A Coincident Thinning Index for Keratoconus Identification Using OCT Pachymetry and Epithelial Thickness Maps

Inter-zonal epithelial thickness differences for early keratoconus detection using optical coherence tomography

PURPOSE

To develop and test a parameter for early keratoconus screening by quantifying the localized epithelial thickness differences in keratoconic eyes.

METHODS

The cross-sectional study included 189 eyes of 116 subjects in total: 86 eyes of 54 keratoconus patients with bilateral ectasia and 40 eyes of 20 healthy subjects in the parameter-development dataset and 42 eyes of 21 keratoconus patients with asymmetric ectasia and 21 eyes of 21 healthy subjects in the parameter-validation dataset. Epithelial thickness maps were obtained using anterior segment optical coherence tomography and the inter-zonal epithelial thickness differences were calculated. The developed parameter was tested in keratoconus patients with asymmetric ectasia.

RESULTS

Compared to healthy controls, the inferior-temporal and global inter-zonal epithelial thickness differences were higher not only in eyes with tomographically significant keratoconus (median [interquartile range] of 4.42 [3.13] µm vs. 0.78 [0.42] µm, p < 0.001, and 3.05 [1.51] µm vs. 1.07 [0.26] µm, p < 0.001, respectively), but also in tomographically normal keratoconus fellow eyes (1.36 [0.85] µm vs. 0.78 [0.42] µm, p = 0.005, and 1.31 [0.32] µm vs. 1.07 [0.26] µm, p = 0.01, respectively). The inferior-temporal inter-zonal epithelial thickness differences had an area under the receiver operating characteristic curve (95% confidence interval) of 0.991 (0.972-1) for detecting tomographically significant keratoconus and 0.749 (0.598-0.901) for differentiating between tomographically normal keratoconus fellow eyes and healthy controls.

CONCLUSIONS

The inter-zonal epithelial thickness differences are increased in keratoconus fellow eyes which still have a normal Scheimpflug corneal tomography, and therefore may serve as a useful parameter to detect early ectatic changes.

Corneal Epithelial Thickness Mapping: A Major Review

The corneal epithelium (CE) is the outermost layer of the cornea with constant turnover, relative stability, remarkable plasticity, and compensatory properties to mask alterations in the underlying stroma. The advent of quantitative imaging modalities capable of producing epithelial thickness mapping (ETM) has made it possible to characterize better the different patterns of epithelial remodeling. In this comprehensive synthesis, we reviewed all available data on ETM with different methods, including very high-frequency ultrasound (VHF-US) and spectral-domain optical coherence tomography (SD-OCT) in normal individuals, corneal or systemic diseases, and corneal surgical scenarios. We excluded OCT studies that manually measured the corneal epithelial thickness (CET) (e.g., by digital calipers) or the CE (e.g., by confocal scanning or handheld pachymeters). A comparison of different CET measuring technologies and devices capable of producing thickness maps is provided. Normative data on CET and the possible effects of gender, aging, diurnal changes, refraction, and intraocular pressure are discussed. We also reviewed ETM data in several corneal disorders, including keratoconus, corneal dystrophies, recurrent epithelial erosion, herpes keratitis, keratoplasty, bullous keratopathy, carcinoma in situ, pterygium, and limbal stem cell deficiency. The available data on the potential role of ETM in indicating refractive surgeries, planning the procedure, and assessing postoperative changes are reviewed. Alterations in ETM in systemic and ocular conditions such as eyelid abnormalities and dry eye disease and the effects of contact lenses, topical medications, and cataract surgery on the ETM profile are discussed.

Artificial intelligence for detecting keratoconus

BACKGROUND

Keratoconus remains difficult to diagnose, especially in the early stages. It is a progressive disorder of the cornea that starts at a young age. Diagnosis is based on clinical examination and corneal imaging; though in the early stages, when there are no clinical signs, diagnosis depends on the interpretation of corneal imaging (e.g. topography and tomography) by trained cornea specialists. Using artificial intelligence (AI) to analyse the corneal images and detect cases of keratoconus could help prevent visual acuity loss and even corneal transplantation. However, a missed diagnosis in people seeking refractive surgery could lead to weakening of the cornea and keratoconus-like ectasia. There is a need for a reliable overview of the accuracy of AI for detecting keratoconus and the applicability of this automated method to the clinical setting.

OBJECTIVES

To assess the diagnostic accuracy of artificial intelligence (AI) algorithms for detecting keratoconus in people presenting with refractive errors, especially those whose vision can no longer be fully corrected with glasses, those seeking corneal refractive surgery, and those suspected of having keratoconus. AI could help ophthalmologists, optometrists, and other eye care professionals to make decisions on referral to cornea specialists. Secondary objectives To assess the following potential causes of heterogeneity in diagnostic performance across studies. • Different AI algorithms (e.g. neural networks, decision trees, support vector machines) • Index test methodology (preprocessing techniques, core AI method, and postprocessing techniques) • Sources of input to train algorithms (topography and tomography images from Placido disc system, Scheimpflug system, slit-scanning system, or optical coherence tomography (OCT); number of training and testing cases/images; label/endpoint variable used for training) • Study setting • Study design • Ethnicity, or geographic area as its proxy • Different index test positivity criteria provided by the topography or tomography device • Reference standard, topography or tomography, one or two cornea specialists • Definition of keratoconus • Mean age of participants • Recruitment of participants • Severity of keratoconus (clinically manifest or subclinical) SEARCH METHODS: We searched CENTRAL (which contains the Cochrane Eyes and Vision Trials Register), Ovid MEDLINE, Ovid Embase, OpenGrey, the ISRCTN registry, ClinicalTrials.gov, and the World Health Organization International Clinical Trials Registry Platform (WHO ICTRP). There were no date or language restrictions in the electronic searches for trials. We last searched the electronic databases on 29 November 2022.

SELECTION CRITERIA

We included cross-sectional and diagnostic case-control studies that investigated AI for the diagnosis of keratoconus using topography, tomography, or both. We included studies that diagnosed manifest keratoconus, subclinical keratoconus, or both. The reference standard was the interpretation of topography or tomography images by at least two cornea specialists.

DATA COLLECTION AND ANALYSIS

Two review authors independently extracted the study data and assessed the quality of studies using the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool. When an article contained multiple AI algorithms, we selected the algorithm with the highest Youden's index. We assessed the certainty of evidence using the GRADE approach.

MAIN RESULTS

We included 63 studies, published between 1994 and 2022, that developed and investigated the accuracy of AI for the diagnosis of keratoconus. There were three different units of analysis in the studies: eyes, participants, and images. Forty-four studies analysed 23,771 eyes, four studies analysed 3843 participants, and 15 studies analysed 38,832 images. Fifty-four articles evaluated the detection of manifest keratoconus, defined as a cornea that showed any clinical sign of keratoconus. The accuracy of AI seems almost perfect, with a summary sensitivity of 98.6% (95% confidence interval (CI) 97.6% to 99.1%) and a summary specificity of 98.3% (95% CI 97.4% to 98.9%). However, accuracy varied across studies and the certainty of the evidence was low. Twenty-eight articles evaluated the detection of subclinical keratoconus, although the definition of subclinical varied. We grouped subclinical keratoconus, forme fruste, and very asymmetrical eyes together. The tests showed good accuracy, with a summary sensitivity of 90.0% (95% CI 84.5% to 93.8%) and a summary specificity of 95.5% (95% CI 91.9% to 97.5%). However, the certainty of the evidence was very low for sensitivity and low for specificity. In both groups, we graded most studies at high risk of bias, with high applicability concerns, in the domain of patient selection, since most were case-control studies. Moreover, we graded the certainty of evidence as low to very low due to selection bias, inconsistency, and imprecision. We could not explain the heterogeneity between the studies. The sensitivity analyses based on study design, AI algorithm, imaging technique (topography versus tomography), and data source (parameters versus images) showed no differences in the results.

AUTHORS' CONCLUSIONS

AI appears to be a promising triage tool in ophthalmologic practice for diagnosing keratoconus. Test accuracy was very high for manifest keratoconus and slightly lower for subclinical keratoconus, indicating a higher chance of missing a diagnosis in people without clinical signs. This could lead to progression of keratoconus or an erroneous indication for refractive surgery, which would worsen the disease. We are unable to draw clear and reliable conclusions due to the high risk of bias, the unexplained heterogeneity of the results, and high applicability concerns, all of which reduced our confidence in the evidence. Greater standardization in future research would increase the quality of studies and improve comparability between studies.

New keratoconus staging system based on OCT

PURPOSE

To establish a numerical spectral-domain optical coherence tomography (SD-OCT)-based keratoconus (KC) staging system and compare it with existing KC staging systems.

SETTING

Eye Hospital of Wenzhou Medical University, Wenzhou, China.

DESIGNS

Retrospective case-control study.

METHODS

Scheimpflug tomography, air-puff tonometry, and SD-OCT were performed on 236 normal and 331 KC eyes. All SD-OCT-derived parameters of the corneal epithelium and stroma were evaluated based on their receiver operating characteristic (ROC) curves, area under the curve (AUC), sensitivity, and specificity to discriminate between normal and KC eyes. The best performing parameters were subsequently used to create an OCT-based staging system, which was compared with existing tomographic and biomechanical staging systems.

RESULTS

236 eyes from 236 normal patients and 331 eyes from 331 KC patients of different stages were included. The highest ranked AUC ROC SD-OCT parameters, derived from stroma and epithelium, were stroma overall minimum thickness (ST: AUC 0.836, sensitivity 90%, specificity 67%) and epithelium overall SD (EP: AUC 0.835, sensitivity 75%, specificity 78%). A numerical SD-OCT staging system called STEP including 2 parameters-"ST" and "EP"-with 5 stages was proposed.

CONCLUSIONS

The new SD-OCT-based KC staging system is the first to take the epithelium with its sublayer stroma information into account, showing a strong agreement to the existing staging systems. This system could be incorporated into daily practice, potentially leading to an overall improvement in KC treatment and follow-up management.

Progress of corneal morphological examination combined with biomechanical examination in preoperative screening for keratorefractive surgery

Although corneal refractive surgery has been proven to be excellent in terms of safety and effectiveness, the reduction of postoperative corneal ectasia remains one of the most concerned topics for surgeons. Forme fruste keratoconus (FFKC) is the most important factor that leads to postoperative corneal ectasia, and common preoperative screenings of the condition include corneal morphology examination and corneal biomechanical examination. However, there are limitations to the single morphological examination or biomechanical examination, and the advantages of the combination of the two have been gradually emerging. The combined examination is more accurate in the diagnosis of FFKC and can provide a basis for determining suspected keratoconus. It allows one to measure the true intraocular pressure (IOP) before and after surgery and is recommended for older patients and those with allergic conjunctivitis. This article aims to discuss the application, advantages, and disadvantages of single examination and combined examination in the preoperative screening of refractive surgery, so as to provide a certain reference value for choosing suitable patients for surgery, improving surgical safety, and reducing the risk of postoperative ectasia.

Detection of Corneal Ectasia Using OCT Maps of Pachymetry and Posterior Surface Mean Curvature

PURPOSE

To quantify the abnormal corneal thinning and posterior surface steepening that is observed in keratoconus with an Ectasia Index.

METHODS

Optical coherence tomography (OCT) was used to image the corneas of normal individuals and patients with varying stages of keratoconus (manifest, subclinical, and forme fruste). Maps of corneal pachymetry and posterior surface mean curvature were generated, and an Ectasia Index was calculated by multiplying Gaussian fits obtained from the two types of maps. Repeated five-fold cross-validation was used to evaluate the ability of the Ectasia Index to differentiate between normal and keratoconic eyes. The classification performance of the Ectasia Index was compared to minimum pachymetry and maximum posterior mean curvature.

RESULTS

Thirty-two eyes from 16 normal individuals, 89 eyes from 63 patients with manifest keratoconus, 16 eyes from 15 patients with subclinical keratoconus, and 26 eyes from 26 patients with forme fruste keratoconus were included in the study. During cross-validation, 100% of the eyes with manifest (89 of 89) and subclinical (16 of 16) keratoconus were correctly classified by the Ectasia Index. The average classification accuracy for the forme fruste keratoconus group was 63 ± 21% (16.4 of 26). The specificity for the normal group was 91 ± 10% (29.1 of 32). The Ectasia Index had a higher sensitivity for keratoconus detection and similar specificity in comparison to minimum pachymetry and maximum posterior mean curvature.

CONCLUSIONS

The Ectasia Index could be a valuable additional metric for clinicians to consider when screening for keratoconus. [J Refract Surg. 2022;38(8):502-510.].

Epithelial thickness mapping for corneal refractive surgery

PURPOSE OF REVIEW

As more devices become available that offer corneal epithelial thickness mapping, this is becoming more widely used for numerous applications in corneal refractive surgery.

RECENT FINDINGS

The epithelial thickness profile is nonuniform in the normal eye, being thinner superiorly than inferiorly and thinner temporally than nasally. Changes in the epithelial thickness profile are highly predictable, responding to compensate for changes in the stromal curvature gradient, using the eyelid as an outer template. This leads to characteristic changes that can be used for early screening in keratoconus, postoperative monitoring for early signs of corneal ectasia, and for determining whether further steepening can be performed without the risk of apical syndrome following primary hyperopic treatment. Compensatory epithelial thickness changes are also a critical part of diagnosis in irregular astigmatism as these partially mask the stromal surface irregularities. The epithelial thickness map can then be used to plan a trans-epithelial PRK treatment for cases of irregularly irregular astigmatism. Other factors can also affect the epithelial thickness profile, including dry eye, anterior basement membrane dystrophy and eyelid ptosis.

SUMMARY

Epithelial thickness mapping is becoming a crucial tool for refractive surgery, in particular for keratoconus screening, ectasia monitoring, hyperopic treatment planning, and therapeutic diagnosis and treatment.

Combining Spectral-Domain OCT and Air-Puff Tonometry Analysis to Diagnose Keratoconus

PURPOSE

To investigate the diagnostic capacity of spectral-domain optical coherence tomography (SD-OCT) combined with air-puff tonometry using artificial intelligence (AI) in differentiating between normal and keratoconic eyes.

METHODS

Patients who had either undergone uneventful laser vision correction with at least 3 years of stable follow-up or those who had forme fruste keratoconus (FFKC), early keratoconus (EKC), or advanced keratoconus (AKC) were included. SD-OCT and biomechanical information from air-puff tonometry was divided into training and validation sets. AI models based on random forest or neural networks were trained to distinguish eyes with FFKC from normal eyes. Model accuracy was independently tested in eyes with FFKC and normal eyes. Receiver operating characteristic (ROC) curves were generated to determine area under the curve (AUC), sensitivity, and specificity values.

RESULTS

A total of 223 normal eyes from 223 patients, 69 FFKC eyes from 69 patients, 72 EKC eyes from 72 patients, and 258 AKC eyes from 258 patients were included. The top AUC ROC values (normal eyes compared with AKC and EKC) were Pentacam Random Forest Index (AUC = 0.985 and 0.958), Tomographic and Biomechanical Index (AUC = 0.983 and 0.925), and Belin-Ambrósio Enhanced Ectasia Total Deviation Index (AUC = 0.981 and 0.922). When SD-OCT and air-puff tonometry data were combined, the random forest AI model provided the highest accuracy with 99% AUC for FFKC (75% sensitivity; 94.74% specificity).

CONCLUSIONS

Currently, AI parameters accurately diagnose AKC and EKC, but have a limited ability to diagnose FFKC. AI-assisted diagnostic technology that uses both SD-OCT and air-puff tonometry may overcome this limitation, leading to improved treatment of patients with keratoconus. [J Refract Surg. 2022;38(6):374-380.].

Differentiating Between Contact Lens Warpage and Keratoconus Using OCT Maps of Corneal Mean Curvature and Epithelial Thickness

PURPOSE

To formulate an Epithelial Modulation index to differentiate between eyes with contact lens warpage and keratoconus.

METHODS

Normal eyes and eyes with either contact lens warpage or keratoconus were scanned by a Fourier-domain optical coherence tomography (OCT) system. Maps of epithelial thickness and anterior surface mean curvature were generated and converted to deviation maps by subtracting the average maps from a healthy population. The Epithelial Modulation index was defined as the covariance between the two types of deviation maps. A logistic regression model was used to classify eyes as non-keratoconus (normal or warp-age) or keratoconus (manifest, subclinical, or forme fruste).

RESULTS

The average Epithelial Modulation index value for normal eyes was -0.6 ± 1.0 µm/m. Eyes with keratoconus were characterized by coincident high anterior surface mean curvature and low epithelial thickness, resulting in a high Epithelial Modulation index (manifest: 103.0 ± 82.9 µm/m, subclinical: 37.0 ± 23.0 µm/m, forme fruste: 7.3 ± 13.2 µm/m). The Epithelial Modulation index was closer to normal for eyes with warpage (-1.9 ± 4.0 µm/m). The classification accuracy of the Epithelial Modulation index during five-fold cross-validation of the logistic regression model was 100 ± 0% for normal eyes and 99.0 ± 2.0% for eyes with warpage. The accuracy was 100 ± 0%, 100 ± 0%, and 53.1 ± 1.5% for the manifest, subclinical, and forme fruste keratoconus groups, respectively.

CONCLUSIONS

The Epithelial Modulation index is useful in distinguishing eyes with secondary epithelial modulation (keratoconus) from those with primary epithelial deformation (contact lens-related warpage). [J Refract Surg. 2022;38(2):112-119.].

Combined biomechanical and tomographic keratoconus staging: Adding a biomechanical parameter to the ABCD keratoconus staging system

PURPOSE

This retrospective cross-sectional study evaluated the potential of an additional biomechanical parameter 'E' as an addition to the tomographic ABCD ectasia/keratoconus (KC) staging.

METHODS

The Corvis Biomechanical Factor (CBiF) represents the modified linear term of the Corvis Biomechanical Index (CBI) developed based on 448 KC corneas from the Homburg Keratoconus Center (HKC). The CBiF range was divided into five stages (E0 to E4) to create a grading system according to the ABCD stages. Stage E0 was characterized by values smaller than the 2.5 percentile. The thresholds were created by dividing the CBiF range between the 2.5 and 97.5 percentiles into four groups of equal values (E1-E4). The frequency distribution of 'E' was analysed and independently validated based on another 860 KC corneas dataset from Milano and Rio de Janeiro (MR). The relationship between 'E' and the ABCD staging was analysed by cross-tabulation. The specificity of 'E' was assessed based on healthy controls (112|851) from both datasets (HKC|MR).

RESULTS

'E' was normally distributed with E0 = 37|30, E1 = 86|200, E2 = 155|354, E3 = 101|206, E4 = 69|70 in the KC group and 96.4%|90.5% of the controls classified E0 in the HKC|MR dataset, respectively. Cross-tabulation revealed that 'E' was most comparable to posterior corneal curvature ('B') in both datasets, while showing a trend towards more advanced stages in comparison to anterior corneal curvature ('A') and thinnest corneal thickness ('C').

CONCLUSION

The novel Corvis-derived parameter 'E' provides a biomechanical staging for ectasia/KC potentially enhancing the ABCD staging and may detect abnormalities before tomographic changes, which requires further studies.