《计算机应用》唯一官方网站 ›› 2025, Vol. 45 ›› Issue (2): 624-632.DOI: 10.11772/j.issn.1001-9081.2024010039

• 多媒体计算与计算机仿真 • 上一篇    

基于双编码器双解码器GAN的低剂量CT降噪模型

上官宏1,2(), 任慧莹1, 张雄1, 韩兴隆1, 桂志国2, 王燕玲2,3   

  1. 1.太原科技大学 电子信息工程学院,太原 030024
    2.动态测试技术省部共建国家重点实验室(中北大学),太原 030051
    3.山西财经大学 信息学院,太原 030006
  • 收稿日期:2024-01-16 修回日期:2024-04-13 接受日期:2024-04-15 发布日期:2024-05-09 出版日期:2025-02-10
  • 通讯作者: 上官宏
  • 作者简介:任慧莹(1998—),女,黑龙江齐齐哈尔人,硕士研究生,主要研究方向:医学图像处理、深度学习
    张雄(1973—),男,山西神池人,教授,硕士,主要研究方向:模式识别、医学图像处理、视频目标跟踪
    韩兴隆(1995—),男,河南禹州人,硕士,主要研究方向:医学图像处理、深度学习
    桂志国(1972—),男,天津蓟州人,教授,博士,主要研究方向:图像处理、图像重建
    王燕玲(1981—),女,山西吕梁人,讲师,博士,主要研究方向:图像处理、数据分析。
  • 基金资助:
    国家自然科学基金资助项目(62001321);山西省基础研究计划项目(202103021224265);太原科技大学研究生教育创新项目(SY2022015)

Low-dose CT denoising model based on dual encoder-decoder generative adversarial network

Hong SHANGGUAN1,2(), Huiying REN1, Xiong ZHANG1, Xinglong HAN1, Zhiguo GUI2, Yanling WANG2,3   

  1. 1.School of Electronic Information Engineering,Taiyuan University of Science and Technology,Taiyuan Shanxi 030024,China
    2.State Key Laboratory of Dynamic Measurement Technology (North University of China),Taiyuan Shanxi 030051,China
    3.School of Information,Shanxi University of Finance and Economics,Taiyuan Shanxi 030006,China
  • Received:2024-01-16 Revised:2024-04-13 Accepted:2024-04-15 Online:2024-05-09 Published:2025-02-10
  • Contact: Hong SHANGGUAN
  • About author:REN Huiying, born in 1998, M. S. candidate. Her research interests include medical image processing, deep learning.
    ZHANG Xiong, born in 1973, M. S., professor. His research interests include pattern recognition, medical image processing, video target tracking.
    HAN Xinglong, born in 1995, M. S. His research interests include medical image processing, deep learning.
    GUI Zhiguo, born in 1972, Ph. D., professor. His research interests include image processing, image reconstruction.
    WANG Yanling, born in 1981, Ph. D., lecturer. Her research interests include image processing, data analysis.
  • Supported by:
    National Natural Science Foundation of China(62001321);Fundamental Research Program of Shanxi Province(202103021224265);Graduate Education Innovation Project of Taiyuan University of Science and Technology(SY2022015)

摘要:

近年来,生成对抗网络(GAN)用于低剂量计算机断层成像(LDCT)图像降噪已经表现出显著的性能优势,成为该领域的研究热点。然而,GAN的生成器对LDCT图像中噪声和伪影分布的感知能力不足,导致网络的降噪性能受限。因此,提出一种基于双编码器双解码器生成对抗网络(DualED-GAN)的低剂量CT降噪模型。首先,提出由一对编解码器构成伪影像素级特征提取通道,用于估计LDCT中的伪影噪声;其次,提出由另外一对编解码器构成伪影掩码信息提取通道,用于估计伪影的强度和位置信息;最后,采用伪影图像质量标签图辅助估计伪影的掩码信息,可以为伪影像素级特征提取通道提供补充特征,进而提高GAN降噪网络对伪影噪声分布强度的敏感性。实验结果表明,在mayo测试集上与次优模型DESD-GAN(Dual-Encoder-Single-Decoder based Generative Adversarial Network)相比,所提模型的平均峰值信噪比(PSNR)提高了0.338 7 dB,平均结构相似性度(SSIM)提高了0.002 8。可见,所提模型在伪影抑制、结构保留与模型鲁棒性方面均有更好的表现。

关键词: 低剂量计算机断层成像, 生成对抗网络, 编码器, 解码器, 降噪

Abstract:

In recent years, Generative Adversarial Network (GAN) used for Low-Dose Computed Tomography (LDCT) image denoising has shown significant performance advantages, becoming a hot topic in the field. However, the insufficient perception ability of GAN generator for the noise and artifact distribution in LDCT images leads to limit the denoising performance. To address this issue, an LDCT denoising model based on a Dual Encoder-Decoder GAN (DualED-GAN) was proposed. Firstly, a pair of encoder-decoder was proposed to form an artifact pixel-level feature extraction channel for estimating the artifact noise in LDCT images. Secondly, another pair of encoder-decoder was proposed to form an artifact mask information extraction channel for estimating the intensity and location information of artifacts. Finally, the artifact image quality label maps were used to assist in estimating the mask information of artifacts, so that supplementary features were provided for the artifact pixel-level feature extraction channel, thereby enhancing the sensitivity of the GAN denoising network to the distribution intensity of artifact noise. Experimental results show that compared with the sub-optimal model DESD-GAN(Dual-Encoder-Single-Decoder based Generative Adversarial Network), the proposed model increases the average Peak Signal-to-Noise Ratio (PSNR) by 0.338 7 dB, and the average Structural Similarity Index Measure (SSIM) by 0.002 8 on mayo test set. It can be seen that the proposed model performs better in all terms of artifact suppression, structural preservation, and model robustness.

Key words: Low-Dose Computed Tomography (LDCT), Generative Adversarial Network (GAN), encoder, decoder, denoising

中图分类号: