磁探测电阻抗成像的若干关键技术研究

发布时间：2018-11-09 09:40

【摘要】：生物体可以看作是由具有不同电特性并且按照一定空间位置分布的多种生物组织所构成的混合导体，生物组织的空间电导率分布可以提供解剖结构、功能活动和生理病理状态的有用信息。电阻抗成像技术以生物体内电阻抗的分布或变化为成像目标体，是一种集形态、结构和功能成像于一体的新一代无创生物医学检测与成像技术。它通过对成像目标体外加一定的安全激励电流，测量成像目标体表面的边界电压，从而重构得到其内部的阻抗分布。然而电阻抗成像通过电极以接触的方式测量边界电压，不仅获得的测量信息量小，并且测量精度还受到许多干扰因素的影响，例如皮肤表面湿度与粗糙程度以及测量时所施加的压力等。磁探测电阻抗成像通过表面电极向成像目标体注入激励电流，利用磁场传感器非接触测量成像目标体周围的磁场，然后根据磁场数据重构得到电阻抗分布图像。磁探测电阻抗成像作为电阻抗成像的一个分支，采用非接触测量的方式测量成像目标体周围的磁场来重建其内部电阻抗分布，磁场传感器的定位精度高，并且测量传感器不与成像目标体接触。磁探测电阻抗成像技术克服了电阻抗成像接触式测量的缺点，又结合了电阻抗成像的廉价、无创和连续图像监护等诸多优点。本文致力于磁探测电阻抗成像的研究，探讨了正问题、逆问题、测量方式优化和数据采集等若干关键问题，为磁探测电阻抗成像技术的发展奠定了基础。首先描述了磁探测电阻抗成像的正问题，并且利用有限元方法和已知的电流边界条件求解得到目标体内部的电压和电流密度分布，然后根据Biot-Savart定律获得目标体外部的磁感应强度数据。在此基础上，比较了各种电极模式下的正问题计算结果，并且提出了环形电极模式。环形电极避免了电流的扩散效应，使得磁探测电阻抗成像简化成磁探测电流密度成像，并且缩短了测量时间和图像重建时间，为快速成像奠定了基础。介绍了磁探测电阻抗成像的电流密度类算法，磁通密度类算法和差分成像方法。针对磁探测电阻抗成像逆问题的不适定性，提出基于全变差正则化的图像重建算法和适用于稀疏图像重建的全变差正则化和1范数正则化的混合正则化算法。与常用的2范数正则化算法相比，全变差正则化不仅改善了磁探测电阻抗成像逆问题的病态性，还具有良好的保边缘性，使得重构图像的介质之间的边界更加清晰，，提高了重建图像的质量。混合正则化算法使得重建图像不仅具有清晰的边界，而且保证了稀疏特性，更好地实现了目标的定位。仿真结果表明了提出的正则化算法的有效性，推动了磁探测电阻抗成像技术的研究和发展。激励电极的大小、位置、数量和测量点的位置、数量等测量配置参数会影响磁探测电阻抗成像的质量。首先利用奇异值分析方法对不同测量配置参数进行了评估和优化，然后利用减少冗余法对每个测量点提供的信息量进行测量和评估，去掉提高较少信息量的冗余测量点，在减少测量点的情况下几乎不损失有用信息，从而对测量配置进行了进一步的优化。图像重建实验验证了奇异值分析和减少冗余法的有效性。结果表明适当地增加激励电极的数量，测量点分布的圆周数和测量点数目都可以增加有效奇异值个数，另外将激励电极尽量靠近感兴趣区域也可以增加有效奇异值个数。奇异值分析和减少冗余法是有效地寻找最优测量配置的方法，为磁探测电阻抗成像实验设计提供了有效而可靠的工具。建立了一套基于数据采集卡的磁探测电阻抗成像数据采集系统，包括恒流激励源、磁场传感器、后续信号放大和滤波电路、信号采集及控制单元和机械扫描装置等，并且利用LabVIEW设计了基于多测量点的自动数据采集和保存系统。在此基础上以离散模型、生理盐水模型和琼脂模型为实验对象，采用环形电极和横向电流加载模式进行了相关实验，验证了磁探测电阻抗成像的可行性，将磁探测电阻抗成像向前推进了一步。
[Abstract]:The organism can be regarded as a mixed conductor composed of a plurality of biological tissues with different electrical characteristics and distributed according to a certain spatial position, and the spatial conductivity distribution of the biological tissue can provide useful information of the anatomical structure, the functional activity and the physiological and pathological state. The electrical impedance imaging technology is a new generation of non-invasive biomedical detection and imaging technology integrating form, structure and function. and the boundary voltage of the surface of the imaging target body is measured by adding a certain safety excitation current to the imaging target body, so that the impedance distribution inside the imaging target body is reconstructed. However, the electrical impedance imaging measures the boundary voltage in such a way that the electrode is in contact, not only the amount of measurement information obtained is small, and the measurement accuracy is also affected by many interference factors, such as the surface humidity and roughness of the skin, and the pressure applied at the time of measurement, and the like. the magnetic detection electric impedance imaging is used for injecting the excitation current to the imaging target body through the surface electrode, the magnetic detection electro-impedance imaging is used as a branch of the electrical impedance imaging, the magnetic field around the imaging target body is measured in a non-contact measurement way to reconstruct the internal electrical impedance distribution, the positioning accuracy of the magnetic field sensor is high, and the measurement sensor is not in contact with the imaging target body The magnetic detection electro-impedance imaging technology overcomes the shortcomings of the contact measurement of the electrical impedance imaging, and also combines the low-cost, non-invasive and continuous image monitoring of the electrical impedance imaging. This paper is devoted to the research of the electrical impedance imaging of magnetic detection, and probes into some key problems, such as positive and inverse problems, optimization of measurement methods and data acquisition, and lays a foundation for the development of the magnetic detection electric impedance imaging