A Method To Determine Stiffness and IOP-induced Stress and Strain In-vivo

Abstract

 

Purpose:

 

Due to recent advances in optical coherence tomography (OCT) some geometric characteristics of the optic nerve head (ONH) and their changes due to increases in intraocular pressure (IOP) are measurable (M) or potentially measurable (PM) in-vivo (Table). Other characteristics of the ONH, such as the tissue mechanical properties and the IOP-induced stresses and strains in the lamina cribrosa (LC) and peripapillary sclera, remain unmeasurable (UM) in-vivo. This limits insight into the biomechanical effects of IOP on the ONH. Our goal was to use modeling to develop predictive models for UM from M and, if necessary, PM.

 

Methods:

 

Using a previously reported parameterized finite element model we simulated an increase in IOP of 10 mmHg on each of 4646 models of the human ONH representing a normal population (Table). We analyzed the resultant stresses and strains within the LC and sclera and 9 measures of deformation. Multivariate adaptive regression splines were fit to predict the UM as a function of M and PM and quality of fit assessed using 10-fold cross-validation. Multivariate adaptive regression splines were fit to predict the UM as a function of M and PM and quality of fit assessed using 10-fold cross-validation.

 

Results:

 

Excellent fits were obtained to predict all UM as a function of M and PM with R²s between 0.957 and 0.998. For example, using only M the predicted median tensile strains were accurate to within 0.06% in the sclera and within 0.13% in the lamina in 95% of cases. Adding PM improved the accuracy to 0.04% and 0.09% respectively. Using both M and PM the predictions of moduli were accurate to within 0.33 MPa for the sclera and 0.12 MPa for the LC, whereas the predictions of stress were within 2.1 kPa for the sclera and 1.3 kPa for the LC.

 

Conclusions:

 

In computational models we have predicted the scleral moduli with an accuracy of 0.33 MPa. This compares well with the measurements in the literature which have an uncertainty on the order of MPa. Most notable, however, is that we were able to predict lamina moduli, and IOP-induced stresses and strains which are not measurable in-vivo.  

 

Keywords: computational modeling • intraocular pressure • lamina cribrosa