In this regard, the loss function plays a crucial role in guiding the image generation process, encompassing how much a synthetic image differs from a real image. To grasp highly complex and non-linear textural relationships in the training process, this work presents a loss function that leverages the intrinsic multi-scale nature of the Gray-Level-Co-occurrence Matrix (GLCM). Although the recent advances in deep learning have demonstrated superior performance in classification and detection tasks, we hypothesize that its information content can be valuable when integrated into GANs' training. To this end, we propose a differentiable implementation of the GLCM suited for gradient-based optimization. Our approach also introduces a self-attention layer that dynamically aggregates the multi-scale texture information extracted from the images. We validate our approach by carrying out extensive experiments in the context of low-dose CT denoising, a challenging application that aims to enhance the quality of noisy CT scans. We utilize three publicly available datasets, including one simulated and two real datasets. The results are promising as compared to other well-established loss functions, being also consistent across three different GAN architectures. The code is available at:

Image and Video Processing,Computer Vision and Pattern Recognition,Machine Learning

What problem does this paper attempt to address?

The paper primarily focuses on addressing the problem of denoising low-dose computed tomography (LDCT) images. Specifically, the research team is concerned with improving the denoising performance by enhancing the loss function in generative adversarial networks (GANs).

In existing GAN-based methods, although significant progress has been made, there are still limitations in capturing complex image relationships. Therefore, the paper proposes a new Multi-Scale Texture Loss Function (MSTLF), which aims to enhance the model's ability to capture complex texture relationships by leveraging the inherent multi-scale nature of the Gray-Level Co-occurrence Matrix (GLCM).

The main contributions of MSTLF include:
1. Proposing a loss function that can extract texture information from different spatial and angular scales, effectively embedding this texture information into the GAN-based denoising algorithm.
2. Introducing a novel aggregation rule based on the self-attention mechanism to effectively and dynamically merge multi-scale texture information.
3. Developing a new GLCM implementation method based on soft assignment, making GLCM differentiable and compatible with gradient optimization.
4. Conducting extensive experimental validation in both paired and unpaired scenarios, and testing the method's effectiveness on three public datasets.
5. Testing the method on three different GAN architectures, demonstrating its generalizability to the backbone structures used.

In summary, the goal of this paper is to improve the quality of low-dose CT images by enhancing the loss function in GANs, particularly by reducing noise while maintaining or improving the clarity of image details and structures.