Research On Lung Electrical Impedance Tomography Based On Generalized Finite Element Method

Electrical Impedance Tomography(EIT)is a non-destructive,non-radiative,functional medical imaging technique with the capability of images monitoring.EIT can provide important information related with the physiology and pathology variations of organs and tissue.Lung electrical impedance tomography is fit to monitor the lung's ventilation function on patients with respiratory failure in ICU.It can detect the property of the tissue in the lung and can give alarm when the alveolar excessive expansion,pulmonary collapse,pulmonary edema or emphysema occurs during mechanical ventilation.It also can provide specific information for the parameter optimization strategy of mechanical ventilation therapy.It may help to avoid lung injury during mechanical ventilation.The application of EIT in electrical property information detection and imaging of the lungs,can contribute to the real-time monitoring of pulmonary ventilation function.It also may help to identify,diagnoses and treat the related lung disease as soon as possible.This work is supported by the National,Hebei Provincial and Tianjin Provincial Natural Science Foundations of China.In this thesis,the numerical model of electrical impedance tomography,image reconstruction algorithm and hardware system design,etc.are intensively and systematically studied.The main research work of the dissertation is as follows:1.To solve the problem of low numerical accuracy in the EIT forward solver based on traditional Finite Element Method(FEM),the Generalized Finite Element Method(GFEM)is used in the modeling of EIT forward problem in the thesis.Both 2-D and 3-D lung's models were adopted in the calculation and simulation of the EIT forward problem.The accuracy of the EIT forward solver was improved with the generalized finite element method.2.The influence of field deformation on the EIT forward problem was studied.The respiratory movement and posture change can cause field deformation in lung EIT.The numerical models based both conformal and non-conformal transformation were formed and were used in the simulation of the potential distribution on the boundary electrodes with the adjacent driving adjacent measurement protocol.The potential change on the boundary electrode caused by the field deformation is analyzed,which can lays the foundation for accurate reconstruction of the lung's EIT images.3.In order to overcome the limitations,such as the presence of excessive smoothing or step artifacts of the existing EIT image reconstruction methods,the High-order Variation Regularization Algorithm(HOVRA)is adopted in constructing the objective function of the two order generalized variation regularization method,and the difference curvature is used as the basis for the selection of the regularization parameter in different regions during the reconstruction process.In the reconstructed image based on HOVRA,the sharp change of the edge area of the lung was maintained,and the step artifacts of the smooth region were reduced,and the reconstructed image is obtained with high quality.4.In this thesis,we outline a method to combine EIT with mechanical ventilation in order to generate clinically meaningful parameters by means of analogy analysis.The mapping relationship between the EIT conductivity and the gas volume fraction in the respiratory system were determined with the single,double and multi-compartment modeling.Then one can get the volume and respiratory flow parameters of the lung based on the proposed compartment models with numerical simulation.The relationship between the reconstructed conductivity in EIT and lung volume was obtained which is expected to provide meaningful parameters for clinical pulmonary physiology.5.A lung EIT hardware system was designed and accomplished with modular configuration.Both the physical and in vivo EIT imaging experiments were conducted using the designed EIT hardware system with both phantom models and during an imitation of protective mechanical ventilation therapy on adult subjects.The images were reconstructed with the proposed high-order regularization variation algorithm.The results showed that the objects in the phantom model can be localized by the designed EIT system.From the point of view of physical experiment,the feasibility and practicability of the proposed numerical model,reconstruction algorithm and hardware system were verified.The results showed that the impedance information of the human thorax during respiration could be acquired safely and efficiently with the EIT system designed.And the capability of the EIT system for functional imaging was realized.