Article Goldmann Tonometer Prism with an …doclibrary.com/MSC167/PRM/CATS_TVST4311.pdf1 TVST j 2016 j Vol. 5 j No. 5 j Article 4 This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives

DOI: 10.1167/tvst.5.5.4

Article

Goldmann Tonometer Prism with an Optimized ErrorCorrecting Applanation Surface

Sean McCafferty1–3, Garrett Lim1, William Duncan1,3, Eniko Enikov4,and Jim Schwiegerling2,3

1 Intuor Technolgies, Tucson, AZ, USA2 University of Arizona Department of Ophthalmology, Tucson, AZ, USA3 University of Arizona College of Optical Science, Tucson, AZ, USA4 University of Arizona College of Engineering - Department of Mechanical and Aerospace, Tucson, AZ, USA

Correspondence: Sean J. McCaff-erty, Intuor Technologies, LLC, 6422E. Speedway Boulevard, Suite 100,Tucson, AZ 85710, USA. e-mail:[email protected]

Received: 24 March 2016Accepted: 12 July 2016Published: 9 September 2016

Keywords: glaucoma; intraocularpressure; corneal biomechanics;tonometry

Citation: McCafferty S, Lim G, Dun-can W, Enikov E, Schwiegerling J.Goldmann tonometer prism with anoptimized error correcting applana-tion surface. Trans Vis Sci Tech. 2016;5(5):4, doi:10.1167/tvst.5.5.4

Purpose: We evaluate solutions for an applanating surface modification to theGoldmann tonometer prism, which substantially negates the errors due to patientvariability in biomechanics.

Methods: A modified Goldmann or correcting applanation tonometry surface (CATS)prism is presented which was optimized to minimize the intraocular pressure (IOP)error due to corneal thickness, stiffness, curvature, and tear film. Mathematicalmodeling with finite element analysis (FEA) and manometric IOP referenced cadavereyes were used to optimize and validate the design.

Results: Mathematical modeling of the optimized CATS prism indicates anapproximate 50% reduction in each of the corneal biomechanical and tear filmerrors. Manometric IOP referenced pressure in cadaveric eyes demonstratessubstantial equivalence to GAT in nominal eyes with the CATS prism as predictedby modeling theory.

Conclusion: A CATS modified Goldmann prism is theoretically able to significantlyimprove the accuracy of IOP measurement without changing Goldmann measure-ment technique or interpretation. Clinical validation is needed but the analysisindicates a reduction in CCT error alone to less than 62 mm Hg using the CATS prismin 100% of a standard population compared to only 54% less than 62 mm Hg errorwith the present Goldmann prism.

Translational Relevance: This article presents an easily adopted novel approach andcritical design parameters to improve the accuracy of a Goldmann applanatingtonometer.

Introduction

Glaucoma is the disease most commonly associatedwith intraocular pressure (IOP). However, IOP isarguably the second-most critical metric next to visualacuity for assessing the ocular health of an individual.Therefore, accurate, repeatable, and interpatient com-parable measurements of IOP often are helpful andoccasionally critical for the treatment of all oculardisease processes as well as for routine screening examsby any eye care professional. Intraocular pressuremeasurement not only is critical to the accurate diagnosisof ocular disease, but also is a necessary guide to effectivetreatment strategies. The IOP is assessed routinely on themajority of patient visits to eye care professionals,

including ophthalmologists and optometrists, whichtotal approximately 50,000 in the United States andapproximately 450,000 worldwide.1,2 Glaucoma alone isa chronic and potentially debilitating disease requiringlifelong treatment. This disease currently affects 2.2millionAmericans, with 3.3millionmore expected by theyear 2020. Glaucoma is now the leading cause ofblindness in the aging Hispanic and African Americanpopulations, and nearly three times as common inAfrican Americans as in White Americans. Worldwide,there were 60.5 million people with various types ofglaucoma in 2010; this figure is expected to increase to79.6 million by 2020.3 Patients still go blind and suffersignificant debilitating vision loss from glaucoma due tomisdiagnosis and mismanagement.4

1 TVST j 2016 j Vol. 5 j No. 5 j Article 4

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.Downloaded from tvst.arvojournals.org on 02/03/2020

https://creativecommons.org/licenses/by-nc-nd/4.0/

For almost 60 years, Goldmann applanationtonometry (GAT) remains the standard for mea-surement of IOP.5,6 Numerous significant errors inthe GAT IOP measurements (mm Hg) have beendescribed previously. Those errors are due to patientvariability in corneal thickness (67 mm Hg), cornealrigidity (68 mm Hg), corneal curvature (62 mmHg), and corneal tear film (65 mm Hg).7–9 Thecombined errors of patient variable parameters arepotentially sight threatening to a large population ofpatients, such as those with glaucoma or undiag-nosed ocular hypertension from other causes, yetcurrently this is the best method we have clinically.Despite the inherent shortcomings identified in theGAT, nothing has improved upon its accuracy, cost-effectiveness, and ease of use. This problem wasbrought into the spotlight by the findings of theOcular Hypertension Treatment Study (OHTS),which noted that pressure readings tend to beoverestimated in thick, and underestimated inthinner, corneas. These errors lead to a misdiagnosisof glaucoma.10 Since the OHTS findings, thestandard of practice has changed to include ameasurement of central corneal thickness (CCT)with a nomogram to correct the pressure for theCCT. Additionally, the effects of laser-assisted insitu keratomileusis (LASIK) surgery render accurateIOP measurement by the GAT problematic.11 TheCCT correction has been partially effective butunreliable due to the other potential corneal biome-chanical and tear film errors. Attempts were made tomeasure and quantify the various error componentsto correct the GAT measurement and yield astandard IOP reading comparable between pa-tients.12 However, the process in practice is errorprone and cumbersome, leading to very limitedclinical adoption with the exception of CCT. Otherdirect measurements of IOP that potentially reduceerror have been developed, such as the dynamiccontour tonometer (DCT) which provides for aconstant appositional force of 1 g on a concavesurface, which contains a central miniaturizedpiezoresistive pressure sensor. This device is verysimilar to a tonopen tonometer but adds the constant1 g of force which partially negates biomechanicalerrors by allowing the corneal deformation force tobe partly resisted by the portion of the contactsurface that does not measure IOP. The DCT similarto the error correcting noncontact ocular responseanalyzer (ORA) is not widely used and has hadminimal clinical acceptance for routine IOP mea-surement.

Central corneal thickness or corneal thickness ingeneral is a geometric quantity affecting the rigidity ofthe cornea.8 The cornea is assumed by the Imbert-Fick principle to be an infinitely thin membrane thatby definition has no shear rigidity, only strength intension.5,8 The ‘‘rigidity’’ of the cornea also is affectedby the corneal curvature. A steeply curved corneamust be ‘‘bent’’ more to applanate against thetonometer prism overestimating the IOP. Converselya flat cornea, such as in someone who has hadLASIK, underestimates the IOP. Also, the intrinsicmaterial property of the cornea (the modulus ofelasticity – both Young’s and shear) greatly affect the‘‘rigidity’’ of the cornea.13–15 All of these rigidity-affecting components together increase the force onthe tonometer prism, which is attributed to IOP but,in fact, have no direct relation to IOP, hence the error.Finally, attraction created by the surface tension inthe tear film (which also is extremely variable inpatients) was theorized to negate much of the‘‘rigidity’’ error.16,17 However, no clinical quantifica-tion of this highly variable attractive capillary forcehas been demonstrated in its effect on IOP.

The CATS tonometer prism is a modification ofthe GAT prism. The CATS prism optimizes thecorneal applanating surface of the flat surface GATprism. The function of the CATS prism, includingforce to pressure conversion supplied by the GATarmature, remains unchanged. The optimized CATSprism is designed to measure the same pressure asGAT prism under ‘‘nominal’’ conditions. The ‘‘nom-inal’’ conditions include a standard average cornealthickness, curvature, rigidity, and tear film. However,approximately 50% of the patient population do nothave a ‘‘nominal’’ cornea.7,9,10,12 In these patients, thevariability in each of these parameters inducessignificant individual and combined error in GATIOP measurement, as mentioned previously.7–9 Eventhe errors due to corneal thickness alone, which is buta fraction of the total error, are sight threatening.10

For this reason, CCT correction was adopted as astandard of practice.10 Although several other sepa-rate measurements and error corrections were pro-posed, they have been too cumbersome to be adoptedclinically.12 The CATS tonometer prism, illustrated inFigures 1 and 2, can significantly reduce all of theidentified measurement errors using the exact samemeasurement apparatus (with a modified prism),practitioner protocol, and measurement techniquewithout calculations, increased time, and at minimalcost.


McCafferty et al.

Downloaded from tvst.arvojournals.org on 02/03/2020

Methods

Two methods were used to assess the GAT: thefinite element modeling (FEM), and cadaveric humaneye benchtop testing. These methods are described indetail below.

