Research On Image Reconstruction Of Electrical Impedance Tomography Based On Quantum-behaved Particle Swarm Optimization Algorithm

Electrical Impedance Tomography (EIT) is a new noninvasive medical functionalimaging technology. EIT uses electrode matrix placed on human surface to injectexcitation currents into human body and measures the electrical response signal, andthen gets the electrical characters of human tissues that are related to human body’sphysiological and pathological situation. Compared with the traditional medicalimaging techniques, EIT doesn’t use the nuclide and radial, so it is noninvasive tohuman body and can be repeatedly used in short time. Meanwhile, the EIT equipment isrelatively affordable, and can be operated just in normal environment, so it has a hugemarket. With these advanced characters, EIT technology is considered very attractiveand competitive in the medical field, industrial inspection, nondestructive testing andother industrial sectors.This dissertation aims at EIT image reconstruction problem, by realizing the finiteelement method based on the complete electrode model to compute EIT forwardproblem and studying the research difficulties and pathological of EIT inverse problem.After introducing the advantages and disadvantages of particle swarm optimizationalgorithm and quantum-behaved particle swarm optimization algorithm, QPSOalgorithm is selected to reconstruct EIT image, and amounts of simulations are carriedout to validate the feasibility of the algorithm, and amounts of simulations are carriedout to validate the feasibility of the algorithm.The main work of this dissertation is as follows:①Firstly, based on the biological medicine, the control equation and boundaryconditions of EIT inverse problem are introduced, the finite element method based onthe complete electrode model is inducted. The accuracy of the model is validated bycomparing the measured data and data. Taking into account of accuracy and speed, thefinite element triangulation model of EIT problem and encryption strategy are given.Focusing on the research difficulties and pathological of EIT inverse problem, theanalysis is carried out to lay a good foundation for EIT image reconstruction.②After analyzing the principles, chart-flows of basic particle swarm optimizationalgorithm and several modified PSO algorithm, as well as advantages and disadvantagesof these algorithms, the principle of the quantum-behaved particle swarm optimizationalgorithm is introduced and compared with PSO algorithm. Finally, two classic standard functions are selected to simulate the PSO algorithm and QPSO algorithm, theconclusion show that QPSO algorithm is better than classic PSO algorithm in accuracyand speed and it is laid a good foundation for the improvement of QPSO algorithm andapplication of EIT image reconstruction.③This paper propose the mathematical optimization model of EIT inverseproblem, and choose particle swarm optimization algorithm to realize the imagereconstruction question. In order to enable the search of QPSO to have direction, we useresult of NOSER as initial population. The simulation test is carried out to validate thealgorithm in a closed EIT region and compared with Newton algorithm and PSOalgorithm; EIT imaging reconstruction and test are carried out to combine the practicalapplication, the results show that reconstructed image has a good resolution. Thesoftware platform of OEIT system was established and perfected, which hasmulti-functions including display of FEM, data and image, serial communication,computation of forward and inverse problem, etc.The final part is the summary description, the main research work were listed.Meanwhile, the current disadvantages and the correlative research prospect wereindicated.