Journal of Refractive Surgery

Biomechanics Supplemental Data

Comparison of Corneal Tomography and a New Combined Tomographic Biomechanical Index in Subclinical Keratoconus

Tommy C.Y. Chan, FRCS; Yu Meng Wang, PhD; Marco Yu, PhD; Vishal Jhanji, MD, FRCOphth

Abstract

PURPOSE:

To investigate and compare the diagnostic ability of corneal tomography and biomechanical and combined parameters for detection of corneal ectasia.

METHODS:

Consecutive patients with subclinical keratoconus (SCKC) and age-matched controls were included. Only one eye from each patient was selected for analysis. The final D value from the Belin/Ambrósio Enhanced Ectasia Display (BAD) was obtained from the Pentacam (Oculus Optikgeräte, Wetzlar, Germany). The tomographic biomechanical index (TBI) was derived from the Pentacam and Corvis ST (Oculus Optikgeräte). Classification analysis between normal and subclinical keratoconus (SCKC) was evaluated using receiver operating characteristic (ROC) curves. The area under the ROC curve (AUC) and partial AUC (pAUC) with specificity of 80% or greater were compared.

RESULTS:

Twenty-three eyes with SCKC and 37 normal eyes were included. All Pentacam-derived parameters (P < .001) and all but two Corvis ST–derived parameters (P < .020) were significantly different between normal and SCKC eyes. A significant difference was found in the final D value (P ≤ .020) and TBI (P ≤ .040) between normal and SCKC eyes. For differentiating normal and SCKC eyes, TBI and BAD final D value demonstrated the highest AUC (0.925 and 0.786, respectively) and pAUC (0.150 and 0.088, respectively). TBI demonstrated 84.4% sensitivity and 82.4% specificity using a cut-off of 0.16. Comparative analysis between these parameters showed that AUC and pAUC of TBI were significantly higher than all parameters from Pentacam (P ≤ .032).

CONCLUSIONS:

In the current study, combined use of tomographic and biomechanical parameters demonstrated a higher capability in differentiating normal and SCKC eyes when compared to tomographic analysis alone.

[J Refract Surg. 2018;34(9):616–621.]

Abstract

PURPOSE:

To investigate and compare the diagnostic ability of corneal tomography and biomechanical and combined parameters for detection of corneal ectasia.

METHODS:

Consecutive patients with subclinical keratoconus (SCKC) and age-matched controls were included. Only one eye from each patient was selected for analysis. The final D value from the Belin/Ambrósio Enhanced Ectasia Display (BAD) was obtained from the Pentacam (Oculus Optikgeräte, Wetzlar, Germany). The tomographic biomechanical index (TBI) was derived from the Pentacam and Corvis ST (Oculus Optikgeräte). Classification analysis between normal and subclinical keratoconus (SCKC) was evaluated using receiver operating characteristic (ROC) curves. The area under the ROC curve (AUC) and partial AUC (pAUC) with specificity of 80% or greater were compared.

RESULTS:

Twenty-three eyes with SCKC and 37 normal eyes were included. All Pentacam-derived parameters (P < .001) and all but two Corvis ST–derived parameters (P < .020) were significantly different between normal and SCKC eyes. A significant difference was found in the final D value (P ≤ .020) and TBI (P ≤ .040) between normal and SCKC eyes. For differentiating normal and SCKC eyes, TBI and BAD final D value demonstrated the highest AUC (0.925 and 0.786, respectively) and pAUC (0.150 and 0.088, respectively). TBI demonstrated 84.4% sensitivity and 82.4% specificity using a cut-off of 0.16. Comparative analysis between these parameters showed that AUC and pAUC of TBI were significantly higher than all parameters from Pentacam (P ≤ .032).

CONCLUSIONS:

In the current study, combined use of tomographic and biomechanical parameters demonstrated a higher capability in differentiating normal and SCKC eyes when compared to tomographic analysis alone.

[J Refract Surg. 2018;34(9):616–621.]