Finite Element Modeling

A GAT estimates IOP by applanation of thecornea to a specified area. Based on the Imbert-Fickprinciple,5 the force to pressure conversion assumesthat the IOP is uniquely responsible for the forcerequired to applanate the cornea force.18,19 In reality,the structure of the eye contributes significantly to theapplanation force.7–10 Moreover, its contribution willvary based on the specimen. It is known that themeasured IOP is affected by material properties andgeometry of the eye.7,8,12–14

Through computer modeling simulations, infor-mation is gained about tonometry that is not possiblefrom theoretical and experimental consideration. Toproperly guide these simulations, appropriate as-sumptions about the physical behavior of corneatissue were made. The tissue of the cornea is anassembly of cells with complex anatomies andstructural properties. In simulation, tissue as acontinuum was analyzed, with inhomogeneous mate-rial properties. For the purposes of this study, themodels were assumed to have three variable materialproperties: (1) cornea substrate elastic modulus, (2)collagen elastic modulus, and (3) relative collagenthickness. These materials were assigned to particularphysical entities in a virtual assembly, and optimizedto match real-world behavior.

A basic theory of GAT is the Imbert-Fickprinciple, as shown in Equation 1. This simpleequation states that the reaction force of the eye, F,is a linear function of the IOP, P. The reaction forcealso depends on the force to deform the cornea tissue,T, and the cross-sectional contact area of thetonometer surface, A. In this study, the normal IOP,P0, was 16.0 mm Hg.

FðPÞ ¼ TðdÞ þ PAðdÞ ð1Þ

The contact area is a function of the displacementdepth, d. In this study, the modeled cornea had aspherical radius of 7.800 mm, and the tonometer hada cylindrical radius of 1.53 mm. This resulted in themaximum displacement of 0.147 mm, and themaximum contact area was 7.354 mm2. The calcula-tion of the contact area, A, as a function of thespherical radius of the cornea, R, and the verticaldisplacement, d, is shown in Equation 2.

AðdÞ ¼ pð2Rdþ d2Þ ð2Þ

In GAT, the measured IOP, PGAT, is a linearfunction of the reaction force. It also depends on acalibration reaction force F(P), which is compared to

Figure 1. Sagittal cross-section of the CATS tonometer prismapplanating surface.

Figure 2. Photograph of the concave–convex CATS tonometerprism applanating surface on the prototype illustrating theoptically polished and contoured surface.


McCafferty et al.


the normal cornea F550(P0), where the 550 refers tothe nominal 550 lm CCT, and P0 is the nominal IOP.This is shown in Equation 3.

PGAT ¼ P0FðPÞ

F550ðP0Þ

� �ð3Þ

The virtual models were designed in AutodeskInventor LT 2015 and simulated in AutodeskSimulation Mechanical 2015 (Autodesk, San Rafael,CA). Several simulations were executed to determinethe sensitivities of the CATS tonometer to variouscornea properties. These included: (1) IOP, (2)Young’s modulus, (3) CCT, and (4) and centralcornea curvature (CC). Each of these were simulatedso as to be comparable to results from other studies inthis field.14,21–25

Geometric and constitutive models were selectedbased on the results of previous studies.14,21,25 Thematerial properties were determined via analyses offinite element simulations. The effects of the variousgeometric aspects of the cornea were measured andstudied in previous studies. Since the publishedcorneal material properties vary widely, the specificproperties were chosen to approximate known reac-tions to GAT diagnostics. The force required forapplanation of a normal cornea under normalconditions was set near 1.6 g. The cornea was allowedto contribute only 30% of this applanation forceunder nominal conditions, with the rest coming fromthe IOP. The finite element mesh density was set sothat the perimeter of the applanation area would beprecise to within 30 lm, but with a measurementtolerance of no finer than 0.1 g. A nominal cornea has

a CCT of 550 lm, central radius of curvature of 7.8mm, P value of 0.82, and width of 11.0 mm. The Pvalue is a metric of eccentricity of an ellipsoid. Asphere has a P value of 1. A prolate ellipsoid, such asthe human cornea, has a P value less than 1.

The sensitivities to the various properties wereanalyzed parametrically. Those of material andphysical properties were simulated with a normalgeometry. Those of geometric properties requiredunique virtual models. The virtual applanation of thenormal cornea with the CATS prism is shown inFigure 3.

The key to the optimization process was minimi-zation of intracorneal stress during applanatingdeformation, which occurs at the perimeter of thecontact area. Minimization of the maximum stressalso reduces the corneal contribution to the measuredIOP. Comparison of the GAT and optimized CATSprism designs indicates reduced transition-zone stressconcentration. Alternatively, this also is minimizationof the second derivative or the rate of change ofcurvature of the cornea.

