基于神经网络拟合的腹部化学交换饱和转移成像  

CEST Imaging of the Abdomen with Neural Network Fitting

作　　者：王志超 张记磊 赵羽 华婷 汤光宇[4] 李建奇[1] WANG Zhi-chao;ZHANG Ji-lei;ZHAO Yu;HUA Ting;TANG Guang-yu;LI Jian-qi(Shanghai Key Laboratory of Magnetic Resonance,School of Physics and Electronic Science,East China Normal University,Shanghai 200062,China;Philips Healthcare,Shanghai 200040,China;Institute of Imaging Science,Vanderbilt University,Tennessee 37232,USA;Department of Radiology,Shanghai Tenth People's Hospital,Shanghai 200072,China)

机构地区：[1]华东师范大学物理与电子科学学院,上海市磁共振重点实验室,上海200062 [2]飞利浦医疗公司,上海200040 [3]Institute of Imaging Science,Vanderbilt University,Tennessee 372324 [4]上海市第十人民医院放射科,上海200072

出　　处：《波谱学杂志》2022年第1期33-42,共10页Chinese Journal of Magnetic Resonance

基　　金：华东师范大学“幸福之花”基金资助项目。

摘　　要：磁共振成像(Magnetic Resonance Imaging,MRI)化学交换饱和转移(Chemical Exchange Saturation Transfer,CEST)技术在临床诊断中展现了巨大的潜力,但在腹部成像中受到主磁场偏移量大的挑战,而且利用传统的非对称性分析法得到的酰胺质子转移(Amide Proton Transfer,APT)成像对比度受到核奥氏增强(Nuclear Overhauser Enhancement,NOE)效应的干扰.本文提出了一种基于神经网络拟合的CEST后处理方法,对每个像素采集得到的Z谱特征进行识别,不需要额外序列扫描即可得到背景参考Z谱与主磁场偏移量,用以校正和获得理想的Z谱,并进一步分离得到源自APT效应与NOE效应的信号.鸡蛋清和健康志愿者腹部成像结果显示,本文提出的基于神经网络的CEST后处理方法效果较好.Chemical exchange saturation transfer(CEST)imaging shows great potential in clinical applications.However,CEST imaging is challenging in abdomen due to the large B_(0) shift.Meanwhile,the nuclear Overhauser enhancement(NOE)effect contaminates amide proton transfer(APT)image when using the conventional asymmetry analysis.In this paper,a CEST post-processing approach based on neural network fitting was proposed.Through recognizing the characteristics of acquired Z-spectrum,the background reference Z-spectrum and the B_(0) offset were obtained and used to correct the acquired Z-spectrum.The APT effect and NOE effect could be calculated by subtracting the background reference Z-spectrum from the acquired Z-spectrum.The proposed CEST post-processing approach was validated by the egg white imaging and the abdomen imaging on four healthy volunteers.

关 键 词：磁共振成像(MRI) 化学交换饱和转移(CEST) 酰胺质子转移(APT) 核奥氏增强(NOE) 神经网络 

分 类 号：O482.53[理学—固体物理]