Keratoconus is characterized by progressive corneal thinning leading to irregular astigmatism and visual loss.1 The diagnosis of keratoconus is made by a combination of clinical history and typical corneal tomographic findings.2 Moderate to severe keratoconus is relatively easy to identify, but subclinical disease remains a diagnostic challenge. Early diagnosis of subclinical keratoconus (SCKC) can enhance safety of corneal refractive surgery and detection of progressive changes, which further allows timely application of corneal cross-linking to prevent worsening of the disease.3

Scheimpflug imaging is one of the most commonly used investigative modalities in clinical practice for detection of corneal ectasia. The Pentacam (Oculus Optikgeräte, Wetzlar, Germany) comprises a rotating Scheimpflug camera featuring anterior corneal curvature–based topometric indices and posterior corneal and thickness–based Belin/Ambrósio Enhanced Ectasia Display (BAD) and maximum Ambrósio's relational thickness (ARTmax). It has been reported that up to 10% of SCKC cases were undetected by this technology.4

The Corvis ST (Oculus Optikgeräte), which also uses Scheimpflug imaging, enables measurement of corneal deformation responses under a standardized air-puff indentation to provide biomechanical measurements. Biomechanical abnormalities have been demonstrated in the normal fellow eye of patients with unilateral keratoconus.5 Our previous comparative study between corneal tomography and corneal dynamic response showed fair diagnostic capability for detection of early ectatic changes using parameters obtained from the Corvis ST.6

Recently, a new combined Scheimpflug-based tomographic biomechanical index (TBI) was proposed to enhance detection of corneal ectasia.7 The current study aimed to investigate and compare the diagnostic ability of corneal tomography and TBI to differentiate SCKC from normal eyes.

Patients and Methods

This was a cross-sectional study conducted at the Refractive Surgery Clinic of a university hospital between December 2015 and September 2016. An informed consent was obtained from all study participants. An institutional review board approved the conduct of the study. The study adhered to the tenets of the Declaration of Helsinki.

Consecutive patients with keratoconus were examined. The criteria used for diagnosis of keratoconus included presence of central corneal thinning, conical protrusion, Vogt's striae or Fleischer's ring on slit-lamp examination, topographic pattern consistent with keratoconus, average corneal power greater than 49.00 diopters (D), and higher order aberrations (HOAs) greater than 1.50 μm.8 Eyes with stromal scarring, history of corneal hydrops, or corneal surgery such as cross-linking were excluded from the study.

Eyes that were classified as having SCKC had either atypical or suspect topography findings that did not meet the diagnostic criteria for keratoconus, with average corneal power of 49.00 D or less or HOAs of 1.50 μm or less in either eye or normal topography but obvious keratoconus in the contralateral eye.9

Age-matched participants with healthy corneas were recruited from a refractive surgery clinic. These participants had no ocular abnormality except myopia and myopic astigmatism, a corrected distance visual acuity of 20/20 or better, and a stable refraction for more than 1 year. Participants with unremarkable slit-lamp examination and normal topography, including a typical axial topography pattern and an average corneal power of 47.75 D or less, were included.

All patients were advised to discontinue rigid contact lens wear for 4 weeks and soft contact lens wear for 2 weeks before ocular examination. A single experienced investigator (YMW) obtained all measurements between 10 AM and 4 PM.

Corneal Tomography

The Pentacam was used to obtain corneal tomography images for all study participants. It captures 100 slit images in 2 seconds with a slit depth of 14 mm by rotating along the optical axis from 0° to 360°. Pentacam's digital camera (1.45-megapixel) and slit illumination system (475-nm monochromatic slit of light) automatically rotates around the corneal apex to capture cross-sectional Scheimpflug images of the anterior eye, each separated by 3.6°.

The following data were obtained from the Pentacam: average keratometry (Km), astigmatism, best fit sphere (BFS) from the anterior and posterior corneal surface, minimum corneal thickness (CTmin), corneal thickness at apex (CTapex), index of surface variance (ISV), index of vertical asymmetry (IVA), keratoconus index (KI), center keratoconus index (CKI), ARTmax, and BAD.

Corneal Deformation Response and TBI

The corneal deformation response was captured using the Corvis ST. It captures the corneal deformation response under an air-puff (size: 3.06 mm; intraocular pressure: 60 mm Hg) with a horizontal distance of 8 mm centered at the apex of the cornea. A total of approximately 140 cross-sectional images of the cornea are collected over a metered collimated air-puff for 30 ms. By tracing the corneal boundaries at individual image frames, corneal deformation response parameters are automatically measured by a built-in software.

The following predefined parameters were analyzed in the Corvis ST: time from the initiation of air-puff until the first (A1T) and second (A2T) applanation; corneal velocity during the first (A1V) and second (A2V) applanation; maximum deformation amplitude (HCDA); time of maximum deformation (T HC); radius of curvature at highest concavity (RC); peak distance (PD) and integrated radius (IR); deflection amplitude of the first and second applanation; and highest concavity (A1DfA, A2DfA, and HCDfA, respectively). Deflection represents movement of the cornea, and deformation describes deflection together with whole eye movement.

In addition, the following parameters were measured: maximal change of arc length, which describes the change in arc length within 3.5 mm in both directions from corneal apex at the time of highest concavity (CTapex); DA ratio 1 and 2, which represent maximum value of the ratio between deformation amplitude at the apex and at 1 and 2 mm from central cornea, respectively; maximum inverse radius (Max Inv Rad), which describes the maximal value of radius of curvature at the time of highest concavity; stiffness parameter at first applanation (SPA1), which is defined as resultant pressure (adjusted pressure at A1 minus a biomechanically corrected intraocular pressure value) divided by deflection amplitude at A1; and horizontal Ambrósio's relational thickness value (ARTh) and TBI, which is a combined parameter based on Scheimpflug-based corneal tomography and biomechanical assessments.

Statistical Analysis

Only one eye with the lower average keratometry was selected for analysis for normal participants. As for patients with subclinical keratoconus, the less severe eye with lower average keratometry value was selected.

Statistical analysis was performed using R 3.2.5 software (R Foundation, Vienna, Austria). Classification analyses between normal and subclinical keratoconus were evaluated using receiver operating characteristic (ROC) curves. The area under the ROC curve (AUC) and partial AUC (pAUC) with specificity of 80% or greater for each classifying parameter was compared based on bootstrap resampling with 200 replicates, one eye from each participant was sampled with replacement in each bootstrap replicate. AUCs provide an overall comparison of the whole ROC curves with an AUC of 1 representing a perfect classification at any level of specificity. A common classifying system in which AUC 0.5 to 0.6 = fail; 0.6 to 0.7 = poor; 0.7 to 0.8 = fair; 0.8 to 0.9 = good; and 0.9 to 1.0 = excellent is adopted. Although pAUCs focus on the comparison of ROC curves at the sector with specificity of 80% or greater, a pAUC of 0.2 represents a perfect classification at any level of specificity of 80% or greater. Because a low level of specificity is usually irrelevant to practical use, it is believed that the pAUC is more relevant to clinical practice. The mean and median values for each parameter measured for all eyes were estimated. Comparison of median values in each parameter between groups was performed using Mann–Whitney U tests. A P value of less than .05 was considered statistically significant.

Results

The study included 23 eyes with subclinical keratoconus and 37 normal eyes. The mean age was 32.4 ± 8.4 years (range: 16 to 53 years) with no difference between groups (P = .779). Parameters obtained from the Pentacam are shown in Table A(available in the online version of this article). There was no difference in Km, astigmatism, BFS from the anterior and posterior cornea, and CKI and KI between normal and SCKC eyes (P ≥ .097). Significant differences were found in ISV, IVA, CTmin, CTapex, ARTmax, and final D value of BAD between normal and SCKC eyes (P ≤ .007). Parameters measured with the Corvis ST and the combined TBI are shown in Table A. Significant differences were found in A1T, A1V, DA ratio 1, DA ratio 2, ARTh, RC, IR, Max Inv Rad, SPA1, TBI, and CBI between normal and SCKC eyes (P ≤ .011).

Parameters Obtained for Normal and Subclinical Keratoconus

Table A:

Parameters Obtained for Normal and Subclinical Keratoconus

Parameters from the Corvis ST and Pentacam were analyzed for differentiating normal and SCKC eyes. The TBI and BAD final D value demonstrated the highest AUC (0.925 and 0.786, respectively) and pAUC (0.150 and 0.088, respectively) from the two devices (Table 1). Comparative analysis between these parameters showed that AUC and pAUC of TBI were significantly higher than all five parameters from the Pentacam (P ≤ .032). On the other hand, BAD final D value demonstrated a similar AUC and pAUC compared to the remaining parameters from the Corvis ST (P ≥ .209) except ARTh (Table 2). The ROC curves for BAD final D value and TBI are shown in Figure 1.

AUC and pAUC With Specificity ≥ 80% for Classification Between Normal and Subclinical Keratoconus

Table 1:

AUC and pAUC With Specificity ≥ 80% for Classification Between Normal and Subclinical Keratoconus

Comparisons of AUC and pAUC Between Pentacam and Corvis ST Parameters and the Combined TBI for Differentiating Subclinical Keratoconus From Normal

Table 2:

Comparisons of AUC and pAUC Between Pentacam and Corvis ST Parameters and the Combined TBI for Differentiating Subclinical Keratoconus From Normal

Receiver operating curve for the Belin/Ambrósio Enhanced Ectasia Display (BAD) final D value from Pentacam and combined tomographical biomechanical index (TBI) from the Corvis ST and Pentacam. The Pentacam and Corvis ST are manufactured by Oculus Optikgeräte, Wetzlar, Germany. CBI = Combined Biomechanical Index

Figure 1.

Receiver operating curve for the Belin/Ambrósio Enhanced Ectasia Display (BAD) final D value from Pentacam and combined tomographical biomechanical index (TBI) from the Corvis ST and Pentacam. The Pentacam and Corvis ST are manufactured by Oculus Optikgeräte, Wetzlar, Germany. CBI = Combined Biomechanical Index

For the 23 eyes with SCKC recruited in this study, the average corrected distance visual acuity was −0.05 ± 0.06 logMAR. The average spherical equivalent manifest refraction and astigmatism were −4.73 ± 3.98 and −1.19 ± 1.23 D, respectively.

Discussion

The current study showed that the BAD final D value and TBI had high diagnostic capability to differentiate keratoconus from normal cornea. As for the differentiation of SCKC from normal, the TBI had a better diagnostic capability compared to the BAD. Scheimpflug tomography has been used widely for diagnosis of corneal ectatic disorders. Previous studies have reported that the BAD final D value had 100% sensitivity and 100% specificity in differentiating keratoconic eyes from normal eyes using a cut-off of 2.1. However, the sensitivity and specificity dropped to 60% and 90%, respectively, in cases with SCKC.10 Another study demonstrated that both the BAD final D value and ARTmax had an AUC of 0.95 for differentiating keratoconic eyes from normal eyes, whereas the corresponding AUC was 0.93 for differentiating SCKC from normal eyes.4 In eyes with forme fruste keratoconus, which was defined as the fellow eye of clinical keratoconus with normal clinical and corneal tomographical findings,11 our previous study reported an AUC of 0.76 for detection of forme fruste keratoconus from normal eyes with a sensitivity of 53% and specificity of 80%.6 On the other hand, keratometric asymmetry and topometric index were shown to be better in discriminating forme fruste cases from normal rather than the BAD and ARTmax in another comparative study.12 It is important to understand that there is limited agreement with the use of the terms “subclinical,” “forme fruste,” and “unilateral” keratoconus in the literature.13 It has been suggested that signs of keratoconus will develop over time in most cases of unilateral keratoconus.14

Corneal biomechanical failure seems to form the basis of keratoconus,15 and the ability to quantify corneal biomechanical properties may therefore help in its early detection. Iatrogenic keratectasia is also regarded as biomechanical decompensation after refractive surgery in susceptible eyes. Hence, detection of biomechanical abnormality together with tomographical findings could enhance preoperative screening and safety of the procedure. The Corvis ST is a commercially available imaging device capable of monitoring corneal deformation during a constant air-puff applanation with an aim to measure parameters that denote the biomechanical properties of the cornea. Previous studies using the Corvis ST have reported a greater HCDA in keratoconic corneas compared to normal corneas with AUCs ranging from 0.77 to 0.88.16–20 In vivo biomechanical analysis using a single parameter may not be sufficient for diagnosis of keratoconus.21 Therefore, the combined Corvis Biomechanical Index (CBI) was created, aiming to have a better diagnostic capacity for detection of SCKC.22 The CBI was able to demonstrate biomechanical abnormality in eyes with forme fruste keratoconus that had normal tomographic findings.5 Our previous comparative study between corneal dynamic and tomographic analysis showed similar diagnostic efficacy between the CBI and final D of the BAD in separating forme fruste keratoconus from normal eyes. Comparable AUC was observed between the CBI (AUC = 0.785) and BAD final D value (AUC = 0.757) with sensitivities of 63% and 53% given a common specificity of 80%.6

The TBI is a combined parameter based on Scheimpflug-based corneal tomography and biomechanical assessments. It was generated by the random forest method with the leave-one-out cross-validation method, and was demonstrated to enhance the capacity for detecting SCKC with an AUC of 0.985 and 0.839 for the TBI and BAD final D value, respectively.7 We found that the AUC for the TBI (0.925) was significantly higher than the BAD final D value (0.786), supporting the enhanced diagnostic capability of the TBI for subclinical keratoconus. The current study provided evidence to support the use of the TBI in early diagnosis of corneal ectasia, especially in refractive surgery screening.

The current study was limited by its small sample size. However, we demonstrated the feasibility of use of combined tomographic and corneal dynamic analysis for differentiating normal and SCKC eyes. Future studies are warranted to corroborate the findings of our study.

The topography findings of SCKC are presented in Figures AW (available in the online version of this article).

Corneal topography findings of patient 1 with subclinical keratoconus.

Figure A.

Corneal topography findings of patient 1 with subclinical keratoconus.

Corneal topography findings of patient 2 with subclinical keratoconus.

Figure B.

Corneal topography findings of patient 2 with subclinical keratoconus.

Corneal topography findings of patient 3 with subclinical keratoconus.

Figure C.

Corneal topography findings of patient 3 with subclinical keratoconus.

Corneal topography findings of patient 4 with subclinical keratoconus.

Figure D.

Corneal topography findings of patient 4 with subclinical keratoconus.

Corneal topography findings of patient 5 with subclinical keratoconus.

Figure E.

Corneal topography findings of patient 5 with subclinical keratoconus.

Corneal topography findings of patient 6 with subclinical keratoconus.

Figure F.

Corneal topography findings of patient 6 with subclinical keratoconus.

Corneal topography findings of patient 7 with subclinical keratoconus.

Figure G.

Corneal topography findings of patient 7 with subclinical keratoconus.

Corneal topography findings of patient 8 with subclinical keratoconus.

Figure H.

Corneal topography findings of patient 8 with subclinical keratoconus.

Corneal topography findings of patient 9 with subclinical keratoconus.

Figure I.

Corneal topography findings of patient 9 with subclinical keratoconus.

Corneal topography findings of patient 10 with subclinical keratoconus.

Figure J.

Corneal topography findings of patient 10 with subclinical keratoconus.

Corneal topography findings of patient 11 with subclinical keratoconus.

Figure K.

Corneal topography findings of patient 11 with subclinical keratoconus.

Corneal topography findings of patient 12 with subclinical keratoconus.

Figure L.

Corneal topography findings of patient 12 with subclinical keratoconus.

Corneal topography findings of patient 13 with subclinical keratoconus.

Figure M.

Corneal topography findings of patient 13 with subclinical keratoconus.

Corneal topography findings of patient 14 with subclinical keratoconus.

Figure N.

Corneal topography findings of patient 14 with subclinical keratoconus.

Corneal topography findings of patient 15 with subclinical keratoconus.

Figure O.

Corneal topography findings of patient 15 with subclinical keratoconus.

Corneal topography findings of patient 16 with subclinical keratoconus.

Figure P.

Corneal topography findings of patient 16 with subclinical keratoconus.

Corneal topography findings of patient 17 with subclinical keratoconus.

Figure Q.

Corneal topography findings of patient 17 with subclinical keratoconus.

Corneal topography findings of patient 18 with subclinical keratoconus.

Figure R.

Corneal topography findings of patient 18 with subclinical keratoconus.

Corneal topography findings of patient 19 with subclinical keratoconus.

Figure S.

Corneal topography findings of patient 19 with subclinical keratoconus.

Corneal topography findings of patient 20 with subclinical keratoconus.

Figure T.

Corneal topography findings of patient 20 with subclinical keratoconus.

Corneal topography findings of patient 21 with subclinical keratoconus.

Figure U.

Corneal topography findings of patient 21 with subclinical keratoconus.

Corneal topography findings of patient 22 with subclinical keratoconus.

Figure V.

Corneal topography findings of patient 22 with subclinical keratoconus.

Corneal topography findings of patient 23 with subclinical keratoconus.

Figure W.

Corneal topography findings of patient 23 with subclinical keratoconus.

References

  1. Jhanji V, Sharma N, Vajpayee RB. Management of keratoconus: current scenario. Br J Ophthalmol. 2011;95:1044–1050. doi:10.1136/bjo.2010.185868 [CrossRef]
  2. Naderan M, Shoar S, Kamaleddin MA, Rajabi MT, Naderan M, Khodadadi M. Keratoconus clinical findings according to different classifications. Cornea. 2015;34:1005–1011. doi:10.1097/ICO.0000000000000537 [CrossRef]
  3. Kymionis GD, Kontadakis GA, Hashemi KK. Accelerated versus conventional corneal crosslinking for refractive instability: an update. Curr Opin Ophthalmol. 2017;28:343–347. doi:10.1097/ICU.0000000000000375 [CrossRef]
  4. Ruiseñor Vázquez PR, Galletti JD, Minguez N, et al. Pentacam Scheimpflug tomography findings in topographically normal patients and subclinical keratoconus cases. Am J Ophthalmol. 2014;158:32–40. doi:10.1016/j.ajo.2014.03.018 [CrossRef]
  5. Vinciguerra R, Ambrósio R Jr, Roberts CJ, Azzolini C, Vinciguerra P. Biomechanical characterization of subclinical keratoconus without topographic or tomographic abnormalities. J Refract Surg. 2017;33:399–407. doi:10.3928/1081597X-20170213-01 [CrossRef]
  6. Wang YM, Chan TCY, Yu M, Jhanji V. Comparison of corneal dynamic and tomographic analysis in normal, forme fruste keratoconic, and keratoconic eyes. J Refract Surg. 2017;33:632–638. doi:10.3928/1081597X-20170621-09 [CrossRef]
  7. Ambrósio R Jr, Lopes BT, Faria-Correia F, et al. Integration of Scheimpflug-based corneal tomography and biomechanical assessments for enhancing ectasia detection. J Refract Surg. 2017;33:434–443. doi:10.3928/1081597X-20170426-02 [CrossRef]
  8. McMahon TT, Szczotka-Flynn L, Barr JT, et al. A new method for grading the severity of keratoconus: the Keratoconus Severity Score (KSS). Cornea. 2006;25:794–800. doi:10.1097/01.ico.0000226359.26678.d1 [CrossRef]
  9. Klyce SD. Chasing the suspect: keratoconus. Br J Ophthalmol. 2009;93:845–847. doi:10.1136/bjo.2008.147371 [CrossRef]
  10. Muftuoglu O, Ayar O, Hurmeric V, Orucoglu F, Kilic I. Comparison of multimetric D index with keratometric, pachymetric, and posterior elevation parameters in diagnosing subclinical keratoconus in fellow eyes of asymmetric keratoconus patients. J Cataract Refract Surg. 2015;41:557–565. doi:10.1016/j.jcrs.2014.05.052 [CrossRef]
  11. Saad A, Gatinel D. Topographic and tomographic properties of forme fruste keratoconus corneas. Invest Ophthalmol Vis Sci. 2010;51:5546–5555. doi:10.1167/iovs.10-5369 [CrossRef]
  12. Bae GH, Kim JR, Kim CH, Lim DH, Chung ES, Chung TY. Corneal topographic and tomographic analysis of fellow eyes in unilateral keratoconus patients using Pentacam. Am J Ophthalmol. 2014;157:103–109. doi:10.1016/j.ajo.2013.08.014 [CrossRef]
  13. Gomes JA, Tan D, Rapuano CJ, et al. Global consensus on keratoconus and ectatic diseases. Cornea. 2015;34:359–369. doi:10.1097/ICO.0000000000000408 [CrossRef]
  14. Li X, Rabinowitz YS, Rasheed K, Yang H. Longitudinal study of the normal eyes in unilateral keratoconus patients. Ophthalmology. 2004;111:440–446. doi:10.1016/j.ophtha.2003.06.020 [CrossRef]
  15. Scarcelli G, Besner S, Pineda R, Yun SH. Biomechanical characterization of keratoconus corneas ex vivo with Brillouin microscopy. Invest Ophthalmol Vis Sci. 2014;55:4490–4495. doi:10.1167/iovs.14-14450 [CrossRef]
  16. Ali NQ, Patel DV, McGhee CN. Biomechanical responses of healthy and keratoconic corneas measured using a noncontact Scheimpflug-based tonometer. Invest Ophthalmol Vis Sci. 2014;55:3651–3659. doi:10.1167/iovs.13-13715 [CrossRef]
  17. Tian L, Ko MW, Wang LK, et al. Assessment of ocular biomechanics using dynamic ultra high-speed Scheimpflug imaging in keratoconic and normal eyes. J Refract Surg. 2014;30:785–791. doi:10.3928/1081597X-20140930-01 [CrossRef]
  18. Bak-Nielsen S, Pedersen IB, Ivarsen A, Hjortdal J. Dynamic Scheimpflug-based assessment of keratoconus and the effects of corneal cross-linking. J Refract Surg. 2014;30:408–414. doi:10.3928/1081597X-20140513-02 [CrossRef]
  19. Peña-Garcia P, Peris-Martínez C, Abbouda A, Ruiz-Moreno JM. Detection of subclinical keratoconus through non-contact tonometry and the use of discriminant biomechanical functions. J Biomech. 2016;49:353–363. doi:10.1016/j.jbiomech.2015.12.031 [CrossRef]
  20. Chan TC, Wang YM, Yu M, Jhanji V. Comparison of corneal dynamic parameters and tomographic measurements using Scheimpflug imaging in keratoconus. Br J Ophthalmol. 2018;102:42–47. doi:10.1136/bjophthalmol-2017-310355 [CrossRef]
  21. Steinberg J, Katz T, Lucke K, Frings A, Druchkiv V, Linke SJ. Screening for keratoconus with new dynamic biomechanical in vivo scheimpflug analyses. Cornea. 2015;34:1404–1412. doi:10.1097/ICO.0000000000000598 [CrossRef]
  22. Vinciguerra R, Ambrósio R Jr, Elsheikh A, et al. Detection of keratoconus with a new biomechanical index. J Refract Surg. 2016;32:803–810. doi:10.3928/1081597X-20160629-01 [CrossRef]

