Research On The Modeling Method Of Human Chest Electrical Impedance Imaging

Lung injury is a common thoracic surgical disease that mainly caused by various internal and external factors,such as severe infection,trauma,shock,inhalation of harmful gases,poisoning,etc.Research suggests that lung injury can lead to pulmonary edema,atelectasis and other symptoms.The postoperative monitoring and hospital stay of the patients are prolonged.The severe cases may lead to acute respiratory distress syndrome(ARDS)and even acute respiratory failure.Its mortality rate is as high as 50%-70%.Lung disease is a serious threat to human health.Therefore,real-time monitoring of lung function status is important.Electrical impedance tomography(EIT)is an emerging technology for medical imaging with non-invasive,safe,real-time imaging and other advantages,which is of great significance for the early diagnosis and treatment of lung injury.At present,the study of thorax electrical impedance tomography at home and abroad mostly use circular,elliptical model or the uniform model formed by approximating the shape of the human thorax instead of thorax structure.However,due to the specificity of human thorax contour,measurement errors and imaging errors are introduced by using these unified models.At present,electrical impedance tomography of lung injury is mostly performed by two-dimensional images reconstruction.So that the diagnosis of lung injury in small areas is limited.Aiming at these problems,in this paper,we research on the electrical impedance tomography for lung injury.The EIT models based on the real structure of human thorax are constructed.We perform 3D reconstructed images based on these models.Furthermore,evaluation index of lung injury is optimized based on the reconstruction result.The main work of this paper is as follows:1.For the specificity of human thorax contour,we extracted the prior information of human body structure based on the CT image to optimize 2D thorax forward problem model.An efficient subdivision method for inverse problem of EIT is also proposed.Simulations revealed that the method proposed in this paper can effectively reduce the imaging error.Compared with the traditional modeling methods,the image quality is improved.In addition,conductivity distribution of human thorax is reconstructed based on(Generalized Minimal Residual Algorithm,GMRES)algorithm to improve the imaging quality.2.The true 3D thorax model is constructed based on CT scan sequence.The excitation-measurement multilayer electrode mode is optimized.We study the effects of different excitation methods on imaging results and select the optimal excitation method through simulation experiments.3.The lung models with different severity of lung injury are established.The 3D conductivity distribution of human thorax is reconstructed based on GMRES algorithm.A method for calculating lung injury evaluation index based on 3D images is proposed.Compared with traditional method which is proposed based on 2D reconstruction through simulation experiments,the results show that new method can effectively improve the accuracy of lung injury assessment.