Purchase this article with an account.
Pooja Khamar, Mathew Francis, RUSHAD SHROFF, Natasha Pahuja, Roshan T, Rohit Shetty, HIMANSHU MATALIA, Abhijit Sinha Roy; A diagnostic biomechanical framework to distinguish keratoconus and normal eyes using air-puff applanation. Invest. Ophthalmol. Vis. Sci. 2017;58(8):4347.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To develop a diagnostic classification score for keratoconus using waveform analyses of corneal deformation
44 normal (44 patients) eyes, 62 suspect (fellow eyes of unilateral keratconus patient) eyes and 143 keratoconus (143 patients) were evaluated with Pentacam and Corvis-ST (OCULUS Optikgerate Gmbh, Germany). A spring and dashpot model (Matalia et al., J Ref. Surg., 2016; 32: 486-93) was used to derive corneal stiffness and other waveform derived variables. The model used segregated corneal and extra-tissue deformation from the total deformation amplitude (DA) reported by Corvis-ST. Based on the biomechanical model corneal stiffness was defined as linear [kc (constant)] and a non-linear measure [kc (mean)]. Multivariate linear regression analyses was performed. Area under the curve (AUC), sensitivity and specificity of the adjusted variables were obtained after including the suspect eyes in the keratoconus group. Cone location magnitude index (CLMI) was also calculated.
Analysis showed peak DA (1.08±0.02 vs. 1.18±0.02), applanation 1 time (7.25±0.01 vs. 7.20±0.01 msec), applanation 2 time (21.5±0.04 vs. 21.58±0.02 msec), kc (constant) [109.4±1.64 vs. 98.37±1.33 N/m], kc (mean) [107.53±2.02 vs. 93.38 N/m] and corneal peak DA [0.95±1.95×10-2 vs. 1.01±1.12×10-2 N/m] were significantly different between normal and keratconus eyes (p<0.001). Calculated and measured parameters of the fellow eyes (suspect) were similar to the keratconus group, particularly grade 1 (p>0.05). The AUC was highest in kc (constant) and kc (mean) (0.97 and 0.98 respectively) with sensitivity (>95%)and specificity (>88%). The logistic regression with the combination of kc parameters (constant and mean) improved AUC to1.0 (cut-off probability score of 0.82 between normal and keratoconus eyes including suspect eyes, 99.2%, 100.0%).
The suspect as well as keratoconus eyes could be differentiated from normal eyes based on corneal stiffness accurately. Further studies with eyes from different populations are needed to validate the scoring system based on corneal stiffness.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.
This PDF is available to Subscribers Only