A Study Of Magnetocardiogram Forward Problem Based On Physiological Model

As a non-invasive detection technique,magnetocardiogram is an important method to detect the body surface magnetic field produced by the heart which owns diagnostic value of coronary artery disease,arrhythmia and other heart diseases.In this dissertation,we focus on the forward problem of magnetocardiogram,namely to simulate the distribution of the body surface magnetic field basing on the personal personalized physiological model.It is helpful for further study of the pathogenesis of cardiac pathology and can provide the priori knowledge for the subsequent inverse problem.Meanwhile,the simulation study is ultimately to diagnose the cardiac electrophysiology by the real magnetocardiogram.So,the measurement of real magnetocardiogram is necessary.Thus,this paper also involves some study about measurement.In this paper,the main research work is as follow:(1)We established a personalized physiological model,which is composed of mathematical model and geometric model.As for mathematical model,one is the cardiac source model.The modified diffusion equation is applied to establish the cardiac electrophysiological diffusion model.The other is the cardiac magnetic field model.Based on the quasi-static Maxwell equations,the transmembrane potential(TMP)in the cardiac source model is converted into the cardiac source current density to calculate the surface magnetic field.As for the geometric model,we constructed the individual 3D geometric model based on magnetic resonance imaging(MRI)data of real human body and some relevant image segmentation algorithm and the volume rendering algorithm were applied.(2)Based on the personalized physiological model,the simulation of spatial TMP distribution and the body surface magnetic field were carried out by finite element method on the COMSOL Multiphysics platform.In order to verify the accuracy of finite element method in calculating the forward problem of magnetocardiogram,a simplified cardiac source model and a simplified geometriccardiac magnetic field model were designed.And we figured out the numerical solution and the analytical solution respectively for the error analysis.(3)We implemented the reconstruction of real magnetocardiographic signal.In this paper,the original signal sampled at one single channel of the magnetometer is reconstructed into a magnetocardiographic signal.Further,the sampling period need to be reduced from 5ms to 2ms which raised the time resolution of magnetocardiographic signal to obtain more physiological information.In the future,the two-dimensional real magnetocardiogram map can be drawn by the magnetocardiographic signal,which is rebuilt by the original signals sampling at each single channel.