Diagnostic Performance of Three-Dimensional Endothelium/Descemet Membrane Complex Thickness Maps in Active Corneal Graft Rejection

Main issues in penetrating keratoplasty

This review explores contemporary challenges in penetrating keratoplasty (PK), focusing on technical intricacies, technological advancements, and strategies for preventing graft rejection. A systematic literature search from January 2018 to July 2023 was conducted across PubMed, Cochrane, Web of Science, Scopus, and EMBASE. The inclusion criteria comprised studies on PK and its comparison with other corneal pathologies, with emphasis on keratoconus (KC). Two independent reviewers screened studies, extracting relevant data. The review covers PK evolution, highlighting infra-red femtosecond lasers' impact on graft shapes, minimizing astigmatism, and enhancing wound healing. Graft rejection, a primary complication, is examined, detailing risk factors and preventive measures. Preoperative considerations, diagnostic techniques for rejection, and PK in KC are discussed. Postoperative care's significance, including intraocular pressure monitoring and steroid administration, is emphasized. The paper concludes with a comprehensive approach to prevent graft rejection, involving topical and systemic medications. An outlook on evolving monoclonal antibody research is presented. As the field progresses, personalized approaches and ongoing therapeutic exploration are expected to refine strategies, enhancing PK outcomes.

Correlation of Clinical Fibrillar Layer Detection and Corneal Thickness in Advanced Fuchs Endothelial Corneal Dystrophy

Central subendothelial geographic deposits are formed as a fibrillar layer (FL) in advanced Fuchs endothelial corneal dystrophy (FECD). Previous studies demonstrated a significant decrease in corneal endothelial cell (CEC) density and an increase in focal corneal backscatter in the FL area. The present study investigated the association of the FL with edema formation and its localization. Patients (n = 96) presenting for Descemet membrane endothelial keratoplasty (DMEK) for advanced FECD were included. Slit-lamp biomicroscopy with FECD grading was followed by Scheimpflug imaging with en face backscatter analysis and pachymetric analysis. FL dimensions were measured, and correlation with pachymetric values was performed. An FL was detected in 74% of all eyes (n = 71). Pachymetric values in FL-positive versus FL-negative eyes were for corneal thickness at the apex (ACT) 614 ± 52 µm and 575 ± 46 µm (p = 0.001), for peripheral corneal thickness at 1 mm (PCT_1mm) 616 ± 50 µm and 580 ± 44 µm (p = 0.002), for PCT_2mm 625 ± 48 µm and 599 ± 41 µm (p = 0.017), for PCT_3mm 651 ± 46 µm and 635 ± 40 µm (p = 0.128) and for PCT_4mm 695 ± 52 µm and 686 ± 43 µm (p = 0.435), respectively. Correlation analysis indicated a weak correlation for the FL maximum vertical caliper diameter with ACT and PCT_1mm values but no further relevant correlations. In FL-positive eyes, increased focal corneal backscatter and increased corneal thickness showed primarily central and inferotemporal localization. In conclusion, Scheimpflug imaging shows an association of the FL with increased corneal thickness in advanced FECD and shows localization of the FL and increased corneal thickness in the central and inferotemporal region. This may provide important information for progression assessment and therapeutic decision making in FECD patients in the future.

Comparison of Autonomous AS-OCT Deep Learning Algorithm and Clinical Dry Eye Tests in Diagnosis of Dry Eye Disease

Objective

To evaluate a deep learning-based method to autonomously detect dry eye disease (DED) in anterior segment optical coherence tomography (AS-OCT) images compared to common clinical dry eye tests.

Methods

In this study, 27,180 AS-OCT images were prospectively collected from 151 eyes of 91 patients. Images were used to train and test the deep learning model. Masked cornea specialist ophthalmologist diagnoses were used as the gold standard. Clinical dry eye tests were performed on patients in the DED group to compare the results of the model. The dry eye tests performed were tear break-up time (TBUT), Schirmer's test, corneal staining, conjunctival staining, and Ocular Surface Disease Index (OSDI).

Results

Our deep learning model achieved an accuracy of 84.62%, sensitivity of 86.36%, and specificity of 82.35% in the diagnosis of DED. The positive likelihood ratio was 4.89, and the negative likelihood ratio was 0.17. The mean DED probability score was 0.81 ± 0.23 in the DED group and 0.20 ± 0.27 in the healthy group (P < 0.01). The deep learning model accuracy in the diagnosis of DED was significantly better than that of corneal staining, conjunctival staining, and Schirmer's test (P < 0.05). There was no significant difference between the deep learning diagnostic accuracy and that of the OSDI and TBUT.

Conclusion

Based on preliminary results, reliable autonomous diagnosis of DED with our deep learning model was achieved, when compared with standard dry eye clinical tests that correlated significantly more or similarly to diagnoses made by cornea specialist ophthalmologists.

Separate Detection of Stromal and Epithelial Corneal Edema on Optical Coherence Tomography Using a Deep Learning Pipeline and Transfer Learning

The accurate detection of corneal edema has become a topic of growing interest with the generalization of endothelial keratoplasty. Despite recent advances in deep learning for corneal edema detection, the problem of minimal edema remains challenging. Using transfer learning and a limited training set of 11 images, we built a model to segment the corneal epithelium, which is part of a three-model pipeline to detect corneal edema. A second and a third model are used to detect edema on the stroma alone and on the epithelium. A validation set of 233 images from 30 patients consisting of three groups (Normal, Minimal Edema and important Edema) was used to compare the results of our new pipeline to our previous model. The mean edema fraction (EF), defined as the number of pixels detected as edema divided by the total number of pixels of the cornea, was calculated for each image. With our previous model, the mean EF was not statistically different between the Normal and Minimal Edema groups (p = 0.24). With the current pipeline, the mean EF was higher in the Minimal Edema group compared to the Normal group (p < 0.01). The described pipeline constitutes an adjustable framework for the detection of corneal edema based on optical coherence tomography and yields better performances in cases of minimal or localized edema.

Prediction of corneal graft rejection using central endothelium/Descemet's membrane complex thickness in high-risk corneal transplants

To determine whether measurements of Endothelium/Descemet complex thickness (En/DMT) are of predictive value for corneal graft rejection after high-risk corneal transplantation, we conducted this prospective, single-center, observational case series including sixty eyes (60 patients) at high risk for corneal graft rejection (GR) because of previous immunologic graft failure or having at least two quadrants of stromal vascularization. Patients underwent corneal transplant. At 1st, 3rd, 6th, 9th, and 12th postoperative month, HD-OCT imaging of the cornea was performed, and the corneal status was determined clinically at each visit by a masked cornea specialist. Custom-built segmentation tomography algorithm was used to measure the central En/DMT. Relationships between baseline factors and En/DMT were explored. Time dependent covariate Cox survival regression was used to assess the effect of post-operative En/DMT changes during follow up. A longitudinal repeated measures model was used to assess the relationship between En/DMT and graft status. Outcome measures included graft rejection, central Endothelium/Descemet's complex thickness, and central corneal thickness (CCT). In patients with GR (35%), the central En/DMT increased significantly 5.3 months (95% CI: 2, 11) prior to the clinical diagnosis of GR, while it remained stable in patients without GR. During the 1-year follow up, the rejected grafts have higher mean pre-rejection En/DMTs (p = 0.01), compared to CCTs (p = 0.7). For En/DMT ≥ 18 µm cut-off (at any pre-rejection visit), the Cox proportional hazard ratio was 6.89 (95% CI: 2.03, 23.4; p = 0.002), and it increased to 9.91 (95% CI: 3.32, 29.6; p < 0.001) with a ≥ 19 µm cut-off. In high-risk corneal transplants, the increase in En/DMT allowed predicting rejection prior to the clinical diagnosis.

Machine learning helps improve diagnostic ability of subclinical keratoconus using Scheimpflug and OCT imaging modalities

Purpose

To develop an automated classification system using a machine learning classifier to distinguish clinically unaffected eyes in patients with keratoconus from a normal control population based on a combination of Scheimpflug camera images and ultra-high-resolution optical coherence tomography (UHR-OCT) imaging data.

Methods

A total of 121 eyes from 121 participants were classified by 2 cornea experts into 3 groups: normal (50 eyes), with keratoconus (38 eyes) or with subclinical keratoconus (33 eyes). All eyes were imaged with a Scheimpflug camera and UHR-OCT. Corneal morphological features were extracted from the imaging data. A neural network was used to train a model based on these features to distinguish the eyes with subclinical keratoconus from normal eyes. Fisher’s score was used to rank the differentiable power of each feature. The receiver operating characteristic (ROC) curves were calculated to obtain the area under the ROC curves (AUCs).

Results

The developed classification model used to combine all features from the Scheimpflug camera and UHR-OCT dramatically improved the differentiable power to discriminate between normal eyes and eyes with subclinical keratoconus (AUC = 0.93). The variation in the thickness profile within each individual in the corneal epithelium extracted from UHR-OCT imaging ranked the highest in differentiating eyes with subclinical keratoconus from normal eyes.

Conclusion

The automated classification system using machine learning based on the combination of Scheimpflug camera data and UHR-OCT imaging data showed excellent performance in discriminating eyes with subclinical keratoconus from normal eyes. The epithelial features extracted from the OCT images were the most valuable in the discrimination process. This classification system has the potential to improve the differentiable power of subclinical keratoconus and the efficiency of keratoconus screening.

Automated diagnosis and staging of Fuchs' endothelial cell corneal dystrophy using deep learning

Background

To describe the diagnostic performance of a deep learning algorithm in discriminating early-stage Fuchs’ endothelial corneal dystrophy (FECD) without clinically evident corneal edema from healthy and late-stage FECD eyes using high-definition optical coherence tomography (HD-OCT).

Methods

In this observational case-control study, 104 eyes (53 FECD eyes and 51 healthy controls) received HD-OCT imaging (Envisu R2210, Bioptigen, Buffalo Grove, IL, USA) using a 6 mm radial scan pattern centered on the corneal vertex. FECD was clinically categorized into early (without corneal edema) and late-stage (with corneal edema). A total of 18,720 anterior segment optical coherence tomography (AS-OCT) images (9180 healthy; 5400 early-stage FECD; 4140 late-stage FECD) of 104 eyes (81 patients) were used to develop and validate a deep learning classification network to differentiate early-stage FECD eyes from healthy eyes and those with clinical edema. Using 5-fold cross-validation on the dataset containing 11,340 OCT images (63 eyes), the network was trained with 80% of these images (3420 healthy; 3060 early-stage FECD; 2700 late-stage FECD), then tested with 20% (720 healthy; 720 early-stage FECD; 720 late-stage FECD). Thereafter, a final model was trained with the entire dataset consisting the 11,340 images and validated with a remaining 7380 images of unseen AS-OCT scans of 41 eyes (5040 healthy; 1620 early-stage FECD 720 late-stage FECD). Visualization of learned features was done, and area under curve (AUC), specificity, and sensitivity of the prediction outputs for healthy, early and late-stage FECD were computed.

Results

The final model achieved an AUC of 0.997 ± 0.005 with 91% sensitivity and 97% specificity in detecting early-FECD; an AUC of 0.974 ± 0.005 with a specificity of 92% and a sensitivity up to 100% in detecting late-stage FECD; and an AUC of 0.998 ± 0.001 with a specificity 98% and a sensitivity of 99% in discriminating healthy corneas from all FECD.

Conclusion

Deep learning algorithm is an accurate autonomous novel diagnostic tool of FECD with very high sensitivity and specificity that can be used to grade FECD severity with high accuracy.

Diagnostic Performance of 3-Dimensional Thickness of the Endothelium-Descemet Complex in Fuchs' Endothelial Cell Corneal Dystrophy

PURPOSE

To describe the diagnostic accuracy of 3-dimensional (3D) endothelium-Descemet's membrane complex thickness (En-DMT) in Fuchs' endothelial corneal dystrophy (FECD) and determine its potential role as an objective index of disease severity.

DESIGN

Observational case-control study.

PARTICIPANTS

One hundred four eyes of 79 participants (64 eyes of 41 FECD patients and 40 eyes of 38 healthy controls).

METHODS

All participants received high-definition OCT imaging (Envisu R2210; Bioptigen, Buffalo Grove, IL). Fuchs' endothelial corneal dystrophy was classified clinically into early-stage (without edema) and late-stage (with edema) disease. Automatic and manual segmentation of corneal layers was performed using a custom-built segmental tomography algorithm to generate 3D maps of total corneal thickness (TCT) and En-DMT of the central 6-mm cornea. Regional En-DMT, regional TCT, and central-to-peripheral total corneal thickness ratio (CPTR) were evaluated and correlated to the clinical severity of FECD. Intraclass correlation coefficients (ICCs) and Bland-Altman plots were used to assess the reliability of the repeated measurements in all eyes.

MAIN OUTCOME MEASURES

Central-to-peripheral total corneal thickness ratio and average En-DMT and TCT of central, paracentral, and peripheral regions.

RESULTS

In FECD, a significant increase in En-DMT, CPTR_, and TCT was found compared to controls (P < 0.001). For identifying FECD, average En-DMT of paracentral and peripheral regions achieved 94% sensitivity and 100% specificity (cutoffs, 19 μm and 20 μm, respectively), whereas CPTR showed 94% sensitivity with a 73% specificity (cutoff, 0.97). Regarding early-stage FECD, average En-DMT of central zones achieved 92% sensitivity and 97% specificity (cutoff, 18 μm), whereas CPTR showed 90% sensitivity and 88% specificity (cutoff, 0.97). The average En-DMT of central, paracentral, and peripheral regions was correlated highly with FECD clinical stage (Spearman's ρ = 0.813, 0.793, and 0.721, respectively; all P < 0.001), compared with CPTR and mean TCT of paracentral zones (0.672 and 0.481, respectively; P < 0.001). The ICC values ranged from 0.98 (En-DMT) to 0.99 (TCT) with a good agreement between the automatic and manual measurements.

CONCLUSIONS

Regional 3D En-DMT is a novel diagnostic tool of FECD that can be used to quantify the disease severity with excellent reliability.