Application Study Of Modified Differential Evolution Algorithm In Brain Lesion Impedance Tomography

Electrical impedance tomography(EIT)is a non-invasive imaging technique without ionizing radiation.Because of its portability,relatively low cost,and the safety of processing,EIT has extremely high value of theoretical research and practical application.However,EIT image reconstruction is an ill-posed,non-linear inverse problem controlled by Poisson's equation.Traditional image reconstruction algorithm needs to solve nonlinear differential equations,which leads to a large amount of computation and low image reconstruction accuracy.Based on the differential evolution(DE)algorithm and its improved form,this paper makes a deep study of 2D and 3D EIT image reconstruction of brain lesions.The main study contents are as follows:1?Three are three random parameters in the differential evolution algorithm,which are population size(NP)? scaling factor(F)and crossover factor(CR).The value of these parameters plays an important role in the performance of EIT image reconstruction.However,the current parameter selection of differential evolution algorithm is random to some extent,and most of the study of parameter are carried out through the standard test functions,which are not applied to a specific field.To solve these problems,a large number of simulation experiments of the head EIT image reconstruction are performed based on the given objective function and termination conditions,and the influence of each parameter on image reconstruction is analyzed,and the reasonable selection interval of these parameters is given,which provides an effective basis for the application of differential evolution algorithm in EIT image reconstruction.2?Aiming at the characteristics of DE algorithm and two-dimensional EIT image reconstruction,a double mode adaptive differential evolution algorithm(DSaDE)is proposed after the advantages of mutation strategy DE/rand/1/bin and DE/best/1/bin are studied in detail.In the iterative process,one of the two mutationstrategies is alternately selected by certain rules to generate new population individuals,which improves the convergence speed and global search efficiency of the algorithm.In addition,aiming at the problem that it is difficult to obtain the combination of control parameters in DE algorithm,the adaptive adjustment of control parameters is adopted to reduce the sensitivity of the algorithm to the parameters,and to improve the adaptability and solving ability of the algorithm.The contrast experiments of EIT image reconstruction are carried out on three-layer concentric circle model.Compared with other algorithms,the average convergence algebra of DSaDE algorithm is reduced by more than 25%,and the total average error is reduced by more than 35%.The results show that the proposed DSaDE algorithm can significantly improve the convergence speed of the algorithm and the accuracy of image reconstruction,and reduce the error of image reconstruction.3?Aiming at the problem that the basic differential evolution algorithm has a slow convergence rate and may fall into a local optimum,a modified differential evolution(MDE)algorithm is proposed to solve the inverse problem.By introducing random optimal mutation strategy and local enhancement operator,this algorithm can enhance the local search efficiency,and achieve a relative balance between the diversity of population and the convergence speed of algorithm.In order to verify the effectiveness of the algorithm,the impedance reconstruction simulation experiment is carried out on three-layer concentric sphere model and compared with other algorithms.According to the simulation results,the modified algorithm is simple and can reconstruct the impedance image more clearly than other algorithms,and the error of image reconstruction is reduced by more than 30%,so our algorithm can effectively improve the accuracy of image reconstruction.