A Systematic Review of Subclinical Keratoconus and Forme Fruste Keratoconus

Multi-modal imaging for the detection of early keratoconus: a narrative review

Keratoconus is a common progressive corneal disorder that can be associated with significant ocular morbidity. Various corneal imaging techniques have been used for the diagnosis of established cases. However, in the early stages of the disease, which include subclinical keratoconus and forme fruste keratoconus, detection of such cases can be challenging. The importance of detecting such cases is very important because early intervention can halt disease progression, improve visual outcomes and prevent postrefractive surgery ectasia associated with performing corneal refractive procedures in such patients. This narrative review aimed to examine several established and evolving imaging techniques for the detection of early cases of keratoconus. The utilization of combinations of these techniques may further increase their diagnostic ability.

Longitudinal Analysis of Corneal Biomechanics of Suspect Keratoconus: A Prospective Case-Control Study

BACKGROUND

To evaluate the corneal biomechanics of stable keratoconus suspects (Stable-KCS) at 1-year follow-up and compare them with those of subclinical keratoconus (SKC).

METHODS

This prospective case-control study included the eyes of 144 patients. Biomechanical and tomographic parameters were recorded (Corvis ST and Pentacam). Patients without clinical signs of keratoconus in both eyes but suspicious tomography findings were included in the Stable-KCS group (n = 72). Longitudinal follow-up was used to evaluate Stable-KCS changes. Unilateral keratoconus contralateral eyes with suspicious tomography were included in the SKC group (n = 72). T-tests and non-parametric tests were used for comparison. Multivariate general linear models were used to adjust for confounding factors for further analysis. Receiver operating characteristic (ROC) curves were used to analyze the distinguishability.

RESULTS

The biomechanical and tomographic parameters of Stable-KCS showed no progression during the follow-up time (13.19 ± 2.41 months, p > 0.05). Fifteen biomechanical parameters and the Stress-Strain Index (SSI) differed between the two groups (p < 0.016). The A1 dArc length showed the strongest distinguishing ability (area under the ROC = 0.888) between Stable-KCS and SKC, with 90.28% sensitivity and 77.78% specificity at the cut-off value of -0.0175.

CONCLUSIONS

The A1 dArc length could distinguish between Stable-KCS and SKC, indicating the need to focus on changes in the A1 dArc length for keratoconus suspects during the follow-up period. Although both have abnormalities on tomography, the corneal biomechanics and SSI of Stable-KCS were stronger than those of SKC, which may explain the lack of progression of Stable-KCS.

Associations between corneal curvature and other anterior segment biometrics in young myopic adults

To investigate the associations between corneal curvature (CC) and other anterior segment biometrics in young myopic adults. In this retrospective multi-center study, 7893 young myopic adults were included. CC and other anterior segment biometrics were measured by Scheimpflug imaging (Pentacam). CC was defined as SimK at central 3 mm area, and other anterior segment biometrics included white-to-white corneal diameter (WTW), central corneal thickness (CCT), corneal volume (CV) at 3 mm, 5 mm, and 7 mm area, anterior corneal astigmatism (ACA), posterior corneal astigmatism (PCA), anterior corneal eccentricity (ACE) and asphericity (ACAP), posterior corneal eccentricity (PCE) and asphericity (PCAP), anterior chamber depth (ACD), and anterior chamber volume (ACV). Univariate regression analyses were used to assess the associations between CC and other anterior segment biometrics, and multivariate regression analyses were further performed to adjusted for age, gender and spherical equivalent. CC was higher in patients of female gender and higher myopia (all P < 0.05). Eyes in higher CC quartiles had lower WTW, thinner CCT, lower CV at 3 mm and 5 mm, lower ACD, and lower ACV (all P < 0.001), but had larger ACA, larger PCA, less PCE and less PCAP (all P < 0.001), compared to eyes in lower CC quartiles. The trends of CV at 7 mm, ACE and ACAP were inconsistent in different CC quartiles. After adjusting for age, gender and spherical equivalent with multivariate linear regression, CC was positively correlated to CV at 7 mm (βs = 0.069), ACA (βs = 0.194), PCA (βs = 0.187), ACE (βs = 0.072), PCAP (βs = 0.087), and ACD (βs = 0.027) (all P < 0.05), but was negatively correlated to WTW (βs = - 0.432), CCT (βs = - 0.087), CV-3 mm (βs = - 0.066), ACAP (βs = - 0.043), PCE (βs = - 0.062), and ACV (βs = - 0.188) (all P < 0.05). CC was associated with most of the other anterior segment biometrics in young myopic adults. These associations are important for better understanding of the interactions between different anterior segment structures in young myopic patients, and are also useful for the exploration of the pathogenesis of myopia.

Comparison of different corneal imaging modalities using artificial intelligence for diagnosis of keratoconus: a systematic review and meta-analysis

PURPOSE

This review was designed to compare different corneal imaging modalities using artificial intelligence (AI) for the diagnosis of keratoconus (KCN), subclinical KCN (SKCN), and forme fruste KCN (FFKCN).

METHODS

A comprehensive systematic search was conducted in scientific databases, including Web of Science, PubMed, Scopus, and Google Scholar based on the PRISMA statement. Two independent reviewers assessed all potential publications on AI and KCN up to March 2022. The Critical Appraisal Skills Program (CASP) 11-item checklist was used to evaluate the validity of the studies. Eligible articles were categorized into three groups (KCN, SKCN, and FFKCN) and included in the meta-analysis. The pooled estimate of accuracy (PEA) was calculated for all selected articles.

RESULTS

The initial search yielded 575 relevant publications, of which 36 met the CASP quality criteria and were included in the analysis. Qualitative assessment showed that Scheimpflug and Placido combined with biomechanical and wavefront evaluations improved KCN detection (PEA, 99.2, and 99.0, respectively). The Scheimpflug system (92.25 PEA, 95% CI, 94.76-97.51) and a combination of Scheimpflug and Placido (96.44 PEA, 95% CI, 93.13-98.19) had the highest diagnostic accuracy for the detection of SKCN and FFKCN, respectively. The meta-analysis outcomes showed no significant difference between the CASP score and accuracy of the publications (all P > 0.05).

CONCLUSIONS

Simultaneous Scheimpflug and Placido corneal imaging methods provide high diagnostic accuracy for early detection of keratoconus. The use of AI models improves the discrimination of keratoconic eyes from normal corneas.

Corneal biomechanics in early diagnosis of keratoconus using artificial intelligence

Keratoconus is a blinding eye disease that affects activities of daily living; therefore, early diagnosis is crucial. Great efforts have been made toward an early diagnosis of keratoconus. Recent studies have shown that corneal biomechanics is associated with the occurrence and progression of keratoconus. Hence, detecting changes in corneal biomechanics may provide a novel strategy for early diagnosis. However, an early keratoconus diagnosis remains challenging due to the subtle and localized nature of its lesions. Artificial intelligence has been used to help address this problem. Herein, we reviewed the literature regarding three aspects of keratoconus (keratoconus, early keratoconus, and keratoconus grading) based on corneal biomechanical properties using artificial intelligence. Furthermore, we summarized the current research progress, limitations, and possible prospects.

Corneal Biomechanical Properties to Predict Prognosis of Abnormal Tomographic Corneas: A Prospective Cohort Study

PURPOSE

To analyze the corneal biomechanical properties in patients with abnormal corneal tomography (ACT) and predict their stability using the biomechanical stability index (BSI).

DESIGN

Prospective cohort study.

METHODS

Setting: Multicenter study.

STUDY POPULATION

This study included 385 eyes of 278 patients with stable ACT (n = 70), subclinical keratoconus (SKC, n = 65), keratoconus (n = 65), normal controls (NL, n = 142). Forty-three eyes with first-visit ACT were included in a separate cohort (follow-up ACT group).

OBSERVATION PROCEDURE

Tomographical and biomechanical parameters (Pentacam and Corvis ST) were recorded.

MAIN OUTCOME MEASURES

Nonparametric tests were used for comparison. Logistic regression was employed to introduce BSI to separate stable ACT and SKC accurately. An independent dataset of 43 first-visit ACT eyes was followed up for 1 year to validate BSI's accuracy and positive and negative predictive values (PPV, NPV).

RESULTS

The tomographical and biomechanical parameters in patients with Stable ACT remained stable over the follow-up period (12.73 ± 2.57 months, P > .05). Stable ACT had 12/14 biomechanical parameters different (P < .05) from SKC but not different from NL (P > .05). With a cut-off value of 0.585, BSI demonstrated the strongest ability to distinguish between stable ACT and SKC (area under the receiver operating characteristic curve = 0.991), with 93.85% sensitivity and 97.14% specificity. During the 1-year follow-up of 43 eyes (follow-up ACT group), 30 remained stable. The accuracy, PPV, and NPV of the BSI were 95.35%, 100%, and 93.75%, respectively.

CONCLUSIONS

Biomechanical properties of patients with stable abnormal tomography corneas were stronger than SKC and close to normal corneas, which may explain the reason for tomographic stability. The BSI may be useful for predicting disease progression in patients with ACT and the possible management of corneal cross-linking at the first visit.

Subclinical Keratoconus Detection and Characterization Using Motion-Tracking Brillouin Microscopy

PURPOSE

To characterize focal biomechanical alterations in subclinical keratoconus (SKC) using motion-tracking (MT) Brillouin microscopy and evaluate the ability of MT Brillouin metrics to differentiate eyes with SKC from normal control eyes.

DESIGN

Prospective cross-sectional study.

PARTICIPANTS

Thirty eyes from 30 patients were evaluated, including 15 eyes from 15 bilaterally normal patients and 15 eyes with SKC from 15 patients.