technology. Based on the method of finite element method and known current boundary condition, the voltage and current density distribution inside the object body are obtained, and then the magnetic induction outside the object body is obtained according to the Biot-Savart law. Based on this, the calculation of positive problems in various electrode modes is compared, and the ring is put forward. The ring-shaped electrode avoids the diffusion effect of the current, so that the magnetic detection electric impedance imaging is simplified into the magnetic detection current density imaging, the measurement time and the image reconstruction time are shortened, The current density class of the magnetic detection electric impedance imaging is introduced, and the flux density class is calculated. An image reconstruction algorithm based on total variation regularization and a total variation regularized and 1-norm regularization for sparse image reconstruction are proposed for the discomforts of the inverse problem of the magnetic detection electrical impedance imaging. Compared with the conventional two-norm regularization algorithm, the total variation regularization not only improves the ill-condition of the inverse problem of the magnetic detection electro-impedance imaging, but also has good edge-preserving property, so that the boundary between the medium of the reconstructed image is more clear and improved, The quality of the reconstructed image is reconstructed. The hybrid regularization algorithm makes the reconstructed image not only have a clear boundary, but also the sparse properties, and better The simulation results show that the proposed regularization algorithm is effective, and the magnetic detection electro-impedance imaging is promoted. Research and development of technology. The size, position, number of excitation electrodes and the location and quantity of measuring points will affect the measurement and configuration parameters. The method of singular value analysis is used to evaluate and optimize the different measurement configuration parameters, then the information quantity provided by each measurement point is measured and evaluated by the method of reducing the redundancy, and the improvement is removed. a redundant measurement point with less information, with little loss of useful information in the event of a reduction in the measurement point, The configuration has been further optimized. The image reconstruction experiment verifies the singular value. The results show that the number of effective singular values can be increased by appropriately increasing the number of excitation electrodes, the number of the circumference of the distribution of the measuring points and the number of measuring points can increase the number of effective singular values, and the excitation electrodes are as close as possible to the region of interest. The number of effective singular values can be increased. The singular value analysis and the reduction of the redundancy method are the effective way to find the optimal measurement configuration, which is an experimental design of the magnetic detection electrical impedance imaging. The invention provides an effective and reliable tool, and a magnetic detection electric impedance imaging data acquisition system based on a data acquisition card is established, Set and control unit and mechanical scanning device, etc., and design the multi-test based on LabVIEW Based on the discrete model, physiological saline model and agar model as the experimental object, the experimental results are carried out by using the model of the discrete model, the physiological saline model and the agar model, and the feasibility of the magnetic detection of the electrical impedance imaging is verified.
【学位授予单位】：天津大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：R310

【参考文献】