Purchase this article with an account.
D. Luce; Methodology for Keratoconus Risk Indices Using Ocular Response Analyzer (ORA) Waveform Analysis. Invest. Ophthalmol. Vis. Sci. 2010;51(13):4983.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To present a biomechanical method for determining keratoconus (KC) indices capable of detecting early stage keratoconus.
In addition to four original ORA parameters (IOPg, IOPcc, CH and CRF), 37 primitive mathematical parameters are calculated of which ten are used to determine an independent waveform score that is used to select a "best" measurement from multiple measurements taken for each eye. "Best" measurement waveforms from 813 normal (NORM), 129 clinical forme fruste keratoconus (FFKC) and 192 keratoconus eyes were then analyzed using a linear, normalized combination of 6 individual parameter scores. Normalization used a mathematical construct that forces the mean values of the reference populations NORM and composite KC to 1 and 0 respectively. Scoring parameters were selected using univariate t-test p values. To establish baseline values for the KC indices, clinically (non-ORA) classified normal (NORM), forme fruste (FFKC), mild (MIKC), moderate (MOKC) and severe (SEKC) keratoconus eyes were scored using the baseline means and standard deviations. The clinical populations were then modeled as normal distributions that are then used to calculate probabilities (KC indices) that a measurement is characterized by each of the 5 clinical populations.
Mean scores and standard deviations for the NORM, FFKC, MIKC, MOKC and SEKC groups are 1.0 (0.40), 0.59(.40), 0.14(0.52), -0.35(0.38), and -0.63(0.26). T-test values for all groups are p< 1E-39. For the FFKC eyes the t-test p value of the 6 parameter score is 5.7E-40. ROC sensitivity and specificity are 92% and 65% respectively for the FFKC versus NORM eyes.
ORA waveform analysis strongly separates all normal and KC groups, including the NORM, FFKC and MIKC eyes where diagnosis is most critical. Scores of all five non-ORA clinical keratoconus populations are in the independent (non-ORA) clinically defined sequence, supporting the conclusion that the test procedure described here can provide valuable information to assist in the identification of early stage compromised corneas.
This PDF is available to Subscribers Only