METHODS

All patients underwent Scheimpflug tomography and MT Brillouin microscopy using a custom-built device. Mean and minimum MT Brillouin values within the anterior plateau region and anterior 150 μm were generated. Scheimpflug metrics evaluated included inferior-superior (IS) value, maximum keratometry (Kmax), thinnest corneal thickness, asymmetry indices, Belin/Ambrosio display total deviation, and Ambrosio relational thickness. Receiver operating characteristic (ROC) curves were generated for all Scheimpflug and MT Brillouin metrics evaluated to determine the area under the ROC curve (AUC), sensitivity, and specificity for each variable.

MAIN OUTCOME MEASURES

Discriminative performance based on AUC, sensitivity, and specificity.

RESULTS

No significant differences were found between groups for age, sex, manifest refraction spherical equivalent, corrected distance visual acuity, Kmax, or KISA% index. Among Scheimpflug metrics, significant differences were found between groups for thinnest corneal thickness (556 μm vs. 522 μm; P < 0.001), IS value (0.29 diopter [D] vs. 1.05 D; P < 0.001), index of vertical asymmetry (IVA; 0.10 vs. 0.19; P < 0.001), and keratoconus index (1.01 vs. 1.05; P < 0.001), and no significant differences were found for any other Scheimpflug metric. Among MT Brillouin metrics, clear differences were found between control eyes and eyes with SKC for mean plateau (5.71 GHz vs. 5.68 GHz; P < 0.0001), minimum plateau (5.69 GHz vs. 5.65 GHz; P < 0.0001), mean anterior 150 μm (5.72 GHz vs. 5.68 GHz; P < 0.0001), and minimum anterior 150 μm (5.70 GHz vs. 5.66 GHz; P < 0.001). All MT Brillouin plateau and anterior 150 μm mean and minimum metrics fully differentiated groups (AUC, 1.0 for each), whereas the best performing Scheimpflug metrics were keratoconus index (AUC, 0.91), IS value (AUC, 0.89), and IVA (AUC, 0.88).

CONCLUSIONS

Motion-tracking Brillouin microscopy metrics effectively characterize focal corneal biomechanical alterations in eyes with SKC and clearly differentiated these eyes from control eyes, including eyes that were not differentiated accurately using Scheimpflug metrics.

FINANCIAL DISCLOSURE(S)

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Enhanced morphological assessment based on interocular asymmetry analysis for keratoconus detection

PURPOSE

To clarify the interocular asymmetry of corneal morphological descriptors and evaluate its discriminant ability of keratoconus (KC).

METHODS

This retrospective study recruited 344 normal participants and 290 KC patients, randomized to training and validation datasets. Interocular correlation and agreement were evaluated on 44 corneal morphological descriptors derived from Schiempflug tomography. Logistic regression models were constructed using binocular data and of which diagnostic performance was evaluated using the area under receiver operating characteristics curve (AUC), net reclassification improvement (NRI), and integrated discrimination improvement (IDI).

RESULTS

Interocular agreement of corneal descriptors is better in the normal than in KC except for dimensions of cornea and anterior chamber. The interocular asymmetry increases along with the severity of KC. Interocular asymmetry in maximum anterior keratometry, mean anterior keratometry and higher-order aberrations of anterior surface show high AUC above 0.950. Binocular logistic regression index reaches an AUC of 0.963 with high specificity (95.2%) and brings gain to monocular parameters in distinguishing the normal eyes from KC (NRI = 0.080 (0.042 ~ 0.118), P < 0.001) and IDI = 0.071 (0.049 ~ 0.092), P < 0.001). Interocular asymmetry benefits even more in subclinical keratoconus (SKC) detection reflected by NRI (0.4784 (0.2703-0.6865), P < 0.001) and IDI (0.2680 (0.1495-0.3866), P < 0.001) measures.

CONCLUSION

Interocular asymmetry is a well-characterized feature of KC and related to the severity. It is feasible to apply the interocular asymmetry in diagnosis of KC and SKC as a replenishment of monocular parameters and in progression tracking.

Altered Corneal Biomechanics According to the Biomechanical E-Staging in Pellucid Marginal Degeneration

PURPOSE

The purpose of this study was to investigate corneal biomechanics in pellucid marginal degeneration (PMD) compared with healthy controls using Corvis ST (Oculus, Germany) by using the new biomechanical E-staging (based on the Corvis Biomechanical Factor, the linearized Corvis Biomechanical Index) together with tomographic parameters.

METHODS

Corneal biomechanical and topographic data of 75 eyes of 75 patients with PMD and 75 eyes of 75 age-matched and sex-matched healthy controls were investigated. Topographic parameters (K1, K2, Kmax, central corneal thickness (CCT), and Belin/Ambrósio Deviation Index (BAD-D) were evaluated in dependence of and correlated with the biomechanically defined E-stages. Biomechanical parameters were also recorded for the 2 groups.

RESULTS

Patients with PMD showed higher K2, Kmax, BAD-D, and Corvis Biomechanical Factor values and a lower CCT compared with healthy controls (P < 0.001). The E-stage was positively correlated with K1, K2, Kmax, BAD-D, and intraocular pressure difference and negatively correlated with CCT. Stage-dependent analysis revealed a significant increase in K1, K2, Kmax (P < 0.001), and BAD-D (P = 0.041) in stage E3 compared with E0 and a significant decrease in stage E2 in CCT (P = 0.009) compared with E0.

CONCLUSIONS

This study showed that patients with PMD may have a reduced corneal stiffness compared with healthy controls which worsens with increasing E-stage. Significant changes in topographic parameters were observed at stage E2 for CCT and at stage E3 for K1, K2, Kmax, and BAD-D when compared with stage E0.

Performance of Corvis ST Parameters Including Updated Stress-Strain Index in Differentiating Between Normal, Forme-Fruste, Subclinical, and Clinical Keratoconic Eyes

PURPOSE

This study seeks to evaluate the ability of the updated stress strain index (SSIv2) and other Corvis ST biomechanical parameters in distinguishing between keratoconus at different disease stages and normal eyes.

DESIGN

Diagnostic accuracy analysis to distinguish disease stages.

METHODS

1084 eyes were included and divided into groups of normal (199 eyes), forme fruste keratoconus (FFKC, 194 eyes), subclinical keratoconus (SKC, 113 eyes), mild clinical keratoconus (CKC-Ⅰ, 175 eyes), moderate clinical keratoconus (CKC-Ⅱ, 204 eyes), and severe clinical keratoconus (CKC-Ⅲ, 199 eyes). Each eye was subjected to a Corvis ST examination to determine the central corneal thickness (CCT), biomechanically corrected intraocular pressure (bIOP), SSIv2 (updated stress-strain index), and other 8 Corvis parameters including the stress-strain index (SSIv1), stiffness parameter at first applanation (SP-A1), first applanation time (A1T), Ambrósio relational thickness to the horizontal profile (ARTh), integrated inverse radius (IIR), maximum deformation amplitude (DAM), ratio between deformation amplitude at the apex and at 2 mm nasal and temporal (DARatio2), and Corvis biomechanical index (CBI). The sensitivity and specificity of these parameters in diagnosing keratoconus were analyzed through receiver operating characteristic curves.

RESULTS

Before and after correction for CCT and bIOP, SSIv2 and ARTh were significantly higher and IIR and CBI were significantly lower in the normal group than in the FFKC group, SKC group and the 3 CKC groups (all P < .05). There were also significant correlations between the values of SSIv2, ARTh, IIR, CBI, and the CKC severity (all P < .05). AUC of SSIv2 was significantly higher than all other Corvis parameters in distinguishing normal eyes from FFKC, followed by IIR, ARTh and CBI.

CONCLUSION

Corvis ST's updated stress-strain index, SSIv2, demonstrated superior performance in differentiating between normal and keratoconic corneas, and between corneas with different keratoconus stages. Similar, but less pronounced, performance was demonstrated by the IIR, ARTh and CBI.

Assessing progression limits in different grades of keratoconus from a novel perspective: precision of measurements of the corneal epithelium

BACKGROUND

To assess repeatability and reproducibility of corneal epithelium thickness (ET) measured by a spectral-domain optical coherence tomographer (SD-OCT)/Placido topographer (MS-39, CSO, Florence, Italy) in keratoconus (KC) population at different stages, as well as to determine the progression limits for evaluating KC progression.

METHODS

A total of 149 eyes were enrolled in this study, with 29 eyes in the forme fruste keratoconus (FFKC) group, 34 eyes in the mild KC group, 40 eyes in the moderate KC group, and 46 eyes in the severe KC group. Employing the within-subject standard deviation (Sw), test-retest variability (TRT), coefficient of variation (CoV), and intraclass correlation coefficient (ICC) to evaluate intraoperator repeatability and interoperator reproducibility.

RESULTS

The repeatability and reproducibility of MS-39 in patients with KC were acceptable, according to ICC values ranging from 0.732 to 0.954. However, patients with more severe KC and progressive peripheralization of the measurement points had higher TRTs but a thinning trend. The current study tended to set the cut-off values of mild KC, moderate KC, and severe KC to 4.9 µm, 5.2 µm, and 7.4 µm for thinnest epithelium thickness (TET). When differences between follow-ups are higher than those values, progression of the disease is possible. As for center epithelium thickness (CET), cut-off values for mild KC, moderate KC, and severe KC should be 2.8 µm, 4.4 µm, and 5.3 µm. This might be useful in the follow-up and diagnosis of keratoconus.

CONCLUSIONS

This study demonstrated that the precision of MS-39 was reduced in measuring more severe KC patients and more peripheral corneal points. In determining disease progression, values should be differentiated between disease-related real changes and measurement inaccuracies. Due to the large difference in ET measured by MS-39 between various stages of disease progression, it is necessary to accurately grade KC patients to avoid errors in KC clinical decision-making.

A Comprehensive Wavefront Assessment of Keratoconus Using an Integrated Scheimpflug Corneal Tomographer/Hartmann-Shack Wavefront Aberrometer

OBJECTIVES

To characterize higher-order aberrations (HOAs) in different severities of keratoconus (KC) from the anterior and posterior corneal surfaces and whole eye using an integrated Scheimpflug corneal tomographer/Hartmann-Shack wavefront aberrometer.

METHODS

This study included eyes with clinical KC, topographic KC (no clinical signs), fellow eyes with very asymmetric ectasia with normal topography and no clinical signs (VAE-NT), and control eyes. Corneal and ocular wavefront aberrations were obtained using an integrated Scheimpflug tomographer/Hartmann-Shack wavefront aberrometer. The diagnostic capability of distinguishing VAE-NT from the control was also tested.

RESULTS

This study included 68 eyes with clinical KC, 44 with topographic KC, 26 with VAE-NT, and 45 controls. Clinical KC had significantly greater total HOAs and coma from the anterior and posterior corneal surfaces and whole eye than the other groups ( P <0.05). Although topographic KC had significantly greater values in all wavefront parameters than the control ( P <0.05), ocular and corneal HOAs did not differ between the VAE-NT and control groups. The coma from the anterior cornea in topographic KC was significantly greater than that in VAE-NT ( P <0.05); the coma from the posterior cornea and whole eye did not differ. Total HOAs from the anterior corneal surface exhibited the highest area under the receiver operating characteristic curve value of 0.774 (sensitivity, 73%; specificity, 78%).

CONCLUSION

A comprehensive wavefront assessment can be used to quantitatively evaluate corneal and ocular HOAs across various severity of KC. Total HOAs from the anterior corneal surface exhibited the potential ability in distinguishing VAE-NT from the control eyes.

Longitudinal Evaluation of Biomechanical Indices in Fellow Eyes of Patients With Keratoconus Classified as Having Very Asymmetric Ectasia With Normal Topography

PURPOSE

To evaluate the biomechanical longitudinal variability and progression of tomographically normal fellow eyes of patients with keratoconus.

METHODS

Of 513 patients with keratoconus, 30 patients with tomographically normal fellow eyes were included in this study. Tomographic and biomechanical parameters of the Pentacam and Corvis ST (Oculus Optikgeräte GmbH) were analyzed in multiple follow-up examinations, including the ABCD grading, Belin/Ambrósio Enhanced Ectasia total deviation index (BAD-D), Corvis Biomechanical Index (CBI), Corvis Biomechanical Factor (CBiF), and Tomographic and Biomechanical Index (TBI). A mixed regression model was applied. The results were compared to a healthy control group (n = 17) and a keratoconus group (n = 20).

RESULTS

Within a maximum observation period of 3.3 years, no fellow eye (0%) showed a progression to tomographically evident keratoconus. No significant change in tomographic or biomechanical parameters was detected over the study period. The indices BAD-D, CBI, CbiF, and TBI exhibited a certain variability over time, whereas the tomographic ABC parameters and maximum keratometry barely changed. This was also shown in the control group and for all parameters in the keratoconus group, except the TBI.

CONCLUSIONS

During the observation period none of the normal fellow eyes progressed to tomographically detectable keratoconus. However, biomechanical parameters CBI, CbiF, and TBI showed pathological values in 43.3% of eyes and certain variability. Subsequent studies with a longer observation period are warranted to confirm the biomechanical trends seen in this study and to rate the ability of single measurements to diagnose early keratoconus. [J Refract Surg. 2024;40(1):e48-e56.].

Parental Corneal Tomographic and Biomechanical Characteristics of Patients With Keratoconus

PURPOSE

To investigate the hereditability of corneal tomographic and biomechanical parameters in keratoconus (KC).

DESIGN

Prospective cohort study.

METHODS

This study was conducted at Qingdao Eye Hospital of Shandong First Medical University in Qingdao, China. Forty-four patients with KC and their biological parents (n = 88) were recruited as the study group. The control group consisted of 84 healthy adults with matched age and gender. Both eyes of each participant underwent clinical examinations, and 1 eye was selected for statistical analysis. Exclusion criteria were as follows: individuals with glaucoma, ocular surgery, systemic diseases known to affect the eyes, or poor cooperation during examination. Subjects were asked to discontinue soft contact lens (CL) wear for 2 weeks and rigid gas permeable CL wear for 4 weeks before ocular examination. All participants underwent a comprehensive assessment including Pentacam Scheimpflug tomography, Corvis ST, visual acuity, refraction examination, axial length, and slitlamp examination for both eyes. Individuals presenting with KC manifestations in at least 1 eye were classified as having KC. A total of 9 Pentacam indices including keratometry in the flat/steep meridian (K1/K2), maximal keratometry (Kmax), thinnest point pachymetry (TP), and maximum/average Ambrósio relational thickness (ARTmax/ARTave), anterior and posterior surfaces elevation of the cornea (Ef/Eb) and total deviation value (Final D), and 21 biomechanical indices were collected. Associations of these factors with KC were evaluated using multiple comparison and binary logistics regression analyses.

RESULTS

Two parents (2.27%) from 2 different families were diagnosed with KC. Parents of patients with KC had thinner corneas with altered corneal biomechanical parameters compared with healthy controls (P < .05). The combined tomographic and biomechanical index demonstrated the highest discriminatory power (area under the receiver operating characteristic curve 0.785) and strong specificity (84.5%). Parental corneal tomographic and biomechanical index, Corvis biomechanical index, and TP were identified as the major influential factors for KC in their offspring by logistic regression analysis, with a 73.3% accuracy in identifying offspring with KC.

CONCLUSIONS

Parental corneal tomographic and biomechanical properties of patients with KC suggest a possible predisposition to KC. A combination of tomography and corneal biomechanics can be helpful in predicting the incidence rate of KC in the offspring of patients with subclinical KC.

Corneal biomechanics in normal and subclinical keratoconus eyes

BACKGROUND

The diagnosis of keratoconus, as the most prevalent corneal ectatic disorder, at the subclinical stage gained great attention due to the increased acceptance of refractive surgeries. This study aimed to assess the pattern of the corneal biomechanical properties derived from Corneal Visualization Scheimpflug Technology (Corvis ST) and evaluate the diagnostic value of these parameters in distinguishing subclinical keratoconus (SKC) from normal eyes.

METHODS

This prospective study was conducted on 73 SKC and 69 normal eyes. Subclinical keratoconus eyes were defined as corneas with no clinical evidence of keratoconus and suspicious topographic and tomographic features. Following a complete ophthalmic examination, topographic and tomographic corneal assessment via Pentacam HR, and corneal biomechanical evaluation utilizing Corvis ST were done.

RESULTS

Subclinical keratoconus eyes presented significantly higher Deformation Amplitude (DA) ratio, Tomographic Biomechanical Index (TBI), and Corvis Biomechanical Index (CBI) rates than the control group. Conversely, Ambrósio Relational Thickness to the Horizontal profile (ARTh), and Stiffness Parameter at the first Applanation (SPA1) showed significantly lower rates in SKC eyes. In diagnosing SKC from normal eyes, TBI (AUC: 0.858, Cut-off value: > 0.33, Youden index: 0.55), ARTh (AUC: 0.813, Cut-off value: ≤ 488.1, Youden index: 0.58), and CBI (AUC: 0.804, Cut-off value: > 0.47, Youden index: 0.49) appeared as good indicators.

CONCLUSIONS

TBI, CBI, and ARTh parameters could be valuable in distinguishing SKC eyes from normal ones.

Multimodal diagnostics for keratoconus and ectatic corneal diseases: a paradigm shift

Different diagnostic approaches for ectatic corneal diseases (ECD) include screening, diagnosis confirmation, classification of the ECD type, severity staging, prognostic evaluation, and clinical follow-up. The comprehensive assessment must start with a directed clinical history. However, multimodal imaging tools, including Placido-disk topography, Scheimpflug three-dimensional (3D) tomography, corneal biomechanical evaluations, and layered (or segmental) tomography with epithelial thickness by optical coherence tomography (OCT), or digital very high-frequency ultrasound (dVHF-US) serve as fundamental complementary exams for measuring different characteristics of the cornea. Also, ocular wavefront analysis, axial length measurements, corneal specular or confocal microscopy, and genetic or molecular biology tests are relevant for clinical decisions. Artificial intelligence enhances interpretation and enables combining such a plethora of data, boosting accuracy and facilitating clinical decisions. The applications of diagnostic information for individualized treatments became relevant concerning the therapeutic refractive procedures that emerged as alternatives to keratoplasty. The first paradigm shift concerns the surgical management of patients with ECD with different techniques, such as crosslinking and intrastromal corneal ring segments. A second paradigm shift involved the quest for identifying patients at higher risk of progressive iatrogenic ectasia after elective refractive corrections on the cornea. Beyond augmenting the sensitivity to detect very mild (subclinical or fruste) forms of ECD, ectasia risk assessment evolved to characterize the inherent susceptibility for ectasia development and progression. Furthermore, ectasia risk is also related to environmental factors, including eye rubbing and the relational impact of the surgical procedure on the cornea.

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.

Localized Corneal Biomechanical Alteration Detected In Early Keratoconus Based on Corneal Deformation Using Artificial Intelligence

PURPOSE

This study aimed to develop a novel method to diagnose early keratoconus by detecting localized corneal biomechanical changes based on dynamic deformation videos using machine learning.

DESIGN

Diagnostic research study.

METHODS

We included 917 corneal videos from the Tianjin Eye Hospital (Tianjin, China) and Shanxi Eye Hospital (Xi'an, China) from February 6, 2015, to August 25, 2022. Scheimpflug technology was used to obtain dynamic deformation videos under forced puffs of air. Fourteen new pixel-level biomechanical parameters were calculated based on a spline curve equation fitting by 115,200-pixel points from the corneal contour extracted from videos to characterize localized biomechanics. An ensemble learning model was developed, external validation was performed, and the diagnostic performance was compared with that of existing clinical diagnostic indices. The performance of the developed machine learning model was evaluated using precision, recall, F1 score, and the area under the receiver operating characteristic curve.

RESULTS

The ensemble learning model successfully diagnosed early keratoconus (area under the curve = 0.9997) with 95.73% precision, 95.61% recall, and 95.50% F1 score in the sample set (n=802). External validation on an independent dataset (n=115) achieved 91.38% precision, 92.11% recall, and 91.18% F1 score. Diagnostic accuracy was significantly better than that of existing clinical diagnostic indices (from 86.28% to 93.36%, all P <0.01).

CONCLUSIONS

Localized corneal biomechanical changes detected using dynamic deformation videos combined with machine learning algorithms were useful for diagnosing early keratoconus. Focusing on localized biomechanical changes may guide ophthalmologists, aiding the timely diagnosis of early keratoconus and benefiting the patient's vision.

The application of corneal biomechanical interocular asymmetry for the diagnosis of keratoconus and subclinical keratoconus

Purpose: To evaluate the interocular consistency of biomechanical properties in normal, keratoconus (KC) and subclinical keratoconus (SKC) populations and explore the application of interocular asymmetry values in KC and SKC diagnoses. Methods: This was a retrospective chart-review study of 331 ametropic subjects (control group) and 207 KC patients (KC group, including 94 SKC patients). Interocular consistency was evaluated using the intraclass correlation coefficient (ICC). Interocular asymmetry was compared between the control and KC groups and its correlation with disease severity was analyzed. Three logistic models were constructed using biomechanical monocular parameters and interocular asymmetry values. The diagnostic ability of interocular asymmetry values and the newly established models were evaluated using receiver operating characteristic curves and calibration curves. Net reclassification improvement (NRI) and integrated discrimination improvement (IDI) were also estimated. Results: The interocular consistency significantly decreased and the interocular asymmetry values increased in KC patients compared with those in control individuals. In addition, the interocular asymmetry values increased with respect to the severity of KC. The binocular assisted biomechanical index (BaBI) had an area under the curve (AUC) of 0.998 (97.8% sensitivity, 99.2% specificity; cutoff 0.401), which was statistically higher than that of the Corvis biomechanical index [CBI; AUC = 0.935, p < 0.001 (DeLong's test), 85.6% sensitivity]. The optimized cutoff of 0.163 provided an AUC of 0.996 for SKC with 97.8% sensitivity, which was higher than that of CBI [AUC = 0.925, p < 0.001 (DeLong's test), 82.8% sensitivity]. Conclusion: Biomechanical interocular asymmetry values can reduce the false-negative rate and improve the performance in KC and SKC diagnoses.

Keratoconus Diagnosis: From Fundamentals to Artificial Intelligence: A Systematic Narrative Review

The remarkable recent advances in managing keratoconus, the most common corneal ectasia, encouraged researchers to conduct further studies on the disease. Despite the abundance of information about keratoconus, debates persist regarding the detection of mild cases. Early detection plays a crucial role in facilitating less invasive treatments. This review encompasses corneal data ranging from the basic sciences to the application of artificial intelligence in keratoconus patients. Diagnostic systems utilize automated decision trees, support vector machines, and various types of neural networks, incorporating input from various corneal imaging equipment. Although the integration of artificial intelligence techniques into corneal imaging devices may take time, their popularity in clinical practice is increasing. Most of the studies reviewed herein demonstrate a high discriminatory power between normal and keratoconus cases, with a relatively lower discriminatory power for subclinical keratoconus.

Screening of sensitive in vivo characteristics for early keratoconus diagnosis: a multicenter study

Purpose: To analyze and compare sensitive in vivo characteristics for screening early keratoconus. Methods: This multicenter, case-control study included 712 eyes, after matching for age and biomechanically corrected intraocular pressure, from three clinics in different cities. The keratoconus (n = 288), early keratoconus (n = 91), and normal cornea (n = 333) groups included eyes diagnosed with bilateral keratoconus, fellow eyes with relatively normal topography with unilateral keratoconus, and normal eyes before refractive surgery, respectively. After adjusting for central corneal thickness, differences in vivo characteristics were analyzed among the three groups. The in vivo characteristics were measured by Pentacam and Corvis ST. Fifty-four indices were evaluated to screen for a sensitive index for the detection of early keratoconus. Results: Significant differences were observed in 26 of the 36 corneal biomechanical indeces between the early keratoconus and normal corneas. The area under the receiver operating characteristic curve of tomographic and biomechanical index, Belin/Ambrósio deviation, and Da in differentiating keratoconus from normal cornea was 1.000. Among the top five indeces of the area under the receiver operating characteristic curve for detecting early keratoconus, the corneal biomechanical-related index accounted for 80% (4/5), including A1 dArc length, highest concavity radius, A2 time, and tomographic and biomechanical index, of which the area under the receiver operating characteristic curve of A1 dArc length was 0.901. Conclusion: A1 dArc length and several corneal biomechanical indices are highly sensitive for the detection of early keratoconus, even in the absence of topographic abnormalities. Ophthalmologists should focus on the clinical application of corneal biomechanics and combine corneal tomography for the timely and accurate detection of early keratoconus.

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.

In vivo corneal elastography: A topical review of challenges and opportunities

Clinical measurement of corneal biomechanics can aid in the early diagnosis, progression tracking, and treatment evaluation of ocular diseases. Over the past two decades, interdisciplinary collaborations between investigators in optical engineering, analytical biomechanical modeling, and clinical research has expanded our knowledge of corneal biomechanics. These advances have led to innovations in testing methods (ex vivo, and recently, in vivo) across multiple spatial and strain scales. However, in vivo measurement of corneal biomechanics remains a long-standing challenge and is currently an active area of research. Here, we review the existing and emerging approaches for in vivo corneal biomechanics evaluation, which include corneal applanation methods, such as ocular response analyzer (ORA) and corneal visualization Scheimpflug technology (Corvis ST), Brillouin microscopy, and elastography methods, and the emerging field of optical coherence elastography (OCE). We describe the fundamental concepts, analytical methods, and current clinical status for each of these methods. Finally, we discuss open questions for the current state of in vivo biomechanics assessment techniques and requirements for wider use that will further broaden our understanding of corneal biomechanics for the detection and management of ocular diseases, and improve the safety and efficacy of future clinical practice.

Combinations of Scheimpflug tomography, ocular coherence tomography and air-puff tonometry improve the detection of keratoconus

PURPOSE

To determine whether combinations of devices with different measuring principles, supported by artificial intelligence (AI), can improve the diagnosis of keratoconus (KC).

METHODS

Scheimpflug tomography, spectral-domain optical coherence tomography (SD-OCT), and air-puff tonometry were performed in all eyes. The most relevant machine-derived parameters to diagnose KC were determined using feature selection. The normal and forme fruste KC (FFKC) eyes were divided into training and validation datasets. The selected features from a single device or different combinations of devices were used to develop models based on random forest (RF) or neural networks (NN) trained to distinguish FFKC from normal eyes. The accuracy was determined using receiver operating characteristic (ROC) curves, area under the curve (AUC), sensitivity, and specificity.

RESULTS

271 normal eyes, 84 FFKC eyes, 85 early KC eyes, and 159 advanced KC eyes were included. A total of 14 models were built. Air-puff tonometry had the highest AUC for detecting FFKC using a single device (AUC = 0.801). Among all two-device combinations, the highest AUC was accomplished using RF applied to selected features from SD-OCT and air-puff tonometry (AUC = 0.902), followed by the three-device combination with RF (AUC = 0.871) with the best accuracy.

CONCLUSION

Existing parameters can precisely diagnose early and advanced KC, but their diagnostic ability for FFKC could be optimized. Applying an AI algorithm to a combination of air-puff tonometry with Scheimpflug tomography or SD-OCT could improve FFKC diagnostic ability. The improvement in diagnostic ability by combining three devices is modest.

Biomechanics of keratoconus: Two numerical studies

BACKGROUND

The steep cornea in keratoconus can greatly impair eyesight. The etiology of keratoconus remains unclear but early injury that weakens the corneal stromal architecture has been implicated. To explore keratoconus mechanics, we conducted two numerical simulation studies.

METHODS

A finite-element model describing the five corneal layers and the heterogeneous mechanical behaviors of the ground substance and lamellar collagen-fiber architecture in the anterior and posterior stroma was developed using the Holzapfel-Gasser-Ogden constitutive model. The geometry was from a healthy subject. Its stroma was divided into anterior, middle, and posterior layers to assess the effect of changing regional mechanical parameters on corneal displacement and maximum principal stress under intraocular pressure. Specifically, the effect of softening an inferocentral corneal button, the collagen-based tissues throughout the whole cornea, or specific stromal layers in the button was examined. The effect of simply disorganizing the orthogonally-oriented posterior stromal fibers in the button was also assessed. The healthy cornea was also subjected to eye rubbing-like loading to identify the corneal layer(s) that experienced the most tensional stress.

RESULTS

Conical deformation and corneal thinning emerged when the corneal button or the mid-posterior stroma of the button underwent gradual softening or when the collagen fibers in the mid-posterior stroma of the button were dispersed. Softening the anterior layers of the button or the whole cornea did not evoke conical deformation. Button softening greatly increased and disrupted the stress on Bowman's membrane while mid-posterior stromal softening increased stress in the anterior layers. Eye rubbing profoundly stressed the deep posterior stroma while other layers were negligibly affected.

DISCUSSION

These observations suggest that keratoconus could be initiated, at least partly, by mechanical instability/damage in the mid-posterior stroma that then imposes stress on the anterior layers. This may explain why subclinical keratoconus is marked by posterior but not anterior elevation on videokeratoscopy.

Revisiting the oil droplet sign in keratoconus: Utility for early keratoconus diagnosis and screening

PURPOSE

An annular dark shadow (ADS) reflex has been observed while performing direct ophthalmoscopy on subjects with keratoconus. This study describes a method that may serve as a diagnostic technique for early keratoconus and may be used as a quantitative measure of severity.

METHODS

Healthy keratoconic subjects and keratoconus suspects underwent corneal tomography and a full ocular examination. Keratoconus severity was graded based on Belin ABCD criteria. An iPhone camera was connected to a direct ophthalmoscope to take a picture of the eye. The height of the ASD was measured using the AutoCAD software. Differences between subject groups were evaluated by chi-squared and Mann-Whitney tests. Spearman correlation compared ocular parameters and the height of the ADS. A multiple stepwise linear regression was used to predict the height of the ADS based on clinical parameters.

RESULTS

Fifty-eight subjects participated in this study: 37 healthy controls (37 eyes) and 21 keratoconics or keratoconus suspects (37 eyes). The ADS was present in all keratoconic and keratoconus-suspect eyes but in none of the controls. The height of the ADS was significantly correlated with keratoconus severity. Front corneal surface root mean square of higher order aberrations, sphere and anterior radius of curvature from the front apex curve are significant predictors of the height of the ADS.

CONCLUSIONS AND RELEVANCE

The ADS may be a useful method to diagnose keratoconus and keratoconus-suspect cases and serve as a grading and follow-up method for tracking disease severity.

Evaluation of the optic nerve head and macular vessel density in keratoconus patients using optical coherence tomography angiography- a cross-sectional study

PURPOSE

Our study evaluated macular and optic disc vascular density (VD) changes in patients diagnosed with keratoconus, as compared with gender and age-matched healthy controls, in relation to the keratoconus stage.

METHODS

This comparative cross-sectional study included adult patients who received a diagnosis of keratoconus at the Cornea Clinic. All scans consisted of two 4.5 mm × 4.5 mm images of the optic disc area and 6 × 6 mm images of the macular area using optical coherence tomography angiography.

RESULTS

The study included 67 patients (26 women (38.8%) and 41 men (61.2%)) and 74 controls (31 women (41.9%) and 43 men (58.1%)). The mean ages of the keratoconus group and the control group were 30.0 ± 8.0 and 31.1 ± 7.5, respectively. Macular VD (whole, superficial, and deep) and peripapillary VD (whole) values were decreased statistically significant compared with the control group (p < 0.01 and p = 0.015, respectively). The keratoconus stage was negatively correlated with both the macular VD (p = 0.048, r = -0.314) and whole-peripapillary VD values (p = 0.03, r = -0.34).

CONCLUSION

Keratoconus might affect the posterior and anterior segments due to similar etiologies, including diminished collagen quality. Therefore, examination of the posterior segment should be performed thoroughly in keratoconus patients.

Biomechanical properties analysis of forme fruste keratoconus and subclinical keratoconus

PURPOSE

To analyze the biomechanical properties of the eye in patients with unilateral keratoconus with normal (forme fruste keratoconus [FFKC]) or abnormal topography (subclinical keratoconus [SKC]).

METHODS

This study included 153 eyes of 153 participants, including 95 eyes of patients with unilateral keratoconus, and 58 eyes of 58 healthy controls. Contralateral eyes with unilateral keratoconus were divided into two groups according to clinical manifestations and global consensus: FFKC (n = 30) and SKC (n = 65). The biomechanical characteristics were analyzed using non-parametric tests; further analysis thereof was performed after adjusting for confounding factors (i.e., intraocular pressure, age, and corneal thickness). Receiver operating characteristic curve (ROC) was used to analyze the ability of the biomechanical parameters to distinguish FFKC from SKC.

RESULTS

Statistically significant differences between the FFKC and SKC groups were found in 9 of the 18 corneal biomechanical parameters analyzed using non-parametric tests. After adjusting for confounding factors, the multivariate analysis still revealed significant statistical differences in A1-time (P = 0.017), integrated radius (IR) (P = 0.024), and tomographic and biomechanical index (TBI, P < 0.001) between the FFKC and SKC groups. Stiffness parameter at first applanation (SP-A1) (Area under ROC [AUROC] = 0.765) demonstrated the strongest distinguishing ability, except for TBI (AUROC = 0.858) and Corvis Biomechanical Index (AUROC = 0.849), however, there was no statistically significant difference in SP-A1 (P = 0.366) between FFKC and SKC.

CONCLUSIONS

Biomechanical parameters A1-time and IR have a high diversity between FFKC and SKC, besides TBI, and may reflect more subtle changes in corneal biomechanical properties (BPs) preceding SP-A1. The BPs of SKC are weaker than FFKC, which might be a basic and clue for the classification and diagnosis of the severity of early keratoconus in terms of biomechanics.

Analysis of the diagnostic accuracy of Belin/Ambrósio Enhanced Ectasia and Corvis ST parameters for subclinical keratoconus

OBJECTIVE

To investigate the diagnostic accuracy of the parameters in the Belin/Ambrósio Enhanced Ectasia Display built in Pentacam, which is designed for the screening of subclinical keratoconus (SKC) built in Pentacam, and the parameters in Corneal visualization Scheimpflug technology (Corvis ST).

METHODS

A retrospective study: The fellow eyes of unilateral keratoconus cases were diagnosed with SKC. Patients presented to Shanxi Eye Hospital with SKC from October 2020 to November 2021 were included as the SKC group, and myopic patients undergoing corneal refractive surgery at the Refractive Surgery Department in our hospital within the same period were included as the control group. The Belin/Ambrósio and Corvis ST parameters were extracted from the system and analyzed using independent samples t test. Receiver operating curves (ROCs) were also created to test the diagnostic accuracy of each parameter.

RESULTS

There were 70 patients (70 eyes) in the SKC group and 137 patients (137 eyes) in the control group. For Corvis ST parameters, Radius (P = 0.021), PachySlope (P = 0.040), SP-A1 (P = 0.002), A2 Deformation Amp. (P = 0.028), A2 Deflection Length (P < 0.001), Max ICR (P = 0.005), DA Ratio Max (1 mm) (P = 0.023), IR (P = 0.016), CBI (P = 0.003) and TBI (P < 0.001) were statistically different between the two groups. For Belin/Ambrósio parameters, PPI min. Axis, ART min, ART max, ART avg, Pachy min, Front K2, Astig, BAD-Df, BAD-Db, BAD-Dp, BAD-Dt, BAD-Da, BAD-D, PPI min, PPI max, PPI max. Axis, PPI avg and Dist.Apex-Thin.Loc. were significantly different between the two groups (all p < 0.001). TBI and BAD-D showed the best diagnostic accuracy, with AUCs of 0.944 and 0.965, respectively.

CONCLUSIONS

Some Belin/Ambrósio and Corvis ST parameters differed between SKC eyes and eyes with normal cornea. TBI and BAD-D showed the ideal diagnostic performance for SKC. In clinical practice, conventional corneal topography could not be replaced by Corvis ST.

KeratoScreen: Early Keratoconus Classification With Zernike Polynomial Using Deep Learning

PURPOSE

We aimed to investigate the usefulness of Zernike coefficients (ZCs) for distinguishing subclinical keratoconus (KC) from normal corneas and to evaluate the goodness of detection of the entire corneal topography and tomography characteristics with ZCs as a screening feature input set of artificial neural networks.

METHODS

This retrospective study was conducted at the Affiliated Eye Hospital of Wenzhou Medical University, China. A total of 208 patients (1040 corneal topography images) were evaluated. Data were collected between 2012 and 2018 using the Pentacam system and analyzed from February 2019 to December 2021. An artificial neural network (KeratoScreen) was trained using a data set of ZCs generated from corneal topography and tomography. Each image was previously assigned to 3 groups: normal (70 eyes; average age, 28.7 ± 2.6 years), subclinical KC (48 eyes; average age, 24.6 ± 5.7 years), and KC (90 eyes; average age, 25.9 ± 5.4 years). The data set was randomly split into 70% for training and 30% for testing. We evaluated the precision of screening symptoms and examined the discriminative capability of several combinations of the input set and nodes.

RESULTS

The best results were achieved using ZCs generated from corneal thickness as an input parameter, determining the 3 categories of clinical classification for each subject. The sensitivity and precision rates were 93.9% and 96.1% in subclinical KC cases and 97.6% and 95.1% in KC cases, respectively.

CONCLUSIONS

Deep learning algorithms based on ZCs could be used to screen for early KC and for other corneal ectasia during preoperative screening for corneal refractive surgery.

Patient selection for corneal topographic evaluation of keratoconus: A screening approach using artificial intelligence

Background

Corneal topography is a clinically validated examination method for keratoconus. However, there is no clear guideline regarding patient selection for corneal topography. We developed and validated a novel artificial intelligence (AI) model to identify patients who would benefit from corneal topography based on basic ophthalmologic examinations, including a survey of visual impairment, best-corrected visual acuity (BCVA) measurement, intraocular pressure (IOP) measurement, and autokeratometry.

Methods

A total of five AI models (three individual models with fully connected neural network including the XGBoost, and the TabNet models, and two ensemble models with hard and soft voting methods) were trained and validated. We used three datasets collected from the records of 2,613 patients' basic ophthalmologic examinations from two institutions to train and validate the AI models. We trained the AI models using a dataset from a third medical institution to determine whether corneal topography was needed to detect keratoconus. Finally, prospective intra-validation dataset (internal test dataset) and extra-validation dataset from a different medical institution (external test dataset) were used to assess the performance of the AI models.

Results

The ensemble model with soft voting method outperformed all other AI models in sensitivity when predicting which patients needed corneal topography (90.5% in internal test dataset and 96.4% in external test dataset). In the error analysis, most of the predicting error occurred within the range of the subclinical keratoconus and the suspicious D-score in the Belin-Ambrósio enhanced ectasia display. In the feature importance analysis, out of 18 features, IOP was the highest ranked feature when comparing the average value of the relative attributions of three individual AI models, followed by the difference in the value of mean corneal power.

Conclusion

An AI model using the results of basic ophthalmologic examination has the potential to recommend corneal topography for keratoconus. In this AI algorithm, IOP and the difference between the two eyes, which may be undervalued clinical information, were important factors in the success of the AI model, and may be worth further reviewing in research and clinical practice for keratoconus screening.

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.].

Fourier analysis on irregular corneal astigmatism using optical coherence tomography in various severity stages of keratoconus

PURPOSE

To investigate the diagnostic capability of Fourier indices in detecting clinical or subclinical keratoconus (KC).

DESIGN

Prospective cross-sectional study METHODS: : The study included 126 eyes with clinical KC (50 KC without any corneal scar, 50 KC with anterior corneal scar, and 26 KC with posterior scar having a history of acute corneal hydrops), 50 with topographic KC (without clinical signs), 50 with pre-topographic KC (normal topography without clinical signs), and 50 controls. Corneal tomographic data were obtained using anterior segment optical coherence tomography (OCT). Fourier analysis decomposed dioptric data from both anterior and posterior corneal surface into spherical, regular astigmatism, asymmetry, and higher-order irregularity components. The discriminating ability of the Fourier indices of pre-topographic KC, topographic KC, and clinical KC from controls were assessed after quantitative Fourier analysis of irregular corneal astigmatism.

RESULTS

Posterior asymmetry and higher-order irregularity components were significantly greater in pre-topographic KC eyes than those in controls (p<0.001 for both), with the highest area under the receiver operating characteristic curve (AUROC) of 0.778 and 0.709, respectively. The same was true for anterior asymmetry, posterior asymmetry, and posterior higher-order irregularity components in topographic KC (AUROC of 0.945, 0.941, and 0.893, respectively), whereas it was >0.948 for all Fourier components in clinical KC.

CONCLUSIONS

Fourier analysis using OCT can evaluate anterior and posterior corneal irregular astigmatism of various KC stages, from very mild to advanced, including severe cases with corneal scar. Irregular astigmatism indices from the posterior corneal surface showed the highest AUROC values in discriminating early KC stages.

Evaluation of artificial intelligence models for the detection of asymmetric keratoconus eyes using Scheimpflug tomography

BACKGROUND

To evaluate artificial intelligence (AI) models based on objective indices and raw corneal data from the Scheimpflug Pentacam HR system (OCULUS Optikgeräte GmbH, Wetzlar, Germany) for the detection of clinically unaffected eyes in patients with asymmetric keratoconus (AKC) eyes.

METHODS

A total of 1108 eyes of 1108 patients were enrolled, including 430 eyes from normal control subjects, 231 clinically unaffected eyes from patients with AKC, and 447 eyes from keratoconus (KC) patients. Eyes were divided into a training set (664 eyes), a test set (222 eyes) and a validation set (222 eyes). AI models were built based on objective indices (XGBoost, LGBM, LR and RF) and entire corneal raw data (KerNet). The discriminating performances of the AI models were evaluated by accuracy and the area under the ROC curve (AUC).

RESULTS

The KerNet model showed great overall discriminating power in the test (accuracy = 94.67%, AUC = 0.985) and validation (accuracy = 94.12%, AUC = 0.990) sets, which were higher than the index-derived AI models (accuracy = 84.02%-86.98%, AUC = 0.944-0.968). In the test set, the KerNet model demonstrated good diagnostic power for the AKC group (accuracy = 95.24%, AUC = 0.984). The validation set also proved that the KerNet model was useful for AKC group diagnosis (accuracy = 94.12%, AUC = 0.983).

CONCLUSIONS

KerNet outperformed all the index-derived AI models. Based on the raw data of the entire cornea, KerNet was helpful for distinguishing clinically unaffected eyes in patients with AKC from normal eyes.

Characteristics of Higher-Order Aberrations in Different Stages of Keratoconus

OBJECTIVES

To characterize higher-order aberrations (HOAs) in clinical and subclinical keratoconus (KC).

METHODS

The study included 33, 36, and 26 patients with clinical, topographic (no clinical signs), and pretopographic (normal topography and no clinical signs) KC and 30 controls. Ocular and corneal HOAs for the 4-mm pupils were measured using a wavefront sensor and expanded up to the sixth order of Zernike polynomials. The magnitudes of trefoil, coma, tetrafoil, secondary astigmatism, and spherical aberration were calculated via Zernike vector analysis and used as HOA parameters along with total HOAs. Area under the receiver operating characteristic curve (AUROC) values for each wavefront parameter for pretopographic KC were compared.

RESULTS

Control eyes and eyes with pretopographic KC had significantly lower ocular or corneal total HOAs and Zernike vector terms than those with clinical KC and topographic KC, except for ocular tetrafoil between topographic KC and pretopographic KC and spherical aberration among all groups. The AUROCs for corneal total HOAs and corneal coma for pretopographic KC and control eyes were 0.781 (100% sensitivity and 47% specificity) and 0.735 (73% sensitivity and 73% specificity), respectively.

CONCLUSION

Corneal total HOAs and corneal coma exhibited a potential ability to discriminate pretopographic KC from normal control eyes.

Long-Term Clinical Outcomes of Small-Incision Femtosecond Laser-Assisted Intracorneal Concave Lenticule Implantation in Patients with Keratoconus

Purpose

The purpose of this study was to evaluate the long-term prognosis of small-incision femtosecond laser-assisted intracorneal concave lenticule implantation (SFII) in correction of human keratoconus.

Methods

This was a prospective study for 11 patients who received SFII after being diagnosed as progressive keratoconus based on the Amsler–Krumeich classification system. Clinical assessment was performed for all the patients prior to and postsurgically at different time points for 5 years. These included uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), biomechanically corrected intraocular pressure (bIOP), corneal topography, anterior segment optical coherence tomography (AS-OCT), confocal microscopy, and biomechanical assessment with Corvis ST.

Results

Comparison of preoperative and 60-month postoperative UDVA and CDVA (P_60months=0.081 and 0.001, respectively), all eyes showed an improvement in CDVA. Corneal topography showed no significant changes in corneal anterior K1, K2, posterior K1, K2, posterior elevation, or corneal densitometry compared with preoperative levels (P > 0.05). Corvis ST showed that central corneal thickness (CCT) and stiffness at applanation 1 (SP-A1) were significantly greater 1 week postsurgically when compared to the baseline (P < 0.05) and remained stable thereafter. The lenticule under the AS-OCT remained transparent throughout the entire postsurgical period. Under confocal microscopy, corneal edema and an increase in cell activation and reflectivity were observed at the lenticule-stromal interface within 1 week postoperatively. These reactions gradually subsided with time within 6 months.

Conclusion

SFII is an effective procedure to prevent the progression of keratoconus due to its minimal invasiveness and capability of maintaining a steady biometry of the cornea.

The false positive rates for detecting keratoconus and potential ectatic corneal conditions when evaluating astigmatic eyes with Scheimpflug Technology

PURPOSE

To quantify the false positive rates for keratoconus (KC) and potential ectatic corneal conditions in highly astigmatism eyes when using published parameters/indices obtained from the Pentacam and Galilei units.

SETTING

Oftalmosalud Instituto de Ojos, Lima, Peru.

DESIGN

Prospective cohort study.

METHODS

67 consecutive eyes with corneal astigmatism > 1.5 D, with a minimum follow ups of 36 months after an uneventful LASIK procedure were included. Indices for KC and other potential ectatic corneal conditions (subclinical KC, forme fruste KC, suspect KC) were obtained using the Pentacam and Galilei Scheimpflug cameras.

MAIN OUTCOME MEASURES

The false positive rates for KC and potential ectatic corneal conditions were measured. Cut off values provided by previous studies and company-based parameters were used to assess the rate of false positivity.

RESULTS

The range of false positive rates for a KC diagnosis depending on the lowest and highest cutoff values were: index of height decentration (61% - 1%), index of surface variance (76% - 0%), Posterior elevation (55% - 0%), maximum Ambrosio Relational thickness (100% - 13%), Belin Ambrosio enhanced ectasia display total deviation value (100% - 4%), Average pachymetric progression index (69% - 3%), Pachymetry at the thinnest point (58% - 1%), CSI Center Surround Index (100%), Differential sector index (51%).

CONCLUSION

The false positive rates for KC and ectatic corneal conditions vary dramatically depending on the cut-off values used. Some indexes used for diagnosis of potential ectatic corneal conditions are inaccurate in normal, highly astigmatic eyes.

Evaluation of corneal topography, pachymetry and higher order aberrations for detecting subclinical keratoconus

PURPOSE

To compare corneal topography, pachymetry and higher order aberrations in keratoconic and normal eyes; to investigate their association in keratoconic eyes; and to determine their diagnostic ability for detecting subclinical keratoconus in a Nepalese population.

METHODS

Ninety-six eyes of 48 keratoconus patients and 50 normal eyes of 50 control subjects were included in this study. The eyes of keratoconus patients were classified into four different study groups: subclinical, stage 1, stage 2 and advanced stage keratoconus. In each eye, corneal topography, pachymetry and corneal aberrometry indices were measured using a Sirius corneal tomographer. The study parameters of keratoconic eyes were compared with normal eyes, and the possible association of corneal aberrometry with topography and pachymetry indices was investigated. The area under curve (AUC) of receiver operating characteristic (ROC) curves along with optimal cutoff values with best sensitivity and specificity were also determined for each index to detect subclinical keratoconus.

RESULTS

All the indices except average keratometry measurements (K_avg and mm_avg ) and spherical aberration (SA) were found to be significantly different in subclinical keratoconus compared to the control group (p < 0.05). In keratoconic eyes, all corneal aberrations were significantly correlated with the topography and pachymetry indices (range of ρ: -0.25 to 0.96; all p < 0.05) except for trefoil and minimum corneal thickness (Thk_min ). All the indices except K_avg , mm_avg and SA showed excellent diagnostic ability (AUC > 0.90) in detecting subclinical keratoconus. The cutoff values proposed for the asymmetry index of the corneal back surface (SI_b ), Strehl ratio of point spread function (PSF), coma and Baiocchi-Calossi-Versaci index of corneal back surface (BCV_b ) each showed excellent sensitivity (100%) and specificity (≥97%).

CONCLUSIONS

Corneal higher order aberrations were found to be significantly elevated in subclinical keratoconus compared to healthy controls. SI_b , PSF, coma and BCV_b were identified as the most powerful Sirius indices for the detection of subclinical keratoconus.

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.].

Mapping of corneal birefringence in thin and asymmetric keratoconus corneas with ultrahigh resolution polarization sensitive OCT

PURPOSE

To evaluate phase retardation (PR) across healthy, thin corneas (< 500 µm), asymmetric and bilateral KC.

SETTING

Narayana Nethralaya, Bangalore.

DESIGN

Observational, cross-sectional.

METHODS

There were four eye groups: healthy (group 1; n=10 eyes), thin corneas with no clinical disease (group 2; n=10 eyes), asymmetric KC (group 3; n=5 eyes) and clinical KC (group 4; n=15 eyes). All eyes were imaged with polarization sensitive OCT (PS-OCT), MS-39 and Corvis-ST. Using PS-OCT, the phase retardation (PR) was analysed in annular regions. The anterior (A-E) and Bowman's (E-B) wavefront aberrations, Epithelium Zernike indices (EZI), total corneal thickness, Corvis biomechanical index (CBI), total biomechanical index (TBI) and Belin-Ambrosio overall deviation index (BAD-D) were analysed.

RESULTS

Only the CBI, TBI, BAD-D, the A-E and E-B aberrations, EZI and total corneal thickness distributions of groups 1, 2 and 3 were similar (p>0.05) but not CCT (p<0.05). The PR distributions clearly showed that the eyes in groups 1, 2 and 3 had a normal corneal birefringence unlike group 4 eyes (p<0.05). The PR map was similar to the preferred orientations of collagen fibers seen in X-ray diffraction ex vivo studies of corneal stroma.

CONCLUSION

The PR distributions may eliminate the uncertainty associated with the stromal status of thin and asymmetric KC corneas. The group 2 and 3 eyes appeared as healthy due to normal corneal birefringence at the time of imaging and longitudinal follow-up of these eyes with PS-OCT may assist in early detection of onset of disease.

Accuracy of Machine Learning Assisted Detection of Keratoconus: A Systematic Review and Meta-Analysis

(1) Background: The objective of this review was to synthesize available data on the use of machine learning to evaluate its accuracy (as determined by pooled sensitivity and specificity) in detecting keratoconus (KC), and measure reporting completeness of machine learning models in KC based on TRIPOD (the transparent reporting of multivariable prediction models for individual prognosis or diagnosis) statement. (2) Methods: Two independent reviewers searched the electronic databases for all potential articles on machine learning and KC published prior to 2021. The TRIPOD 29-item checklist was used to evaluate the adherence to reporting guidelines of the studies, and the adherence rate to each item was computed. We conducted a meta-analysis to determine the pooled sensitivity and specificity of machine learning models for detecting KC. (3) Results: Thirty-five studies were included in this review. Thirty studies evaluated machine learning models for detecting KC eyes from controls and 14 studies evaluated machine learning models for detecting early KC eyes from controls. The pooled sensitivity for detecting KC was 0.970 (95% CI 0.949–0.982), with a pooled specificity of 0.985 (95% CI 0.971–0.993), whereas the pooled sensitivity of detecting early KC was 0.882 (95% CI 0.822–0.923), with a pooled specificity of 0.947 (95% CI 0.914–0.967). Between 3% and 48% of TRIPOD items were adhered to in studies, and the average (median) adherence rate for a single TRIPOD item was 23% across all studies. (4) Conclusions: Application of machine learning model has the potential to make the diagnosis and monitoring of KC more efficient, resulting in reduced vision loss to the patients. This review provides current information on the machine learning models that have been developed for detecting KC and early KC. Presently, the machine learning models performed poorly in identifying early KC from control eyes and many of these research studies did not follow established reporting standards, thus resulting in the failure of these clinical translation of these machine learning models. We present possible approaches for future studies for improvement in studies related to both KC and early KC models to more efficiently and widely utilize machine learning models for diagnostic process.

In Vivo Confocal Microscopy Evaluation in Patients with Keratoconus

Keratoconus is the most common primary corneal ectasia characterized by progressive focal thinning. Patients experience increased irregular astigmatism, decreased visual acuity and corneal sensitivity. Corneal collagen crosslinking (CXL), a minimally invasive procedure, is effective in halting disease progression. Historically, keratoconus research was confined to ex vivo settings. In vivo confocal microscopy (IVCM) has been used to examine the corneal microstructure clinically. In this review, we discuss keratoconus cellular changes evaluated by IVCM before and after CXL. Cellular changes before CXL include decreased keratocyte and nerve densities, disorganized subbasal nerves with thickening, increased nerve tortuosity and shortened nerve fibre length. Repopulation of keratocytes occurs up to 1 year post procedure. IVCM also correlates corneal nerve status to functional corneal sensitivity. Immediately after CXL, there is reduced nerve density and keratocyte absence due to mechanical removal of the epithelium and CXL effect. Nerve regeneration begins after 1 month, with nerve fibre densities recovering to pre-operative levels between 6 months to 1 year and remains stable up to 5 years. Nerves remain tortuous and nerve densities are reduced. Corneal sensitivity is reduced immediately postoperatively but recovers with nerve regeneration. Our article provides comprehensive review on the use of IVCM imaging in keratoconus patients.

Keratoconus: An updated review

Keratoconus is a bilateral and asymmetric disease which results in progressive thinning and steeping of the cornea leading to irregular astigmatism and decreased visual acuity. Traditionally, the condition has been described as a noninflammatory disease; however, more recently it has been associated with ocular inflammation. Keratoconus normally develops in the second and third decades of life and progresses until the fourth decade. The condition affects all ethnicities and both sexes. The prevalence and incidence rates of keratoconus have been estimated to be between 0.2 and 4,790 per 100,000 persons and 1.5 and 25 cases per 100,000 persons/year, respectively, with highest rates typically occurring in 20- to 30-year-olds and Middle Eastern and Asian ethnicities. Progressive stromal thinning, rupture of the anterior limiting membrane, and subsequent ectasia of the central/paracentral cornea are the most commonly observed histopathological findings. A family history of keratoconus, eye rubbing, eczema, asthma, and allergy are risk factors for developing keratoconus. Detecting keratoconus in its earliest stages remains a challenge. Corneal topography is the primary diagnostic tool for keratoconus detection. In incipient cases, however, the use of a single parameter to diagnose keratoconus is insufficient, and in addition to corneal topography, corneal pachymetry and higher order aberration data are now commonly used. Keratoconus severity and progression may be classified based on morphological features and disease evolution, ocular signs, and index-based systems. Keratoconus treatment varies depending on disease severity and progression. Mild cases are typically treated with spectacles, moderate cases with contact lenses, while severe cases that cannot be managed with scleral contact lenses may require corneal surgery. Mild to moderate cases of progressive keratoconus may also be treated surgically, most commonly with corneal cross-linking. This article provides an updated review on the definition, epidemiology, histopathology, aetiology and pathogenesis, clinical features, detection, classification, and management and treatment strategies for keratoconus.

A Multicenter Study of the Distribution Pattern of Posterior-To-Anterior Corneal Curvature Radii Ratio in Chinese Myopic Patients

Estimation of corneal refractive power (CRP) is of crucial importance to refractive and cataract surgery. The ratio of posterior to anterior curvature radii of the cornea (P/A ratio) is one of the key factors to determine the actual CRP (True-K). While the traditional method to calculate the CRP (Sim-K) is based on a constant P/A ratio (0.82), it is suggested that the P/A ratio varies in different people and exhibits a distribution pattern, which may have an impact on the accuracy of CRP estimation and postoperative refractive outcome. In this multicenter study, we aimed to investigate the distribution pattern of the P/A ratio in a large number of myopic patients, and further explore the relationship between P/A ratio and ΔK (the difference between True-K and Sim-K). We found that distribution of the P/A ratio ranged from 0.72 to 0.86 with an average value of 0.82 ± 0.01. The compensation effect of the refractive power of the posterior on the anterior surface of the cornea decreased with the increase of P/A ratio. There was a significant correlation between P/A ratio and ΔK in all eyes (r = 0.9764, P &lt; 0.0001). A change of 0.1 in P/A ratio could cause a change of 0.75 D in ΔK. Our study suggests that the actual P/A ratio should be taken into consideration in refractive and cataract surgery when calculating the CRP and power of the intraocular lens in eyes with significantly deviated P/A ratios.

Universal architecture of corneal segmental tomography biomarkers for artificial intelligence-driven diagnosis of early keratoconus

AIMS

To develop a comprehensive three-dimensional analyses of segmental tomography (placido and optical coherence tomography) using artificial intelligence (AI).

METHODS

Preoperative imaging data (MS-39, CSO, Italy) of refractive surgery patients with stable outcomes and diagnosed with asymmetric or bilateral keratoconus (KC) were used. The curvature, wavefront aberrations and thickness distributions were analysed with Zernike polynomials (ZP) and a random forest (RF) AI model. For training and cross-validation, there were groups of healthy (n=527), very asymmetric ectasia (VAE; n=144) and KC (n=454). The VAE eyes were the fellow eyes of KC patients but no further manual segregation of these eyes into subclinical or forme-fruste was performed.

RESULTS

The AI achieved an excellent area under the curve (0.994), accuracy (95.6%), recall (98.5%) and precision (92.7%) for the healthy eyes. For the KC eyes, the same were 0.997, 99.1%, 98.7% and 99.1%, respectively. For the VAE eyes, the same were 0.976, 95.5%, 71.5% and 91.2%, respectively. Interestingly, the AI reclassified 36 (subclinical) of the VAE eyes as healthy though these eyes were distinct from healthy eyes. Most of the remaining VAE (n=104; forme fruste) eyes retained their classification, and were distinct from both KC and healthy eyes. Further, the posterior surface features were not among the highest ranked variables by the AI model.

CONCLUSIONS

A universal architecture of combining segmental tomography with ZP and AI was developed. It achieved an excellent classification of healthy and KC eyes. The AI efficiently classified the VAE eyes as 'subclinical' and 'forme-fruste'.

Machine Learning Algorithms to Detect Subclinical Keratoconus: Systematic Review

BACKGROUND

Keratoconus is a disorder characterized by progressive thinning and distortion of the cornea. If detected at an early stage, corneal collagen cross-linking can prevent disease progression and further visual loss. Although advanced forms are easily detected, reliable identification of subclinical disease can be problematic. Several different machine learning algorithms have been used to improve the detection of subclinical keratoconus based on the analysis of multiple types of clinical measures, such as corneal imaging, aberrometry, or biomechanical measurements.

OBJECTIVE

The aim of this study is to survey and critically evaluate the literature on the algorithmic detection of subclinical keratoconus and equivalent definitions.

METHODS

For this systematic review, we performed a structured search of the following databases: MEDLINE, Embase, and Web of Science and Cochrane Library from January 1, 2010, to October 31, 2020. We included all full-text studies that have used algorithms for the detection of subclinical keratoconus and excluded studies that did not perform validation. This systematic review followed the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) recommendations.

RESULTS

We compared the measured parameters and the design of the machine learning algorithms reported in 26 papers that met the inclusion criteria. All salient information required for detailed comparison, including diagnostic criteria, demographic data, sample size, acquisition system, validation details, parameter inputs, machine learning algorithm, and key results are reported in this study.

CONCLUSIONS

Machine learning has the potential to improve the detection of subclinical keratoconus or early keratoconus in routine ophthalmic practice. Currently, there is no consensus regarding the corneal parameters that should be included for assessment and the optimal design for the machine learning algorithm. We have identified avenues for further research to improve early detection and stratification of patients for early treatment to prevent disease progression.

Interleukin-6 and tumor necrosis factor-α levels in tear film of Keratoconus patients

Background:

It is hypothesized that increased inflammatory markers in keratoconus (KC) may be one of the causes of corneal damage. The aim of our study was to the measurement of tumor necrosis factor-alpha (TNF-α) and interleukin-6 (IL)-6 in tear of patients with KC and investigate their relationship with the severity of KC.

Materials and Methods:

The current study was performed on KC patients and healthy controls with a case-control setting. Tear levels of TNF-α and IL-6 were measured after collecting the tears from the tear lake using a filter paper via Schirmer I method without anesthesia.

Results:

Eighty-one KC patients (mean age 29.45 ± 5.06 years) and 85 controls (mean age 28.01 ± 5.14 years) were enrolled. The mean levels of IL-6 and TNF-α were 26.77 ± 8.16, and 34.58 ± 9.82 pg/ml in the healthy group and 103.22 ± 51.94, and 183.76 ± 54.61 pg/ml in the KC group, respectively (P < 0.001). There was a significant relationship between the severity of the KC and the mean levels of IL-6 TNF-α in the case group (P < 0.001).

Conclusion:

Our results indicated that the mean levels of IL-6 and TNF-α are significantly higher in KC than the healthy group, and the disease severity was significantly associated with TNF-α and IL-6.

Machine learning with a reduced dimensionality representation of comprehensive Pentacam tomography parameters to identify subclinical keratoconus

PURPOSE

To investigate the performance of a machine learning model based on a reduced dimensionality parameter space derived from complete Pentacam parameters to identify subclinical keratoconus (KC).

METHODS

All 1692 available parameters were obtained from the Pentacam imaging machine on 145 subclinical KC and 122 control eyes. We applied a principal component analysis (PCA) to the complete Pentacam dataset to reduce its parameter dimensionality. Subsequently, we investigated machine learning performance of the random forest algorithm with increasing numbers of components to identify their optimal number for detecting subclinical KC from control eyes.

RESULTS

The dimensionality of the complete set of 1692 Pentacam parameters was reduced to 267 principal components using PCA. Subsequent selection of 15 of these principal components explained over 85% of the variance of the original Pentacam-derived parameters and input to train a random forest machine learning model to achieve the best accuracy of 98% in detecting subclinical KC eyes. The model established also reached a high sensitivity of 97% in identification of subclinical KC and a specificity of 98% in recognizing control eyes.

CONCLUSIONS

A random forest-based model trained using a modest number of components derived from a reduced dimensionality representation of complete Pentacam system parameters allowed for high accuracy of subclinical KC identification.

Bibliometric analysis of the keratoconus literature

CLINICAL RELEVANCE

Clinicians, researchers funding agencies and indeed the general public can benefit from knowledge of the most highly cited papers and most impactful authors, institutions, countries and journals in the field of keratoconus.

BACKGROUND

Bibliometrics relating to the keratoconus literature were derived to enable identification of the most impactful papers published, as well as the leading authors, institutions, countries and journals.

METHODS

A search was undertaken of the titles of papers on the Scopus database to identify keratoconus-related articles. The 20 most highly cited papers were determined from the total list of 4,419 papers found. Rank-order lists by count were assembled for the 'top 20' in each of four categories: authors, institutions, countries and journals. A subject-specific keratoconus-related h-index (h_KC-index) was derived for each constituent of each category to serve as a measure of impact in the field. The top 10 constituents of each category were ranked by h_KC-index and tabulated for consideration.

RESULTS

The h_KC-index of the keratoconus field is 125. The 4,419 papers have been cited a total of 98,010 times, and 18.5% of these papers have never been cited. The most highly cited paper is a general review of keratoconus by Yaron Rabinowitz, who is also the most impactful author in the field (h_KC = 31). The Cedars Sinai Medical Center in the United States produces the most impactful keratoconus-related papers (h_KC = 36), and the United States is the most impactful country (h_KC = 91). The Journal of Cataract and Refractive Surgery is the most impactful journal (h_KC = 55).

CONCLUSION

Keratoconus is a topic of high interest in the clinical and scientific literature. Highly cited papers and impactful authors, institutions, countries and journals are identified.

Increased epithelial backscatter: A novel finding in subclinical and clinical keratoconus

BACKGROUND

To assess alterations in backscatter from the corneal epithelium, anterior stroma and lens surface in eyes with subclinical, mild and moderate keratoconus (KC).

METHODS

In this single-centre, cross-sectional study involving 24 eyes with subclinical KC, 107 eyes with manifest KC (mild = 40 and moderate = 67 eyes) and 90 controls, line densitometry was performed with Pentacam (Oculus Optikgeräte GmbH, Wetzlar, Germany) to obtain simultaneous backscatter values for the corneal epithelium, anterior stroma and anterior lens surface. Backscatter values and Pentacam parameters were used in subsequent statistical analyses.

RESULTS

Eyes with subclinical, mild and moderate KC had similar epithelial and stromal backscatter (P > 0.05) that was significantly increased compared with the controls (P < 0.05). Although anterior lens surface backscatter did not differ between the control and KC groups (P > 0.05), it was significantly higher in the mild and moderate KC groups than in the subclinical KC group (P < 0.05). In the KC group (n = 131) epithelial backscatter was strongly correlated with stromal backscatter (r = 0.911, P < 0.0001).

CONCLUSIONS

Increased epithelial backscatter and a strong correlation with anterior stromal backscatter in the KC groups were consistent with the epithelium-stroma interaction involved in KC pathogenesis. Single-point backscatter analysis can be used with point clouds to construct epithelial and stromal backscatter maps in Pentacam to aid the detection of KC as a novel feature.