Research On Ultrasound Modulated Electrical Impedance Tomography

Aiming at the ill-posed inverse problem and low-resolution reconstructed image of traditional electrical impedance tomography(EIT),based on acousto-electric(AE)effect,ultrasound modulated electrical impedance tomography(UMEIT)is proposed,which utilizes coupling between electric and acoustic modulates.As a new hybrid imaging technique,UMEIT combines the contrast advantage of EIT and resolution advantage of ultrasound imaging.With more additional effective measurement information,UMEIT can reduce the ill-posedness of EIT inverse problem and improve the spatial resolution of reconstructed image.The main work and results are as follows.(1)Aiming at the ill-condition of EIT image reconstruction,a hybrid regularization method combing Tikhonov and total variation and a spatial adaptive total variation regularization method are proposed respectively.With adaptive weighted parameter adjusting the weighting between Tikhonov regularization term and total variation regularization term,the hybrid regularization method has been investigated to get better reconstructed images.With difference curvature adopted as spatial feature indicator,the effectiveness of the spatial adaptive total variation regularization method has been validated.The modification of total variation regularization method,on the one hand,improves the performance of image reconstruction method.On the other hand,this process accomplishes the image reconstruction of single physics field,and lays a good foundation for multi-physics field image reconstruction.(2)Aiming at the issue of tissue AE effect mechanism,tissue AE effect is modeled using the theory of solid mechanics.Based on solid mechanics theory,both bulk compression and thermal expansion of tissue are considered and discussed.Ultrasonic pressure produces the volume change of focal region and leads to the conductivity change.In addition,when ultrasound passes through the tissue,some of the energy of acoustic field is absorbed by the tissue,leading to tissue thermal expansion and conductivity change.From those two aspects,the influence of tissue electrical feature and ultrasonic feature on AE effect has been investigated,and the tissue AE effect is modeled.(3)Aiming at the limitation of exciting power density formula,a formula relating the power density to the measured data is derived again with a more reasonable expression.In additional,from an electrical point of view,the relationship between power density distribution and conductivity distribution is investigated,which enhances the understanding of power density distribution characteristics and provides a theoretical foundation for further investigation on UMEIT image reconstruction.(4)Based on power density,a linearized image reconstruction method is proposed for UMEIT.Relating the power density change to the change in conductivity,Jacobian matrix is employed to make the nonlinear relationship into a linear one.The analytic formulation of this Jacobian matrix is derived and its effectiveness is also verified.In additional,three current excitation patterns: adjacent excitation,vertical excitation and opposite excitation,are considered and tested.Compared with the other two excitation patterns,opposite excitation is of the best performance with the proposed method and recommended to be the first choice for other image reconstruction methods based on power density.(5)A new image reconstruction method based on boundary voltage is proposed for UMEIT.According to optimization method,an objective function of image reconstruction has been directly constructed using boundary voltage of imaging object.Then the conductivity reconstruction is implemented from the boundary voltages.Image reconstruction results show that UMEIT has a faster convergence rate and higher image quality than the conventional EIT,and it is more robust with respect to measurement noise.