Appropriate Objective Functions for Quantifying Iris Mechanical Properties Using Inverse Finite Element Modeling
Appropriate Objective Functions for Quantifying Iris Mechanical Properties Using Inverse Finite Element Modeling
复制标题
DOI：
10.1115/1.4039679
复制
发表时间：
2018-07-01
影响因子：
1.7
通讯作者：
Amini, Rouzbeh
中科院分区：
文献类型：
--
作者：
Pant, Anup D.;Dorairaj, Syril K.;Amini, Rouzbeh
量化虹膜的机械特性是重要的，因为它提供了对青光眼病理生理学的深入了解。最近的离体研究表明，虹膜的机械性能是不同的，在青光眼的眼睛相比，正常的。尽管离体研究的重要性，这种测量是严重限制的诊断和排除治疗策略的发展。随着详细的成像方式的出现，它是可能的，以确定在体内的力学性能，使用逆有限元（FE）建模。逆建模方法需要适当的目标函数来可靠地估计参数。在虹膜的情况下，在临床实践中常规地进行许多测量，例如虹膜弦长（CL）和虹膜宽度（CV）。在这项研究中，我们已经评估了五个不同的目标函数的基础上选择的虹膜生物特征（在存在和不存在临床测量误差），以确定适当的标准进行逆建模。我们的研究结果表明，在实验测量误差的情况下，虹膜CL和CV的组合可以用作目标函数。然而，随着测量误差的增加，采用大量局部位移值的目标函数提供了更可靠的结果。
Quantifying the mechanical properties of the iris is important, as it provides insight into the pathophysiology of glaucoma. Recent ex vivo studies have shown that the mechanical properties of the iris are different in glaucomatous eyes as compared to normal ones. Notwithstanding the importance of the ex vivo studies, such measurements are severely limited for diagnosis and preclude development of treatment strategies. With the advent of detailed imaging modalities, it is possible to determine the in vivo mechanical properties using inverse finite element (FE) modeling. An inverse modeling approach requires an appropriate objective function for reliable estimation of parameters. In the case of the iris, numerous measurements such as iris chord length (CL) and iris concavity (CV) are made routinely in clinical practice. In this study, we have evaluated five different objective functions chosen based on the iris biometrics (in the presence and absence of clinical measurement errors) to determine the appropriate criterion for inverse modeling. Our results showed that in the absence of experimental measurement error, a combination of iris CL and CV can be used as the objective function. However, with the addition of measurement errors, the objective functions that employ a large number of local displacement values provide more reliable outcomes.