RESEARCH PAPER Assessing subject-related variations of the Ocular Response Analyzer parameter calculation DOI:10.1111/cxo.12258 Marta E Rogowska* MSc D Robert Iskander* DSc PhD Henryk T Kasprzak † DSc PhD * Department of Biomedical Engineering, Wroclaw University of Technology, Wroclaw, Poland † Department of Optics, Wroclaw University of Technology, Wroclaw, Poland E-mail: marta.rogowska@pwr.edu.pl Background: The aim was to study the relationships between the output parameters of the Ocular Response Analyzer (ORA) and those calculated from the raw ORA data and to ascertain the subject-related variations of ORA parameter calculation procedures. Method: Six subjects were recruited for a prospective study. Up to 32 measurements by ORA were performed in series on the dominant eye of each subject. A relationship was examined between Goldmann-correlated intraocular pressure values (IOPg) obtained from the stand- ard ORA output and IOPg′ calculated from raw ORA data with a custom-written procedure. The same analysis was carried out for the parameters of corneal hysteresis (CH and CH′). Data and statistical analysis included Epanechnikov kernel smoothing, orthogonal linear regression, hypothesis testing and bootstrap techniques. Results: The group average (mean ± standard deviation) IOPg and CH values were 11.6 ± 1.8 mmHg and 10.7 ± 1.7 mmHg, respectively. A strong correlation was found between IOPg and IOPg′ and also between CH and CH′ parameters. There was a significant (Behrens– Fisher test, p < 0.001) difference between subjects for both IOPg and CH calculations, in terms of the regression slope parameter. Conclusions: Subject-related variations of ORA parameter calculation were demonstrated. This could indicate that currently employed estimators of IOP parameters include unre- ported algorithmic procedures that may lead to biased results. Submitted: 29 April 2014 Revised: 8 October 2014 Accepted for publication: 23 October 2014 Key words: air-puff tonometry, corneal hysteresis, glaucoma, intraocular pressure It is well known that the human cornea has viscoelastic properties 1–3 but comprehen- sive in vivo measurement 4 of such proper- ties is still a matter of the future. Undoubtedly, biomechanical properties of the human cornea affect the pressure measurement in the eye. 4–7 Measurement of intraocular pressure (IOP) is a very important factor in diagnosis of glaucoma. If detected early, glaucoma can be treated before too much damage occurs. 5 Hence, precise and accurate measurement of IOP is of essence. Most of the methods of IOP measurement are based on applanation of the cornea as it is in the case of Goldmann applanation tonometry (GAT), dynamic contour tonometry (DCT) and air-puff techniques. 8 The Ocular Response Analyzer, (ORA, Reichert, Inc, Depew, New York, USA) measures IOP in consid- eration of viscoelastic properties of the human cornea. 4–6 Also, it appears that ORA acquires more accurate measurements of IOP than the Goldmann applanation tonometer. 9–11 ORA uses a rapid (25 ms) air pulse to deform the cornea. The waft induces two applanation events: the first ‘inward’ (peak 1) and second ‘outward’ (peak 2). Both events correspond to two specific values of the air pressure, known as the first applanation pressure (P1) and the second applanation pressure (P2). Subsequently, corneal-compensated IOP (IOPcc), Goldmann-correlated IOP (IOPg), corneal hysteresis (CH) and corneal resistance factor (CRF) are determined, among a number of other parameters of the meas- ured applanation curve. The corneal hys- teresis characterises viscoelastic properties of the cornea. 3,4 It has been demonstrated that the corneal hysteresis can be used as an indicator for disease diagnosis and management. Glau- coma patients have been reported to have a lower CH than normal control subjects 7,12–15 and a lower hysteresis has been associated with increased deformation of the optic nerve surface during transient elevation of IOP. 12 Each ORA measurement is described by a quality index called the waveform score. This parameter is bounded between zero and 10 with lower numbers indicat- ing poorer quality. Lam, Chen and Tse 16 have shown that measurements with a waveform score of less than 3.5 should not be taken into account. Vantomme, Moustaka and Pourjavan 17 arrived at a similar conclusion showing that waveform scores of less than three are not sufficiently accurate. ORA allows up to four measurements in one single acquisition but the repeatability of the instrument, as indicated in our pilot studies, is not always high. This prompted testing the reliability of ORA measurements and their subsequent usefulness for diagnos- ing diseases. 18,19 In this paper, calculations of CH and IOPg parameters were studied. The main goal of this work was to determine the relation- ship between ORA output parameters and those calculated from the raw ORA data using a custom-written procedure and to CLINICAL AND EXPERIMENTAL OPTOMETRY Clin Exp Optom 2015; 98: 348–352 348 Clinical and Experimental Optometry 98.4 July 2015 © 2015 The Authors Clinical and Experimental Optometry © 2015 Optometry Australia