AUC and pAUC With Specificity ≥ 80% for Classification Between Normal and Subclinical Keratoconus

ParameteraAUCpAUCCut-offSpecificitySensitivity
TBI0.9250.1500.1682.4%84.4%
Corvis ST
  ARTh0.8360.129444.082.4%81.3%
  Max inv rad0.7540.0790.1982.4%59.4%
  A1T0.7500.0527.1882.4%46.9%
  RC0.7360.0946.7882.4%62.5%
Pentacam
  BAD final D0.7860.0881.3885.3%53.1%
  IVA0.7810.1250.1588.2%68.8%
  ARTmax0.7590.095386.582.4%65.6%
  CTapex0.7220.058534.582.4%37.5%
  CTmin0.7100.059529.582.4%43.8%

Comparisons of AUC and pAUC Between Pentacam and Corvis ST Parameters and the Combined TBI for Differentiating Subclinical Keratoconus From Normal

PTBI/Corvis STaPentacama

BAD Final DIVAARTmaxCTapexCTmin
AUC comparisonsTBI.001.032.001.002< .001
ARTh.209.527.078.034.024
Max inv rad.616.760.941.658.574
A1T.558.696.900.662.529
RC.399.638.743.845.729
pAUC comparisonsTBI< .001.217.008< .001< .001
ARTh.028.853.110.001< .001
Max inv rad.705.085.571.375.418
A1T.111.016.127.839.788
RC.806.231.944.154.166