The IOP error dependence on tear film is quanti-tatively minimized, but is not included in the FEAmodel. Tear film modeling was completed usingpublications on capillary pressure calculation of a fluidbridge between two curved solids.16,17 The attractiveforce created by the fluid bridge and associated surfacetension of the tear film between the tonometer prismand cornea is reduced in the mathematical model byincreasing the contact angle between the cornea andcontacting prism in the region of the tear film meniscuswith the CATS Tonometer prism. (Fig. 1)

Figure 3. Concave-convex prism surface cross-section of the CATS prism surface in contact with the cornea.


McCafferty et al.


Ultimately, the results of the mathematical mod-eling theoretically optimized the applanation surfaceof the tonometer prism. The optimized surface then isincorporated in the GAT design. Several prototypeswere constructed and their proper design and functionwere verified. The CATS prisms then were placed intoexisting Goldmann tonometer armatures for clinicalevaluation.

Human Cadaveric Eye Testing

Bench testing was performed with the CATSTonometer prism prototypes representative of thedevice to be marketed. The bench testing consisted ofhuman cadaveric eyes to test for accuracy andrepeatability. Specifically, we were examining IOPmeasurement bias and variability comparing theCATS and GAT prism measurements relative to themanometric and pressure gauge measured IOP as atrue value. The objective of this testing was todemonstrate substantial equivalence on a nominalcornea between the series of measurements made bythe GAT prism and the CATS prism at each pressure.

Three enucleated human globes were stabilized in aspecially designed chamber for pressurizing a wholeglobe (Fig. 4) with the cornea exposed.

Standard biological precautions were followedwhen handling eye tissue. The pressure was initiallyset to a true 16 mm Hg via intracameral transducerand the corneal thickness measured via ocularcoherence tonometer (OCT). Intraocular pressuremeasurements were taken using a Perkins-type GATwith the GAT and CATS prisms. Previous studiesdemonstrated that the Perkins tonometer is clinicallyequivalent to the slit-lamp mounted GAT.26 Three

eyes were individually measured five times with eachprism at each of the following nine intracameralpressures (0, 5, 10, 15, 20, 25, 30, 40, 50 mm Hg). Eachmeasurement was rotated counterclockwise 908 froma standard reference axis to account for anyastigmatic errors. For example, three cadaver eyeswere tested four times (alternately rotating 908) at a 5mm Hg intracameral pressure using the GAT and theCATS prisms. A randomization occurred to deter-mine which tonometer prism was used first.

The San Diego Eye Bank (San Diego, CA)provided three freshly enucleated human cadaverglobes that were suitable for corneal transplantation.Eyes were stored at 48C in Optisol chambers untiluse.27 The cadaver eyes were used on the day ofarrival and within hours postmortem. The eyes, agesof the cadavers, and cause of death were recorded.Eyes with a history or evidence of previous anteriorsegment intraocular surgery (except cataract) orcorneal abnormalities were excluded.

Central corneal thickness was measured with aspectral domain ocular coherence tomographer HD-OCT pachymeter (HD-OCT-5000; Zeiss, Jena, Ger-many). All eyes remained epithelized and hydratedwith standard isotonic ocular surgical balanced saltsolution (BSS) (Alcon, Ft. Worth, TX). Balanced SaltSolution was used to hydrate the corneal epitheliumbetween measurements before the application offluorescein solution. A 22-gauge needle with Y-adaptor (Saf-T-Intima, Vialon; Becton, Dickinsonand Company, Franklin Lakes, NJ) was inserted intothe anterior chamber via a separate scleral approach.Extreme care was taken with all penetrations of theeye to avoid touching the endothelium, iris, or lens.The entire globe was mounted in the eye stabilizationdevice shown in Figure 4 embedded in moisturizedgauze facing upward to be measured by the Perkinstonometer (Haag Streit USA, Inc., Mason, OH) fittedwith the GAT and CATS prisms. The use of thePerkins tonometer facilitated the globe’s upwardplacement, which minimized measurement errorsdue to the weight and subsequent distortion of theenucleated globe in the ocular stabilization device.This use of the Perkins in enucleated globes increasedthe accuracy of all measurements. The needle IV tubewas connected to a manometric transducer (DwyerInstruments, Michigan City, IN), an isotonic sodiumchloride solution infusion bottle, and an open-airreference tube (Fig. 5).

Multiple stopcocks also were attached to bleed allbubbles from the system, allowing either open orclosed stopcock techniques. The transducer and

Figure 4. Ocular globe stabilizing apparatus for measuring IOP.


McCafferty et al.


anterior chamber were maintained at the same height.The isotonic sodium chloride solution infusion bottlewas attached to a manually driven intravenous polefor bottle height adjustment. After each series ofmeasurements on an eye, the bottle height waslowered to the initial 4.8 cm. The series was onlyaccepted if the initial and closing manometricpressures are within 61 mm Hg. Figure 6 is thephotograph of the schematic in Figure 5.

Results

Finite element modeling simulations analyzed theeffects of corneal properties on the IOP measurement,and made it possible to design of a set of prismcontact surfaces to improve measurement fidelity. Thedesign optimization process used the verified FEM toflatten the isobaric curves of the simulated IOP withrespect to the error-producing GAT parameters. Theobjective was to create a prism surface that isindependent of corneal thickness, rigidity, cornealcurvature, and tear film. The optimal solution, shownin Figure 7 below, is a polynomial approximatedcentral concavity with a peripheral annular convexitythat minimizes the effect of the error causingparameters.

The polynomial in Figure 7 above was determinedto be the ideal design after extensive FEM optimiza-tion. This profile yielded a CCT sensitivity of 5.0 mm

Hg/mm, which is an improvement of 88.2%, com-pared to the GAT flat prism surface (see Fig. 8). Thecontact surface reduces sensitivity to cornea thicknessby structurally supporting the central section oftissue, which causes the stress to be more evenlydistributed, represented in Figure 9. Von Mises stressis represented for this contour in Figures 8 and 9 asfollows: the top bar is the exterior surface, the bottombar is the interior surface, and the center shape is thesagittal cross-section. The highest stress is producedalong the perimeter of the contact surface.

This sensitivity can be further reduced by closelymatching the curvature of the prism to that of thecornea. However, one must be careful to avoidperfectly matching the cornea anterior surface, asthis would cause the applanation area to be met withzero applied force, thereby providing no useablemeasurement.

The contact surface of the prism was optimizedwith FEA. In this process, the surface was modified tominimize the contribution of CCT to the applanationforce. The sensitivities of the measured IOP to CCTfor the standard GAT and CATS prisms are shown inFigure 10. The shallower slope and lower varianceindicate an improvement of measurement accuracy.Finite element analysis indicates that the CATStonometer has a maximum IOP measurement errorof 62 mm Hg due to variations in subject CCTcompared to 6 5 mm Hg with the standard GATprism.

The CATS tonometer also is optimized to reducesensitivity to an individual’s average corneal modulusof elasticity. Young’s modulus or corneal rigidity canvary up to an order of magnitude in individuals and

Figure 5. Experimental set up for cadaver eye IOP comparison. 1,stop-cock. 2, intracameral needle. 3, tonometer prism (Goldmannor CATS). 4, Perkins tonometer armature. 5, whole human globe. 6,ocular stabilizing apparatus. 7, pressure transducer. 8, pressuretransducer signal pressure read-out in mm Hg. 9, manometer tubepressure measurement. 10, intracameral fluid reservoir.

Figure 6. Experimental set up for cadaver eye IOP comparison.


McCafferty et al.


previous studies have demonstrated that it is age-dependent.7 The simulated sensitivity of each prisms’IOP measurement to modulus of elasticity is shown inFigure 11. Again, the shallower slope indicates thatthe CATS prism is less sensitive to this source oferror. Finite element analysis of the CATS prismindicates a maximum IOP measurement error of 62mm Hg maximum error due to variations of cornealYoung’s modulus compared to 68 mm Hg error withthe standard GAT prism.

The sensitivity to Young’s modulus is codepen-dent with CCT; the slope of the sensitivity is

proportional to the CCT. Therefore, it follows thatcorneal rigidity (resistance to deformation) is depen-dent upon the modulus of elasticity and CCT.Corneal rigidity typically is not corrected forclinically, but could cause more significant errorthan CCT in the measurement of IOP. This also wasdemonstrated by Kotecha et al. in 2006.8 The designalso was analyzed for a reduction in cornealcurvature error dependence. The design effectivelyreduced the corneal curvature error from 62.5 mmHg with the GAT prism to 61.5 mm Hg with theCATS prism.

Figure 7. Optimized radial profile of the CATS tonometer prism surface designed to minimize the corneal mechanical and tear filmhydrostatic contributions to measured IOP in applanation tonometry.

Figure 8. Von Mises stress in cornea applanated to a flat contact surface.


McCafferty et al.


The CATS prism theoretically reduces the sensi-tivity of IOP measurement to the tear film. The effectof the tear film was not included in the finite elementanalysis. Tear film modeling was completed usingpublications on capillary pressure calculation of afluid bridge between two curved solids.16,17 TheCATS prism effectively minimized the contributionfrom capillary tear film adhesion by decreasing theradius of curvature of the CATS surface where thetear film bridged between the two solids prism surfaceand cornea. The calculated attractive force created by

the fluid bridge and associated surface tension of thetear film between the tonometer prism and cornea isreduced approximately 45% for the CATS prismcompared to that of the standard GAT. This isachieved by increasing the contact angle between thecornea and contact surface, as illustrated in Figure 12.The range of values shown for the GAT and CATSprisms is due to variations in cornea geometry. Thisreduction in average attractive force is equivalent to areduction of IOP measurement error of approximate-ly 1.5 mm Hg.

Figure 9. Von Mises stress in applanation with CATS polynomial concave–convex contact surface.

Figure 10. Finite element analysis modeling CCT error sensitivity with constant Young’s modulus, constant corneal curvature, andconstant IOP (note reduced slope indicating decreased error sensitivity with the CATS prism).


McCafferty et al.


Human Cadaveric Eye Testing

The three enucleated globes were measured to havea CCT of 641, 786, and 840 lm in eye1, eye2, andeye3, respectively. The intracameral IOP measure-ments were recorded for each of the cadaveric globesshown in Figures 13A to C. While the GAT prism,CATS prism, and Tonopen measurements were

substantially equivalent, the CATS did trend towardimproved accuracy (in this limited sample size) whencompared to true manometric intracameral pressure.The improved accuracy tended to be in thickercorneas. The increased sensitivity of the CATStonometer prism also appears in the cadaveric eyesto improve accuracy at low IOP (,10 mm Hg). Thepurpose of cadaveric manometric referenced testing

Figure 12. Corneal tear film adhesion force error. Note 45% reduction in tear film error in CATS tonometer prism (0.003 N¼0.3 g force¼3 mm Hg on GAT, approximately).

Figure 11. Finite element analysis modeling corneal rigidity (Young’s modulus) error sensitivity with constant corneal thickness,curvature, and IOP (note reduced slope indicating decreased error sensitivity with the CATS prism).


McCafferty et al.


was to demonstrate general equivalence and repeat-ability. The altered CCT seen in even fresh humancadaveric eyes also indicates possible alterations inother biomechanical properties, such as Young’smodulus. To show, significantly improved IOPmeasurement capabilities with reduced error requiresactual clinical conditions on live human eyes.

Discussion

Today virtually all clinicians have the in-officecapability to measure IOP with a GAT, as a majorityof practitioners consider it the most accurate mea-surement modality and it is used as the referencetonometer by the Food and Drug Administration(FDA). Furthermore, it is a required protocol if thereis any question in the IOP measurements made byother devices. However, GAT errors are well knownto the clinician community. As a result, mostclinicians also use an expensive pachymeter to

measure CCT separately for partial correction ofcorneal thickness error although this correction isincomplete given the totality of other errors.

The CATS prism optimizes the shape of the flatGAT prism surface, without any change in measure-ment technique or examination time, and significantlyaddresses all known mechanical errors in GAT IOPmeasurements; corneal thickness, corneal curvature,corneal rigidity, and corneal tear film. It is reasonableto predict an additional reduction in scleral expansionerror (listed as approximately 2% in the GAT prism)in IOP measurement due to decreased volumedisplacement of the cornea and aqueous fluid withthe CATS prism.

The function of the prism, including force topressure conversion supplied by the GAT armature,remains unchanged. The optimized CATS prism isdesigned to measure the same pressure as the flatsurface GAT prism under ‘‘nominal’’ conditions. The‘‘nominal’’ conditions include a standard average

Figure 13. Measurement of IOP in (3) cadaveric human globes (A), (B), and (C), using the GAT prism, CATS prism, and Tonopen versusmanometric intracameral IOP.


McCafferty et al.


corneal thickness, a standard average corneal curva-ture, a standard average corneal rigidity, and astandard average corneal tear film. However, forapproximately 50% of the population, patient corneavariability in each of these parameters has asignificant individual and combined error in GATIOP measurement, as mentioned previously.7–9 Eventhe errors due to corneal thickness alone, which is buta fraction of the total error, were shown to bepotentially sight threatening.10 For this reason, CCTcorrection has been adopted as a standard ofpractice.10 Although several other separate measure-ments and error corrections were proposed, they havebeen too cumbersome to be adopted clinically.12

The published data indicates that the combinederror in IOP measurement can total 615 to 19 mmHg for patients at the extremes of corneal thickness,rigidity, curvature, and tear film.7 The most commonrecognized error is that due to CCT at a published 67mm Hg, which is but a small portion of the totalpossible error. The effect of an error corrected IOP onclinical decisions can be profound. For example, if weonly consider corneal thickness GAT error at 67 mmHg over a standard distribution of varying thickness-es: The average CCT is approximately 556 6 40 lmstandard deviation.10 Using this distribution, thepercentage of people who’s IOP error is greater than62 mm Hg translates to 46% of all patients withsignificant IOP error from CCT alone. Using theCATS tonometer, the number of patients in which theIOP error is greater than 6 2 mm Hg drops to 0%.The CATS tonometer prism may negate the need forpachymetry measurement and CCT correction.

The CATS tonometer prism is designed tosignificantly reduce all of the identified measurementerrors by 50% using the exact same measurementapparatus (with a modified prism) and the sametechnique without any calculations or increased timeand at a minimal cost. The expense incurred by theclinician is limited to only the replacement of theprism. The prism is designed to operate in any existingGAT or Perkins tonometer system.

Clinical studies have noted a �3.5 mm Hg bias inthe GAT prism when compared to true intracameralmanometric pressures and we also found this bias inour FEM analysis.28 The mathematical modelingpredicted this bias and the cadaver eye study andpreliminary clinical evaluation also indicated thisbias. Therefore, to make it equivocal to the GATprism for an average eye with nominal errorcharacteristics, the decision was made also to includethis bias in the CATS prism. This bias equalization in

the design means that a nominal eye with averagecorneal thickness, average corneal rigidity, andaverage corneal curvature will measure the sameIOP with the GAT and CATS prisms. However, TheCATS measurements will differ significantly fromGAT as the error parameters vary from average,which is significant in likely greater than 50% of thepopulation.7–10,12

The GAT and CATS prisms require a centeredcornea on the prism face to accurately measure IOP.The GAT prism will measure applanated miresimaged through the prism anywhere on the flat prismface, but decentration is listed in the GAT instruc-tions for use as necessary for accurate measurement.Conversely, the CATS tonometer, due to its concave–convex, surface will not allow the mires to intersectunless the prism is centered on the cornea. Under thecadaver eye conditions, the prism was easily centeredfor IOP readings. Within the relatively narrow rangethat the CATS mires do intersect, the effect onpressure measurement appears negligible. This insen-sitivity to small amounts of decentration is supportedby the minimal variability in repeat measurementsequivalent to that seen with the GAT prism. Itremains to be seen in clinical studies the effect undersaccadic and nystagmus conditions.

A CATS prism clinical study is proposed todemonstrate the relative improvement in IOP mea-surement between the GAT and CATS prisms whencorrelated to the known error parameters. The studyinvolves examining the difference in CATS and GATmeasurements in correlation to the individual errorparameter (i.e., CCT). An additional clinical studyhas been proposed to determine the absolute IOPmeasurement improvement with the CATS prismusing an intracameral cannulated pressure transducerplaced in the anterior chamber during cataractsurgery.

Current clinical practice does not correct for errorsdue to corneal rigidity, curvature, and tear film.However, the CATS tonometer demonstrates thecapacity to correct for these errors. This is asignificant improvement in the number of patientswho otherwise would be at risk due to inaccurate IOPmeasurement and subsequent undertreatment orovertreatment. Increasing the fidelity of IOP mea-surement greatly improves patient care and safety andsignificantly reduces healthcare cost. The CATS prismcan provide a single error-corrected measurementwith no additional measurement, calculation, orinterpretation error.


McCafferty et al.


Acknowledgments

Supported in part by NIH SBIR Grant R43EY026821-01 and Arizona Eye Consultants, Tucson,AZ with extensive facilities use. All testing accordingto the Declaration of Helsinki document on humanresearch ethics.

Disclosure: S. McCafferty, Intuor Technologies (I,E); G. Lim, Intuor Technologies (E); W. Duncan,

Intuor Technologies (E); E. Enikov, None; J. Schwie-gerling, Intuor Technologies (I)

References

1. Bureau of Labor Statistics, United States, May2014 - Occupational statistics. 29-1041 Optome-trists.

2. Holden B, Resnikoff S. The role of optometry invision 2020. Community Eye Health. 2002;15:33–36.

3. Quigley H, Broman T. The number of peoplewith glaucoma worldwide in 2010 and 2020. Br JOphthalmol. 2006;90:262–267.

4. Susanna R Jr, De Moraes CG, Cioffi GA, RitchR. Why do people (still) go blind from glaucoma?Trans Vis Sci Tech. 2015;4:1.

