A coarse-to-fine multi-scale feature hybrid low-dose CT denoising network

• Published in: Signal processing. Image communication Vol. 118; p. 117009
• Main Authors: Han, Zefang, Shangguan, Hong, Zhang, Xiong, Cui, Xueying, Wang, Yue
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.10.2023
• Subjects: Generative adversarial networks; Image denoising; Low-dose CT; Multi-resolution inception discriminator; Self-calibration convolution; Self-calibration convolution; Generative adversarial networks; Multi-resolution inception discriminator; Low-dose CT; Image denoising
• ISSN: 0923-5965; 1879-2677
• DOI: 10.1016/j.image.2023.117009

With the growing development and wide clinical application of CT technology, the potential radiation damage to patients has sparked public concern. However, reducing the radiation dose may cause large amounts of noise and artifacts in the reconstructed images, which may affect the accuracy of the clinical diagnosis. Therefore, improving the quality of low-dose CT scans has become a popular research topic. Generative adversarial networks (GAN) have provided new research ideas for low-dose CT (LDCT) denoising. However, utilizing only image decomposition or adding new functional subnetworks cannot effectively fuse the same type of features with different scales (or different types of features). Thus, most current GAN-based denoising networks often suffer from low feature utilization and increased network complexity. To address these problems, we propose a coarse-to-fine multiscale feature hybrid low-dose CT denoising network (CMFHGAN). The generator consists of a global denoising module, local texture feature enhancement module, and self-calibration feature fusion module. The three modules complement each other and guarantee overall denoising performance. In addition, to further improve the denoising performance, we propose a multi-resolution inception discriminator with multiscale feature extraction ability. Experiments were performed on the Mayo and Piglet datasets, and the results showed that the proposed method outperformed the state-of-the-art denoising algorithms. •A coarse-to-fine multi-scale feature hybrid low-dose CT denoising network was proposed.•The problem of LDCT denoising was decomposed into two sub-problems: global feature extraction and local texture feature enhancement.•A self-calibration feature fusion module was designed to expand the receptive field, promote the information exchange between space and channel, and avoid the interference of irrelevant regions.•A shulffe based multi-resolution inception discriminator was designed to realize the multi-scale feature extraction and discrimination.