【作者】 陈波；

【导师】 蒋式勤；

【作者基本信息】 同济大学 ， 控制理论与控制工程， 2007， 硕士

【摘要】 心脏是极其复杂的生理系统。通过检测人体心脏表面磁场信号来诊断心脏疾病是近年来科学技术的新发展。心脏电磁场理论以及心磁信号处理方面的研究受到了很多国家的重视。目前，采用超导量子干涉器检测得到的心脏磁场信号及其微弱，在检测过程中容易受到各种噪声的干扰。除了常规的基于时间的信号滤波需要考虑之外，环境噪声对人体胸腔表面36个测点空间分布的影响也受到了国内外学者的关注，该问题的研究具有一定的理论意义和应用价值。本论文利用人体心脏磁场检测数据，在心脏磁场源模型的基础上，主要研究了两种不同的空间滤波方法。论文完成的主要工作如下：1．在阅读文献的基础上，分析了心磁信号的特点及环境噪声对心磁信号的影响，提出了心脏电磁场检测数据的空间滤波问题。2．建立了基于单磁偶极子模型的空间滤波算法。比较了在理论模型产生的磁场信号上添加不同强度噪声后空间滤波的效果，并通过计算其信噪比与拟合优度，验证了心磁信号空间滤波方法的有效性。3．在单磁偶极子模型的基础上，研究了基于小波变换的空间滤波算法。利用理论图模型产生的心脏磁场信号，比较了三种不同小波变换的空间滤波结果。仿真结果说明7阶Sym小波空间滤波效果较好。4．将单一时刻的空间滤波方法拓展到整个时间域，给出了连续时间心磁测量数据的空间滤波算法。计算机仿真结果验证了该方法的有效性。5．用MATLAB编写了心磁信号空间滤波的程序模块。其中包括单一时刻的空间滤波、连续时间心磁检测信号的空间滤波以及用理论模型产生多通道心磁信号等信号处理与信号分析的功能。6．在验证单磁偶极子数学模型的基础上，用心脏磁场信息分析软件研究了多种心脏磁场源模型的计算和定位，并对计算结果进行了分析。更多还原

【Abstract】 Heart is a complicated physical system. Based on the research of cardiacmagnetism, it has become a new trend of science technology to diagnose heartdiseases via detecting the magnetocardiac signals on the body surface. Manycountries pay more and more attention to the research on cardiac magnetism theoryand signals processing.Nowadays, the magnetocardiac signals detected by SQUID are very weak, andthey can be easily affected by noises during the detected process. Not only regularfiltering methods which are based on time domain but also spatial influence ofenvironmental noises on the 36 measuring points over the thorax need to be involvedinto consideration. The research on magnetocardiac signal’s spatial filter is veryimportant. Based on single magnetic dipole model and the signals detected by SQUID,two different spatial filter methods are deep researched in this paper. The followingstated are the main tasks of this paper:1. Analyze the characters of cardiac magnetic signals and the effect of environmentnoise on the basis of relevant works about spatial filter methods, and put forwardthe question of cardiac magnetic signal’s spatial filter.2. Establish the algorithm of spatial filter on single magnetic dipole model. Estimatedifferent results of spatial filter after adding different magnitude noises on modelmagnetic signals. The SNR(Signal to Noise Ratio) and GOF(Goodness of Fit) arecomputed, and the results prove the validity of spatial filter of cardiac magneticsignals.3. Research the algorithm of spatial filter about wavelet transforms on the base ofsingle magnetic dipole model. Compare the results of three different wavelettransforms’ spatial filter by using the cardiac magnetic signals of theory graphmodel. The result shows wavelet Sym7 is better.4. Expand the single time’s spatial filter methods to the whole time domain. Giveout the arithmetic of continuous spatial filter. This method is proved by a computer simulation.5. The spatial filter module of magnetocardiac signals is programmed in theenvironment of MATLAB GUIDE. It includes spatial filter of single time point,spatial filter of real cardiac magnetic detecting signal and multi-channels cardiacmagnetic signal of theory model.6. Based on proving single magnetic dipole model, kinds of cardiac magnetic sourcemodel are researched with the help of magnetocardiac information analysissoftware.更多还原