《计算机应用》唯一官方网站 ›› 2022, Vol. 42 ›› Issue (7): 2301-2310.DOI: 10.11772/j.issn.1001-9081.2021040700

• 前沿与综合应用 • 上一篇    

用于低剂量CT降噪的伪影感知生成对抗网络

韩泽芳, 张雄(), 上官宏, 韩兴隆, 韩静, 奉刚, 崔学英   

  1. 太原科技大学 电子信息工程学院,太原 030024
  • 收稿日期:2021-05-06 修回日期:2021-10-08 接受日期:2021-10-12 发布日期:2022-07-15 出版日期:2022-07-10
  • 通讯作者: 张雄
  • 作者简介:韩泽芳(1996—),女,山西晋城人,硕士研究生,主要研究方向:医学图像处理
    上官宏(1988—),女,山西乡宁人,副教授,博士,主要研究方向:模式识别、医学图像处理
    韩兴隆(1995—),男,河南禹州人,硕士研究生,主要研究方向:医学图像处理
    韩静(1998—),女,河北衡水人,硕士研究生,主要研究方向:医学图像处理
    奉刚(1994—),男,四川南充人,硕士研究生,主要研究方向:医学图像处理
    崔学英(1978—),女,山西太原人,副教授,博士,主要研究方向:图像处理与重建。
  • 基金资助:
    国家自然科学基金资助项目(62001321);山西省自然科学基金资助项目(201901D111261);山西省研究生教育创新项目(2020SY417)

Artifacts sensing generative adversarial network for low-dose CT denoising

Zefang HAN, Xiong ZHANG(), Hong SHANGGUAN, Xinglong HAN, Jing HAN, Gang FENG, Xueying CUI   

  1. School of Electronic Information Engineering,Taiyuan University of Science and Technology,Taiyuan Shanxi 030024,China
  • Received:2021-05-06 Revised:2021-10-08 Accepted:2021-10-12 Online:2022-07-15 Published:2022-07-10
  • Contact: Xiong ZHANG
  • About author:HAN Zefang, born in 1996, M. S. candidate. Her research interests include medical image processing.
    SHANGGUAN Hong, born in 1988, Ph. D., associate professor. Her research interests include pattern recognition, medical image processing.
    HAN Xinglong, born in 1995, M. S. candidate. His research interest includes medical image processing.
    HAN Jing, born in 1998, M. S. candidate. Her research interests include medical image processing.
    FENG Gang, born in 1994, M. S. candidate. His research interests include medical image processing.
    CUI Xueying, born in 1978, Ph. D., associate professor. Her research interests include image processing and reconstruction.
  • Supported by:
    National Natural Science Foundation of China(62001321);Natural Science Foundation of Shanxi Province(201901D111261);Graduate Education Innovation Project of Shanxi Province(2020SY417)

摘要:

近年来,生成对抗网络(GAN)用于低剂量CT(LDCT)伪影抑制表现出一定性能优势,已成为该领域新的研究热点。由于伪影分布不规律且与正常组织位置息息相关,现有GAN网络的降噪性能受限。针对上述问题,提出了一种基于伪影感知GAN的LDCT降噪算法。首先,设计了伪影方向感知生成器,该生成器在U型残差编解码结构的基础上增加了伪影方向感知子模块(ADSS),从而提高生成器对伪影方向特征的敏感度;其次,设计了注意力判别器(AttD)来提高对噪声伪影的鉴别能力;最后,设计了与网络功能相对应的损失函数,通过多种损失函数协同作用来提高网络的降噪性能。实验结果表明,与高频敏感GAN(HFSGAN)相比,该降噪算法的平均峰值信噪比(PSNR)和结构相似度(SSIM)分别提升了4.9%和2.8%,伪影抑制效果良好。

关键词: 低剂量断层扫描成像, 图像降噪, 生成对抗网络, 方向卷积, 注意力机制

Abstract:

In recent years, Generative Adversarial Network (GAN) has become a new research hotspot in Low-Dose Computed Tomography (LDCT) artifact suppression because of its performance advantages. Due to irregular distribution and strong relevance to the normal tissues of artifacts, denoising performance of the existing GAN-based denoising networks is limited. Aiming at this problem, a LDCT denoising algorithm based on artifacts sensing GAN was proposed. Firstly, an artifacts direction sensing generator was designed. In this generator, on the basis of U-residual encoding and decoding structure, an Artifacts Direction Sensing Sub-module (ADSS) was added to improve the generator’s sensitivity to artifacts direction features. Secondly, the Attention Discriminator (AttD) was designed to improve the ability of distinguishing noise and artifacts. Finally, the loss functions corresponding to the network functions were designed. Through the cooperation of multiple loss functions, the denoising performance of network was improved. Experimental results show that compared to the High-Frequency Sensitive GAN (HFSGAN), the proposed denoising algorithm has the average Peak Signal-to-Noise Ratio (PSNR) and Structural SIMilarity (SSIM) improved by 4.9% and 2.8% respectively, and has good artifact suppression effect.

Key words: Low-Dose Computed Tomography (LDCT), image denoising, Generative Adversarial Network (GAN), orientation convolution, attention mechanism

中图分类号: