Shear modulus reconstruction in dynamic elastography: time harmonic case
Shear modulus reconstruction in dynamic elastography: time harmonic case
复制标题
DOI：
10.1088/0031-9155/51/15/007
复制
发表时间：
2006-08-07
影响因子：
3.5
通讯作者：
Maniatty, Antoinette M.
中科院分区：
文献类型：
--
作者：
Park, Eunyoung;Maniatty, Antoinette M.
本文提出了一种直接反演方法，用于根据时间谐波激励期间内部位移场的动态测量来重建软组织中的弹性剪切模量。假设组织服从简谐运动中几乎不可压缩的、线性的、各向同性的弹性动力学方程。控制方程的有限元离散化被用作基础，并概述了一个程序来消除反问题中对边界条件的需要。在此过程中也会重建静水应力（压力），并考虑在控制方程中忽略此项的影响（这是常见的做法）。该方法不需要迭代，并且可以在域的子区域上执行，从而形成计算高效的方法。进行敏感性研究以研究不同尺寸和剪切模量与背景对比的异常区域的可检测性。该算法在二维域上的模拟数据上进行测试，其中数据在非常精细的网格上生成以获得接近精确的解，然后下采样到类似于实际数据的空间离散化的较粗糙的网格，并添加噪声。结果显示了静水应力项和噪声的影响。还使用涉及模拟组织模型的 MR 测量实验数据进行重建，以演示该算法。
This paper presents a direct inversion approach for reconstructing the elastic shear modulus in soft tissue from dynamic measurements of the interior displacement field during time harmonic excitation. The tissue is assumed to obey the equations of nearly incompressible, linear, isotropic elasto-dynamics in harmonic motion. A finite element discretization of the governing equations is used as a basis, and a procedure is outlined to eliminate the need for boundary conditions in the inverse problem. The hydrostatic stress (pressure) is also reconstructed in the process, and the effect of neglecting this term in the governing equations, which is common practice, is considered. The approach does not require iterations and can be performed on sub-regions of the domain resulting in a computationally efficient method. A sensitivity study is performed to investigate the detectability of abnormal regions of different size and shear modulus contrast from the background. The algorithm is tested on simulated data on a two-dimensional domain, where the data are generated on a very fine mesh to get a near exact solution, then downsampled to a coarser mesh that is similar to the spatial discretization of actual data, and noise is added. Results showing the effect of the hydrostatic stress term and noise are presented. A reconstruction using MR measured experimental data involving a tissue-mimicking phantom is also shown to demonstrate the algorithm.