Theoretical Analysis of an Unknown Change in Corneal Refractive Index Following Corneal Cross-Linking Introduces Artifact in Optical Measurements of Central Corneal Thickness

Abstract

Purpose : Central corneal thickness (CCT) measurements vary following corneal cross-linking (CXL), depending on the instruments used. We sought to explain this discrepancy by investigating the consequences of a proposed unexpected change in corneal refractive index (n_c) on Scheimpflug imaging and optical coherence tomography (OCT) measurements of CCT.

Methods : We developed two theoretical models to test the effect of a change in n_c on CCT. Our Scheimpflug model used existing equations that utilize a side viewing angle of an optical section of cornea illuminated by a central slit to predict CTT. Similarly, our OCT model used existing equations that describe optical path length (OPL) to predict CCT. Predicted CCT was calculated for a range of n_c values in both models. CCT measurement error was defined as the difference between the CCT measurement at the instrument-assumed index of 1.376 (CCT_1.376) and the predicted CCT at other indices (CCT_n). This error was plotted versus n_c to analyze the effect a change in n_c would have on predicted CCT for both models. To quantify the effect, we calculated the percent error in predicted CCT measurements and plotted it versus percent increase of n_c for both models.

Results : Our Scheimpflug model produced a positive association between predicted CCT and n_c. It also suggested a negative association between CCT measurement error(CCT_1.376-CCT_n) and n_c, which was quantified as a 0.822% reduction in CCT for every 1% increase in n_c (Fig1). In contrast, our OCT model produced a negative association between predicted CCT and n_c with a positive association between CCT measurement error and n_c, resulting in a 0.983% increase in CCT for every 1% increase in n_c (Fig2).

Conclusions : Our models suggest that an unknown increase in n_c results in underestimation of CCT from Scheimpflug-based devices, and, conversely, overestimation of CCT from OCT devices. These results suggest that a change in n_c after CXL introduces artifact in optical CCT measurement techniques. Our predictions are consistent with literature reports of CCT 2 weeks after CXL that show a 9.7μm increase when measured with OCT yet a 16μm decrease when measured by Scheimpflug (Antonios et al. 2016).

This is a 2021 ARVO Annual Meeting abstract.

 

For every 1% increase in n_cornea we see a 0.822% decrease in CCT in Scheimpflug

 

For every 1% increase in n_cornea we see a 0.923% increase in CCT in OCT