Parameters Obtained for Normal and Subclinical Keratoconus

ParameterNormalSubclinical KeratoconusPa


Mean ± SDMedianRangeMean ± SDMedianRange
Pentacam
  Km42.92 ± 0.9642.9541.15 to 44.7543.72 ± 2.1543.3040.05 to 47.5.097
  Astigmatism1.37 ± 0.661.150.10 to 2.901.22 ± 0.850.950.30 to 3.70.399
  CTapex567.0 ± 28.8570.5505.0 to 637.0542.9 ± 25.7539.0503.0 to 593.0< .001
  CTmin561.5 ± 29.1564.0500.0 to 635.0537.3 ± 26.1532.0494 to 588< .001
  BFS Front7.95 ± 0.187.967.62 to 8.267.83 ± 0.377.877.16 to 8.49.135
  BFS Back6.48 ± 0.196.486.16 to 6.906.39 ± 0.346.455.75 to 6.94.292
  ISV17.7 ± 4.717.09 to 2922.3 ± 8.320.010 to 42.007
  IVA0.11 ± 0.040.100.04 to 0.260.20 ± 0.090.190.05 to 0.43< .001
  KI1.02 ± 0.021.030.98 to 1.081.04 ± 0.031.030.98 to 1.10.164
  CKI1.01 ± 0.011.011.00 to 1.021.01 ± 0.011.011.00 to 1.04.173
  ARTmax455.6 ± 73.4440.0334.0 to 656.0373.3 ± 74.4354.0210.0 to 489.0< .001
  BAD final D0.931 ± 0.5370.930−0.71 to 2.031.696 ± 0.7231.7550.43 to 2.92< .001
TBI/Corvis ST
  A1DfA0.106 ± 0.0060.1050.095 to 0.1220.107 ± 0.0080.1050.095 to 0.124.817
  A1T7.375 ± 0.2787.3036.832 to 8.2767.177 ± 0.1367.1776.655 to 7.486< .001
  A1V0.159 ± 0.0170.1650.107 to 0.1840.170 ± 0.0180.1700.124 to 0.220.011
  A2DfA0.138 ± 0.0810.1210.100 to 0.6300.236 ± 0.2380.1180.096 to 0.878.997
  A2T21.736 ± 0.54721.86919.122 to 22.72621.384 ± 1.19521.87318.358 to 22.742.434
  A2V (m/s)−0.289 ± 0.041−0.291−0.352 to −0.144−0.250 ± 0.080−0.285−0.403 to −0.033.126
  HCDA1.096 ± 0.0901.1030.880 to 1.3341.124 ± 0.0731.1310.962 to 1.371.062
  HCDfA0.927 ± 0.0830.9340.764 to 1.1180.978 ± 0.1190.9520.794 to 1.356.143
  DA ratio 11.612 ± 0.1461.5951.456 to 2.4901.634 ± 0.0711.6141.518 to 1.847.007
  DA ratio 24.664 ± 1.0094.4903.600 to 10.5085.298 ± 1.5054.8053.54 to 11.148.006
  T HC16.689 ± 0.53616.86315.015 to 17.32516.592 ± 0.46516.63215.708 to 17.556.780
  RC7.418 ± 0.8517.4665.330 to 9.2636.301 ± 1.5386.3212.201 to 8.515< .001
  PD5.128 ± 0.2275.1334.551 to 5.6005.138 ± 0.2675.1394.637 to 5.728.944
  IR8.088 ± 0.8488.0666.094 to 9.5989.312 ± 1.7318.7127.212 to 14.211< .001
  Max Inv Rad0.187 ± 0.0550.1710.145 to 0.4200.281 ± 0.1430.2020.155 to 0.612< .001
  ARTh545.63 ± 98.04547.94392.32 to 811.36410.62 ± 98.91392.50224.03 to 630.73< .001
  SPA199.619 ± 16.19197.97765.328 to 127.39688.834 ± 16.17091.57314.497 to 132.03.007
  TBI0.097 ± 0.0990.0600.010 to 0.5100.633 ± 0.3730.8700.070 to 1.000< .001
  CBI0.040 ± 0.0860.0010.000 to 0.3480.538 ± 0.4610.8070.000 to 1.000< .001
Authors

From the Department of Ophthalmology & Visual Sciences, The Chinese University of Hong Kong, Hong Kong SAR, China (TCYC, YMW, VJ); the Department of Ophthalmology, Hong Kong Sanatorium & Hospital, Hong Kong SAR, China (TCYC); the Department of Mathematics and Statistics, Hang Seng Management College, Hong Kong SAR, China (MY); and UPMC Eye Center, University of Pittsburgh, Pittsburgh, Pennsylvania (VJ).

The authors have no financial or proprietary interest in the materials presented herein.

AUTHOR CONTRIBUTIONS

Study concept and design (VJ); data collection (YMW); analysis and interpretation of data (TCYC, MY); writing the manuscript (TCYC, YMW, VJ); critical revision of the manuscript (MY, VJ); statistical expertise (MY); administrative, technical, or material support (VJ); supervision (VJ)

Correspondence: Vishal Jhanji, MD, FRCOphth, Department of Ophthalmology and Visual Sciences, The Chinese University of Hong Kong, Hong Kong SAR, China. E-mail: jhanjiv@upmc.edu

Received: March 13, 2018
Accepted: June 20, 2018

10.3928/1081597X-20180705-02

Sign up to receive

Journal E-contents