O'Hagan Joseph J, Samani Abbas
Department of Electrical and Computer Engineering, University of Western Ontario, London, ON, Canada.
Phys Med Biol. 2008 Dec 21;53(24):7087-106. doi: 10.1088/0031-9155/53/24/006. Epub 2008 Nov 18.
The elastic and hyperelastic properties of biological soft tissues have been of interest to the medical community as there are several applications where parameters characterizing these properties are critical for a reliable outcome. This includes applications such as surgery planning, needle biopsy and cancer diagnosis using medical imaging. While there has been considerable research on the measurement of the linear elastic modulus of small tissue samples, little research has been conducted for measuring parameters that characterize nonlinear elasticity of tissues included in slice specimens. In this paper, we present a method of measuring the hyperelastic parameters of tissue slice samples with tumours. In this method, to measure the hyperelastic properties of a tumour within a slice sample, the tumour was indented to acquire its force-displacement response while the slice remained intact. To calculate the hyperelastic parameters from the acquired data, we developed two inversion techniques that use the slice nonlinear finite element model as their forward problem solver. One of these techniques was based on nonlinear optimization while the other is a novel iterative technique that processes the variable slopes of the force-displacement response to calculate the hyperelastic parameters. The latter was developed specifically for the Yeoh and the second-order polynomial hyperelastic models, since we found that the other optimization-based inversion technique did not perform well with these models. To validate the proposed techniques, we performed numerical and phantom experiments. While we were able to achieve convergence with wide ranges of parameters of initial guesses to within 1% error with the numerical simulation experiments, we achieved convergence to within errors of around 5% with the tissue mimicking phantoms. Moreover, we successfully applied these techniques to data we acquired from nine pathological breast tissue slice specimens where the goal was to determine the hyperelastic properties of the tumour within the breast tissue slices.
生物软组织的弹性和超弹性特性一直受到医学界的关注，因为在一些应用中，表征这些特性的参数对于可靠的结果至关重要。这包括手术规划、针吸活检以及使用医学成像进行癌症诊断等应用。虽然对于小组织样本的线性弹性模量的测量已经有了大量研究，但对于测量切片样本中组织的非线性弹性特性参数的研究却很少。在本文中，我们提出了一种测量带有肿瘤的组织切片样本超弹性参数的方法。在这种方法中，为了测量切片样本内肿瘤的超弹性特性，在切片保持完整的情况下对肿瘤进行压痕以获取其力 - 位移响应。为了从获取的数据中计算超弹性参数，我们开发了两种反演技术，它们使用切片非线性有限元模型作为其正向问题求解器。其中一种技术基于非线性优化，而另一种是新颖的迭代技术，该技术处理力 - 位移响应的可变斜率以计算超弹性参数。后者是专门为Yeoh和二阶多项式超弹性模型开发的，因为我们发现其他基于优化的反演技术在这些模型上表现不佳。为了验证所提出的技术，我们进行了数值实验和体模实验。在数值模拟实验中，我们能够在初始猜测的广泛参数范围内实现收敛，误差在1%以内，而在组织模拟体模实验中，我们实现了收敛，误差在5%左右。此外，我们成功地将这些技术应用于从九个病理性乳腺组织切片标本中获取的数据，目标是确定乳腺组织切片内肿瘤的超弹性特性。
