Electromagnetic tomography (EMT): image reconstruction based on the inverse problem

作     者：XIONG Hanliang , HE Min, LIU Ze and XU Ling'an (1. School of Automation, Tianjin University, Tianjin 300072, China 2. College of Engineering, Shanghai Maritime University, Shanghai 200135, China 3. Department of Automatic Control, Nothern Jiaotong University, Beijing 100044, China) 

作者机构：School of Automation Tianjin University Tianjin 300072 China College of Engineering Shanghai Maritime University Shanghai 200135 China Department of Automatic Control Nothern Jiaotong University Beijing 100044 China 

出 版 物：《Progress in Natural Science:Materials International》 (自然科学进展·国际材料(英文))

年 卷 期：2003年第6期

页      面：72-76页

学科分类：0831[工学-生物医学工程（可授工学、理学、医学学位）] 100207[医学-影像医学与核医学] 1006[医学-中西医结合] 1002[医学-临床医学] 1001[医学-基础医学(可授医学、理学学位)] 08[工学] 1010[医学-医学技术（可授医学、理学学位）] 100106[医学-放射医学] 100602[医学-中西医结合临床] 10[医学] 

基　　金：the National Natural Science Foundation of China (Grant No.59877018) 

主　　题：electromagnetic theory inverse problem tomography image reconstruction. 

摘      要：Starting from Maxwell s equations for inhomogeneous media, nonlinear integral equations of the inverse problem of the electromagnetic tomography (EMT) are derived, whose kernel is the dyadic Green s function for the EMT sensor with a homogeneous medium in the object space. Then in terms of ill-posedness of the inverse problem, a Tikhonov-type regularization model is established based on a linearization-approximation of the nonlinear inverse problem. Finally, an iterative algorithm of image reconstruction based on the inverse problem and reconstruction images of some object flows for simplified sensor are given. Initial results of the image reconstruction show that the algorithm based on the inverse problem is superior to those based on the linear back-projection in the quality of image reconstruction .