Purpose:
To evaluate the performance of CH, CRF, and 16 custom-derived ORA parameters in distinguishing KC corneas from normal corneas.
Methods:
Sixteen custom parameters were derived from the ORA raw data measurement and calculated with an in-house Matlab program. Fifty-four normal eyes (27 patients) and 49 KC eyes (25 patients) in a comparative case series underwent a complete clinical eye examination, corneal topography, corneal tomography, and ORA evaluation. Receiver operating characteristic (ROC) statistics compared CH, CRF, and custom parameters’ ability to characterize biomechanical properties and distinguish KC. Selected parameters were correlated with KC severity, central corneal thickness (CCT) and maximum corneal curvature (max K).
Results:
Parameters with the highest Area Under the ROC (AUROC) are in figure 1. MinATrough is the minimum applanation value between peaks on the ORA waveform, and MeanTrough is the average value. Hysteresis Loop Area (HLA) is the area within an ORA pressure versus applanation plot. Hysteresis Loop Area complete (HLAc) is the area calculated after inverting part of the HLA plot and completing the curve. MinATrough had the best overall predictive accuracy (cutoff, 50.37; sensitivity, 94.9%; specificity, 91.7%; test accuracy, 93.2%) for detecting KC. MinATrough, MeanTrough, HLA, and HLAc showed decreasing trends with KC severity and Max K and increased with CCT.
Conclusions:
Parameters MinATrough, HLA, MeanTrough, HLAc, and CRF outperformed CH in ability to distinguish KC. MinATrough may be most sensitive to the corneal biomechanical changes associated with KC. Further studies should investigate new parameters’ ability to distinguish KC severity gradations and other pathologies affecting corneal biomechanics.
Keywords: keratoconus • anterior segment • cornea: basic science