Regularization Algorithm Research Of Electrical Impedance Tomography

Electrical Impedance Tomography(EIT) is a kind of medical imaging technology which targets hunmanbeing’s internal body conductivity distribution. There are lots of advantages belong to this technique such as non-invasive and no damage to the target,the imaging is real-time,the system and equipments are simple and lower price. EIT is one of the fouces in the biomedical imaging in the recent three decades and it should have a bright future.The research on EIT techoonlgy has made great progress, but there are still serious ill-posed in solving its inverse problem. Because of the ill-posed quality of receonstructiong is not that good which boudaries are blurred, resolution is poor and can not meet the clinical medicine application requirements. Aiming at solving this problem, this paper mainly does some research in the reconstruction algorithm of electrical impedance tomography. The main work as follows:Firstly, models the forward problem through deducing the electrical impedance tomography’s mathematical model, and uses finite element method to solve the forward problem.Secondly, introduces the mainstream reconstruction algorithms currently and Tikhonov regularization, studies Newton- Raphson algorithm with Tikhonov regularization in inverse problem solving detaily and does the simulation experiments on Matlab platform.Finally, aiming at solving the low resolution and blurry boundaries in electrical impedance tomography where the traditional Tikhonov regularization is used, a new optimization regularization algorithm is proposed in this work. This algorithm takes the total variation of the normalized measured voltage as the regularization and takes account of the inhomogeneous conductivity’s sparse characteristics in electrical impedance image. The two-step iterative threshold algorithm(TwIST) is considered to solve the conductivity distribution. The computer simulation results showed that the studied regularization algorithm for biomedical imaging application can significantly improve the image quality and boundary resolution.