Accuracy of an Air-Puff Dynamic Tonometry Biomarker to Discriminate the Corneal Biomechanical Response in Patients With Keratoconus

Abstract

Purpose: The aim of this study was to assess accuracy of the mean corneal stiffness (k c , N/m) parameter to discriminate between patients with keratoconus and age-matched healthy subjects. Methods: Dynamic Scheimpflug imaging tonometry was performed with Corvis ST (Oculus Optikgeräte GmbH, Germany) in patients with keratoconus (n = 24; study group) and age-matched healthy subjects (n = 32; control). An image processing algorithm was developed to analyze the video sequence of the Corvis ST air-puff event and to determine the geometric and temporal parameters that correlated with the corneal tissue biomechanical properties. A modified 3-element viscoelastic model was used to derive the k c parameter, which represented the corneal tissue resistance to deformation under load. Receiver operating characteristic curves were used to assess the overall diagnostic performance for determining the area under the curve, sensitivity, and specificity of the k c in assessing the corneal tissue deformation to the Corvis ST air-puff event in keratoconus and control eyes. The Corvis Biomechanical Index (CBI) was analyzed for external validation. Results: The k c parameter was significantly different between keratoconus and controls (P < 0.001), ranging from 24.9 ±3.0 to 34.2 ±3.5 N/m, respectively. It was highly correlated with CBI (r = −0.69; P < 0.001); however, the k c parameter had greater specificity (94%) than CBI (75%), whereas the 2 biomarkers had similar area under the curve (0.98 vs. 0.94) and sensitivity (96% vs. 92%) in predicting the occurrence of keratoconus. Conclusions: The k c parameter extracted by video processing analysis of dynamic Scheimpflug tonometry data was highly accurate in discriminating patients with clinically manifest keratoconus compared with controls.