摘要
PET成像在信号采集过程中存在辐射暴露风险,在保证成像质量的前提下需要尽可能降低示踪剂的使用剂量。这将导致PET图像伪影及信噪比低等问题。本文提出一种基于多尺度上下文信息融合的条件生成对抗网络,用于优化PET图像重建,以提高PET图像的图像质量。通过对作为生成器的编码器-解码器网络的跳过连接进行重新设计,提出一种基于多尺度上下文信息融合的跳过连接,使解码器能够在解码过程中获取来自编码器更加丰富的语义特征。采用使生成器网络学习低剂量PET图像的噪声分布的策略,降低了网络的学习难度。在低剂量PET数据集上对所提出的网络进行评估,峰值信噪比为29.948±4.062,结构相似性系数为0.926±0.030,标准均方根误差为0.395±0.211。相比于传统去噪算法Non-Local Mean和Block-Matching 3D以及2种深度学习方法RED-CNN和以U-Net为生成器的条件生成对抗网络,本文所提出的网络均取得了更加优越的性能表现。
There is a risk of radiation exposure in the signal acquisition process of PET images.Under the premise of ensuring the image quality,it is necessary to reduce the tracer dosage as much as possible.But this will cause artifacts in the PET image and lower signal-to-noise ratio.Propose a conditional generative adversarial network(cGAN)based on multi-scale contextual information fusion,which is used to optimize PET image reconstruction to improve the image quality of PET images.By redesigning the skip connection of encoder-decoder network as generator,proposes a skip connection based on multi-scale context information fusion,which enables the decoder to obtain richer semantic features from the encoder during the decoding process.Adopt the strategy of making the generator network learn the noise distribution of low-dose PET images,which reduces the learning difficulty of the network.The proposed network is evaluated in the low-dose PET dataset.The peak signal-to-noise ratio was 29.948±4.062,the structural similarity index was 0.926±0.030,and the normalized root mean square error was 0.395±0.211.Compared with the traditional denoising algorithm Non-Local Mean(NLM)and Block-Matching 3 D(BM3 D)and two deep learning methods RED-CNN and cGAN(U-Net is the generator),the proposed network is superior to the above four methods.
作者
杨昆
杜瑀
钱武侠
薛林雁
刘琨
卢闫晔
Yang Kun;Du Yu;Qian Wuxia;Xue Linyan;Liu Kun;Lu Yanye(College of Quality and Technical Supervision,Hebei University,Baoding 071002,China;National&Local Joint Engineering Research Center of Metrology Instrument and System,Baoding 071002,China;Department of Biomedical Engineering,Peking University,Beijing 100871,China)
出处
《电子测量技术》
北大核心
2021年第7期74-81,共8页
Electronic Measurement Technology
基金
河北省自然科学基金(H2019201378)
基于深度学习的低剂量PET图像增强算法研究(XZJJ201917)项目资助
关键词
图像去噪
低剂量PET图像
条件生成对抗网络
深度学习
image denoising
low dose PET image
conditional generative adversarial network
deep learning