5. Goldmann H, Schmidt T. Uber Applanationsto-nometrie. Ophthalmologica. 1957;134:221.

6. Holden B. The role of optometry in vision 2020.Community Eye Health. 2005;10:3–8.

7. Liu J, Roberts CJ. Influence of cornea biome-chanical properties on intraocular pressure mea-surement: quantitative analysis. J CataractRefract Surg. 2005;31:146–155.

8. Kotecha A, Elsheikh A, Roberts C, Haogang Z,Garway-Heath D. Corneal thickness- and agerelated biomechanical properties of the corneameasured with the ocular response analyzer.Invest Ophthalmol Vis Sci. 2006;47:5337–5347.

9. Whitacre M, Stein R. Sources of error with use ofGoldmann-type tonometers. Surv Ophthalmol.2002;38:1–30.

10. Kass M, Heuer D, Higginbotham E, et al. TheOcular Hypertension Treatment Study: a ran-domized trial determines that topical ocularhypotensive medication delays or prevents theonset of primary open-angle glaucoma. ArchOphthalmol. 2002;120:701–713.

11. McCafferty SJ, Schwiegerling JT, Enikov ET.Corneal surface asphericity, roughness, and

transverse contraction after uniform scanningexcimer laser ablation. Invest Ophthalmol VisSci. 2012;53:1296–1305.

12. Elsheikh A, Alhasso D, Gunvant P, Garway-Heath D. Multiparameter correction equation forGoldmann applanation tonometry. Optom VisSci. 2011;88:E102–E112.

13. Neuburger M, Maier P, Bohringer D, ReinhardT, Jordan J. The impact of corneal edema onintraocular pressure measurements using gold-mann applanation tonometry, Tono-Pen XL,iCare, and ORA: an in vitro model. J Glaucoma.2013;22:584–590.

14. Orssengo G, Pye D. Determination of the trueintraocular pressure and modulus of elasticity ofthe human cornea in vivo. Bull Math Biol. 1999;61:551–572.

15. McCafferty SJ, Schwiegerling JT, Enikov ET.Thermal load from a CO2 laser radiant energysource induces changes in corneal surface aspher-icity, roughness. Invest Ophthalmol Vis Sci. 2012;53:4279–4288.

16. Tselishchev Y, Val’tsifer V. Influence of the typeof contact between particles joined by a liquidbridge on the capillary cohesive forces. Colloid J.2003;65:385–389.

17. Orr F, Scriven L, Rivas A. Pendular ringsbetween solids: meniscus properties and capillaryforce. J Fluid Mech. 1975;67:723–742.

18. Tesser P. Applanation area in Goldmann appla-nation tonometry. Arch Ophthalmol. 2006;124:1664.

19. Moses R. The Goldmann applanation tonometer.Am J Ophthalmol. 1958;46:865–869.

20. Pandolfi A, Holzapfel GA. Three-dimensionalmodeling and computational analysis of thehuman cornea considering distributed collagenfibril orientations. J Biomed Eng. 2008;130:1–12.

21. Elsheikh A, Ross S, Alhasso D, Rama P.Numerical study of the effect of cornea layeredstructure on ocular biomechanics. Curr Eye Res.2009;34:26–35.

22. Kling S, Marcos S. Finite element modeling ofintrastomal ring segment implantation into ahyperelastic cornea. Invest Ophthalmol Vis Sci.2013;54:881–889.

23. Elsheikh A, Wang D, Kotecha A, Brown M,Garway-Heath D. Evaluation of Goldmannapplanation tonometry using a nonlinear finiteelement ocular model. Ann Biomed Eng. 2006;34:1628–1640.

24. Sr’odka W. Biomechanical model of the humaneyeball and its applications. Optica Applicata.2009;39:401–415.


McCafferty et al.


25. Luce D. Determining in vivo biomechanicalproperties of the cornea with an ocular responseanalyzer. J Cataract Refract Surg. 2005;31:156–162.

26. Schneider E, Grehn F. Intraocular pressuremeasurement: comparison of dynamic contourtonometry and Goldmann applanation tonome-try. J Glaucoma. 2006;15:2–6.

27. Kniestedt C, Nee M, Stamper RL. Dynamiccontour tonometry: a comparative study onhuman cadaver eyes. Arch Ophthalmol. 2004;122:1287–1293.

28. A-Yong Y, Su-Fang D, Yun EZ, et al. Correla-tion between corneal biomechanical properties,applanation tonometry and direct intracameraltonometry. Br J Ophthalmol. 2012;96:640–644.


McCafferty et al.


Documents

Article Goldmann Tonometer Prism with an …doclibrary.com/MSC167/PRM/CATS_TVST4311.pdf1 TVST j 2016 j Vol. 5 j No. 5 j Article 4 This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives