基于超声弹性成像技术的生物组织应变估计算法仿真研究

发布时间：2018-04-25 14:29

本文选题：超声弹性成像 + 应变估计　；参考：《哈尔滨工业大学》2012年硕士论文

【摘要】：依赖于组织的分子构成和相应的微观组织结构，生物软组织的弹性属性（或硬度）的变化与软组织的病理变化过程有关，病变组织与正常组织之间存在着弹性属性（或硬度）的差异。通过对生物软组织施加一个激励，使生物软组织发生形变，得到生物软组织压缩前、后的超声回波射频信号，利用信号处理技术及互相关技术，计算组织内部位移分布，估计组织的应变分布以及组织内部的弹性分布。本文基于反射式超声弹性成像系统优化和提出了一维轴向应变估计算法，准确地估计组织内部的弹性模量（或硬度）分布信息。本论文的主要工作有：首先，介绍了超声弹性成像技术中生物软组织的力学基本定律—胡可定律，研究了生物软组织的力学特性，超声弹性成像技术在生物软组织中的应用，并对组织模型进行简化处理，以及生物组织弹性成像的力学模型；研究了超声弹性成像技术的原理和目前常见的几种弹性成像实现技术以及基于静态/准静态压缩超声弹性成像原理的超声弹性成像系统及其分类，比较了反射式超声弹性成像系统和透射式超声弹性成像系统，并研究反射式超声弹性成像系统的成像原理。其次，研究了超声弹性成像的经典算法—互相关算法，以及在互相关算法基础上的传统时域互相关算法，，针对传统方法的不足，设计了计算伸缩系数的方法，来估计组织的应变，提出了利用计算插值点个数估计应变算法，这种算法能够得到正确反应组织内部弹性模量（或硬度）信息的伸缩系数分布图，但是精度较低（1%），并在此基础上优化并提出了计算伸缩系数估计应变算法。再次，总结并提出了用于弹性成像算法的参数选择、插值选择等分析方法。研究了不同窗口长度和重叠率对弹性成像算法的影响，通过仿真实验，确定窗口长度为100和重叠率为0.95为最优参数；对比分析了线性插值算法、立方插值算法、样条插值算法等，选择线性插值算法用于信号拉伸，运行时间较短和精度较高；为了提高仿真实验的速度，根据成像算法的结构特点，实现了算法在实验室研发的生物医学图像处理平台上的并行化，大大缩短了算法的运行时间。最后，本文中的算法进行了仿真实验，对得到的仿真结果进行了论证与讨论；并对本课题进行了总结以及对进一步的研究方向做了展望。
[Abstract]:The changes of elastic properties ( or hardness ) of biological soft tissues depend on the structure of tissues and the corresponding microstructure , and the difference of elastic properties ( or hardness ) between diseased tissue and normal tissue .

Firstly , the mechanical basic law of biological soft tissues in ultrasonic elastography is introduced . The law of Hu ' s law is used to study the mechanical properties of soft tissue , the application of ultrasonic elastography in soft tissues and the simplified treatment of tissue model and the mechanical model of biological tissue elasticity imaging ;
In this paper , the principle of ultrasonic elastography is studied , and several kinds of elastic imaging systems and its classification are studied . The reflective ultrasonic elastography system and the transmission ultrasonic elastography system are compared , and the imaging principle of the reflective ultrasonic elastography system is studied .

Secondly , the classical algorithm - cross - correlation algorithm of ultrasonic elastography and the traditional time - domain cross - correlation algorithm based on cross - correlation algorithm are studied .

Thirdly , the influence of different window length and overlapping rate on the elastic imaging algorithm is studied . Through simulation experiments , it is determined that the window length is 100 and the overlap ratio is 0.95 as the optimal parameter .
The linear interpolation algorithm , cubic interpolation algorithm , spline interpolation algorithm and so on are compared and analyzed , and the linear interpolation algorithm is selected for signal stretching , and the operation time is short and the precision is high ;
In order to improve the speed of simulation experiment , according to the structural characteristics of the imaging algorithm , the parallelizing of the algorithm in the biomedical image processing platform developed by the laboratory is realized , and the running time of the algorithm is greatly shortened .

Finally , the simulation experiment is carried out on the algorithm in this paper , and the simulation results are proved and discussed .
This paper summarizes the subject and prospects for further research direction .

【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R310;TP301.6

【参考文献】