Research On Magnetic Detection Electrical Impedance Imaging Algorithm Based On Structure Prior Information

Magnetic Detection Electrical Impedance Tomography(MDEIT)is a new medical imaging technology,which can detect the magnetic induction intensity distribution outside the imaging body by injecting excitation current into the imaging body,and reconstruct the electrical conductivity distribution image inside the imaging body by using imaging algorithm.MDEIT technology,developed from electrical impedance tomography(electrical impedance tomography,EIT)technology,inherits the EIT's advantages of portable and sustainable monitoring and overcome the disadvantages of small measurement information and limited electrode numbers,etc.At the same time,MDEIT detection system is non-contact measurement system,which has prominent advantages and broad application prospects in the dynamic continuous testing and functional imaging.This paper focuses on the inverse problem algorithm in MDEIT.Based on the analysis of the MDEIT positive problem,the Krylov subspace projection method based on sensitivity matrix is used to solve the inverse problem,and square model and pulmonary simulation model are established to verify the inverse problem.For solving the inverse problem,image relative error,image correlation coefficient,reconstructed image heterogeneous centroid and reconstructed image profile are used to comprehensively evaluate the effectiveness of the Krylov subspace projection method from the aspects of reconstructed image similarity,reconstructed heterogeneous location and heterogeneous shape.Secondly,in order to solve the problems of heterogeneous localization of the image reconstructed by Krylov subspace projection,such as deviation and shape edge blur,etc.,the structure prior information was introduced into MDEIT technology,and the MDEIT image reconstruction algorithm based on the structure prior information is further proposed.Based on the simulation experiment,the phantom experiment is designed and carried out to prove the stability of the reconstruction algorithm based on the structure prior information.Four evaluation indexes including image relative error,image correlation coefficient,reconstructed image heterogeneous centroid and reconstructed image profile are used to evaluate the quality of the reconstructed image of two kinds of algorithms in the simulation experiment and the simulation experiment.The results show that the MDEIT image reconstruction algorithm based on the structure prior information can improve the accuracy of the location and shapes of phantoms in reconstructed images.Finally,for obtaining more accurate heterogeneous position information,a CTMDEIT image fusion algorithm combining image information with functional information is proposed.The fusion algorithms commonly used in image fusion are analyzed,and finally a CT-MDEIT image fusion algorithm based on discrete wavelet transform is proposed.The processed CT image was fused with the reconstructed MDEIT image,and the fusion image quality processing by the three methods was accurately evaluated by combining subjective evaluation with objective evaluation.In the research of inverse problem of MDEIT,reconstruction algorithm based on the structure prior information and the CT-MDEIT image fusion algorithm based on discrete wavelet transform are proposed,which can improve the accuracy of the location and shapes of phantoms in reconstructed images,and lay a foundation for the clinical application of MDEIT.