Cavity Space Detection System And Adaptive Modeling Design

There are no symptoms or obvious symptoms in the early stage of intestinal diseases,and the existing diagnosis methods are limited,so it can not be accurately judged.Because the electrical conductivity and other information of the pathological area are different from that of the normal area,electrical impedance tomography(EIT)can be used to determine the location of the pathological area,and to obtain the size and depth of the pathological tissue under certain circumstances,so as to provide assistance for diagnosis.In this thesis,we use this technique to study the lumen tissue(such as colorectal,esophagus).The main work is as follows:1.According to the common structure of intestinal cavity,the corresponding model is constructed and simulated.Firstly,the mathematical model of the positive problem is introduced through the connection between the positive and the negative problems of EIT,and the formula of solving the voltage distribution is deduced by tetrahedral subdivision unit.Then,three kinds of cavity models are established by COMSOL software according to the relevant data and structure of colorectal.Through the pressure simulation of the finite element model of the cavity,it shows that the sensor can detect whether the electrode is in close contact with the inner wall of the cavity.Then,the part with different conductivity is set on the cavity model to represent the pathological area,and the current excitation simulation is carried out with the normal cavity model.The results show that it is correct to use EIT technology to detect the intestinal tract and establish different structural and morphological models for different intestinal structures.2.Using BP neural network optimized by firefly algorithm for image reconstruction.In this thesis,some non intelligent EIT reconstruction algorithms and intelligent EIT reconstruction algorithms are analyzed and compared,and a BP neural network optimized by firefly algorithm for image reconstruction in this study is proposed.In this thesis,the principle of Optimizing BP neural network by firefly algorithm and the determination of the number of hidden layer nodes are described.Then we use the algorithm,traditional BP neural network and DNN with three hidden layers to reconstruct the image,and compare the results.Results: this algorithm is superior to the traditional BP neural network and DNN with three hidden layers in the area boundary and size of lesions.In order to determine the location of the lesion area and other information,we need to observe it in the three-dimensional image,so three-dimensional reconstruction is carried out on the reconstruction results of three algorithms,and the results are analyzed.The analysis shows that the 3D reconstruction image based on the algorithm in this thesis can be used to observe the information of the lesion area,and it is basically consistent with the location and size of the lesion area in the lumen model.3.In order to verify the idea and algorithm of this thesis,cavity model experiment and agar model experiment are designed.The results of cavity model experiment show that the resistance of colorectal cavity model is negatively related to the current frequency,with the maximum resistance at 50 Hz and the minimum resistance at 5MHz;when the current frequency is less than 2.7k Hz,the resistance first changes slowly with the frequency and then changes violently;when the current frequency is greater than 2.7 k Hz,the curve of resistance changes slowly with frequency,which is similar to that measured by intestinal tissue in vitro.Therefore,in the subsequent experimental measurement,the frequency range of the excitation signal can be set to 2.7k Hz ? 5MHz.In order to verify the algorithm,the hardware design and overall control flow of the cavity space detection system are introduced,and then the agar model with foreign matters is made for agar model experiment.After collecting the response voltage sequence,the impedance detection system is sent to the trained network in this thesis to train and then reconstruct the image.Five cross-section reconstruction images are obtained,and then the 3D reconstruction image is reconstructed by MATLAB.The 3D reconstruction image can clearly reflect the location of the foreign object.It is proved that EIT technology can be used to detect the abnormal parts of the cavity tissue.