Purchase this article with an account.
Anup Pant, Rouzbeh Amini; Determination of the iris mechanical properties using inverse finite element simulation. Invest. Ophthalmol. Vis. Sci. 2015;56(7 ):6140.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To develop a framework to obtain mechanical properties of the iris from clinical imaging modalities for potential diagnosis of glaucoma.
Our previous finite element model of the iris contraction during dilation (IOVS, 53: 1188-1194) was modified to calculate the iris shear modulus (G) and dilator internal stress (σ). In the clinical setups, the iris displacement during miosis/mydriasis can be obtained from imaging methods such as ultrasound biomicroscopy and anterior segment optical coherence tomography. To reliably calculate mechanical properties from such data, we verified our method using pseudo-experimental displacement data obtained from predefined values of 9 KPa for G and 4 KPa for σ in our forward finite element model. An inverse modeling approach based on the genetic algorithm (J.Global. Optim.,11: 341-359) was used to extract the values of G and σ from the pseudo-experimental data. To assess the optimization method sensitivity to the experimental errors, a second set of data was generated by adding random noise within a ±%8 range of the original pseudo-experimental data. In both cases, the norm of displacement between the pseudo-experimental data and the genetically driven ones was chosen as the objective function. The simulations were performed on an HP Intel Xeon machine at the Ohio Supercomputing Center (Columbus,OH).
In both cases, when the initial guessed values were chosen between 8-10 KPa for G and 3-5 KPa for σ, the optimization technique solution accurately matched the values used for generating pseudo-experimental data (i.e.9 KPa for G and 4 KPa for σ). In each case, simulation took approximately 190 minutes (~50 numbers of generations) until the values of the G and σ were within the acceptable numerical error (10-5 mm). The minimum error for the original data was found to be less than 10-8 mm while for the case with added noise, the error was found to be larger (4.6 × 10-6 mm).
In addition to, iris biometrics such as the curvature, thickness, and chord length, iris compressibility has also been identified as a risk factor for glaucoma during dynamics phenomena such as dilation (J Glaucoma,18:173-179;Ophthalmology,117:3-10). The method developed and verified here provides an opportunity to quantify the iris mechanical properties from clinical images and to conduct studies on the potential relation between the iris stiffness and the glaucoma risks.
This PDF is available to Subscribers Only