Low-dose CT image denoising method based on generative adversarial network  

基于生成对抗网络的低剂量CT图像降噪方法

作　　者：JIAO Fengyuan YANG Zhixiu SHI Shaojie CAO Weiguo 焦枫媛;杨志秀;石韶杰;曹卫国(中北大学生物医学成像与影像大数据山西省重点实验室,山西太原030051;中北大学信息与通信工程学院,山西太原030051;中北大学环境与安全工程学院,山西太原030051)

机构地区：[1]Shanxi Provincial Key Laboratory for Biomedical Imaging and Big Data,North University of China,Taiyuan 030051,China [2]School of Information and Communication Engineering,North University of China,Taiyuan 030051,China [3]School of Environment and Safety Engineering,North University of China,Taiyuan 030051,China

出　　处：《Journal of Measurement Science and Instrumentation》2024年第4期490-498,共9页测试科学与仪器（英文版）

基　　金：supported by National Natural Science Foundation of China(No.11802272);China Postdoctoral Science Foundation(No.2019M651085)。

摘　　要：In order to solve the problems of artifacts and noise in low-dose computed tomography(CT)images in clinical medical diagnosis,an improved image denoising algorithm under the architecture of generative adversarial network(GAN)was proposed.First,a noise model based on style GAN2 was constructed to estimate the real noise distribution,and the noise information similar to the real noise distribution was generated as the experimental noise data set.Then,a network model with encoder-decoder architecture as the core based on GAN idea was constructed,and the network model was trained with the generated noise data set until it reached the optimal value.Finally,the noise and artifacts in low-dose CT images could be removed by inputting low-dose CT images into the denoising network.The experimental results showed that the constructed network model based on GAN architecture improved the utilization rate of noise feature information and the stability of network training,removed image noise and artifacts,and reconstructed image with rich texture and realistic visual effect.为解决临床医学诊断中低剂量计算机断层扫描(Computed tomography,CT)图像存在的伪影和噪声问题,提出了一种改进的生成对抗网络(Generative adversarial network,GAN)架构下的图像降噪算法。首先,构造了一个基于Style GAN2的噪声模型,并通过训练该模型来估计真实噪声分布,从而生成与真实噪声分布相似的噪声信息作为实验噪声数据集。然后,构造了一个基于GAN思想的以编码器-解码器(Encoder-decoder)架构为核心的网络模型,用生成的噪声数据集训练该网络模型直至达到最优。最后,将低剂量CT图像输入该降噪网络,可去除低剂量CT图像中的噪声和伪影。实验结果表明,利用所构造的基于GAN架构的网络模型,可提高噪声特征信息利用率和网络训练稳定性,去除图像噪声和伪影,且重建图像纹理丰富,视觉效果逼真。

关 键 词：low-dose CT image generative adversarial network noise and artifacts encoder-decoder atrous spatial pyramid pooling(ASPP) 

分 类 号：R814[医药卫生—影像医学与核医学] TP391.41[医药卫生—放射医学] TP18[医药卫生—临床医学]