Assessment of corneal endothelial cell density in patients with keratoconus not using contact lenses

Deep Learning-Based Automatic Diagnosis of Keratoconus with Corneal Endothelium Image

INTRODUCTION

The primary objective of this study was to develop an end-to-end model that can accurately identify corneal endothelial cells and diagnose keratoconus based on corneal endothelial images acquired from a non-contact specular microscope.

METHODS

This was a retrospective case-control study performed at the Refractive Surgery Center of West China Hospital. A total of 403 keratoconus eyes (221 patients) and 370 myopic eyes (185 normal controls) were consecutively recruited from January 2021 to September 2022. Specular microscopy was used to image and measure the morphometric parameters of the corneal endothelial cells. A Fully Convolutional Network model with a ResNet50 (FCN_ResNet50) was established to perform the endothelial segmentation. The images were then classified using an ensemble machine learning system consisting of four pre-trained deep learning networks: DenseNet121, ResNet50, Inception_v3, and MobileNet_v2. The performance of the models was evaluated based on different metrics, such as accuracy, intersection over union (IoU), and mean IoU.

RESULTS

We established a fully end-to-end deep-learning model for the segmentation of endothelial and diagnosis of keratoconus. For endothelial segmentation, the accuracy of the FCN_ResNet50 model achieved near 90% with mean IoU converging to about 80%. The ensemble machine learning system can achieve over 92% accuracy, and > 98% area under curve (AUC) values to diagnose keratoconus with endothelial cell images. In addition, we constructed a diagnostic model based on deep-learning features and developed an associated nomogram which manifested an excellent performance for diagnosis and monitoring the progression of keratoconus.

CONCLUSIONS

Our research developed an end-to-end model to automatically identify and assess corneal endothelial morphological changes in keratoconus eyes. Moreover, we also constructed a novel nomogram, which can provide valuable information for the diagnosis, monitoring, and management of the disease.

Assessment and Correlation of Corneal Endothelial Cell Changes in Different Stages of Keratoconus in Non-contact Lens Wearers

SIGNIFICANCE

The endothelial changes in keratoconus are of interest because these can affect the surgical plan in management. Previously, many studies have evaluated the endothelium in eyes with keratoconus, but there is no common consensus on change in endothelial cell density (ECD) with studies showing variable results.

PURPOSE

This study aimed to compare and correlate endothelial cell parameters in different stages of keratoconus using specular microscope.

METHODS

This cross-sectional, cohort, comparative study included 162 eyes of 96 patients with keratoconus in one or both eyes. Corneal endothelial cell parameters were assessed with SP-1P specular microscope (Topcon Co., Tokyo, Japan). Corneal topography and thickness data were obtained from Sirius tomographer (Costruzione Strumenti Oftalmici, Florence, Italy). Eyes were divided into keratoconus stage 0 to 4 according to ABCD classification. Comparison and correlation of endothelial cell parameters in different stages were done.

RESULTS

There were 21.6% eyes (35) in stage 0, 29.6% eyes (48) in stage 1, 29.0% eyes (47) in stage 2, and 19.8% eyes (32) in stage 3 of keratoconus. Measurement was not possible in stage 4 keratoconus. Among the stages 0, 1, 2, and 3, the ECD was 3024, 3051, 3025, and 3043 cells/mm2; coefficient of variation was 27.2, 27.6, 26.8, and 27.4; and frequency of hexagon cells was 61.5, 63.7, 63.3, and 62.3, respectively (P > .05). The number of analyzed cells was 232, 209, 185, and 169 in stages 0, 1, 2, and 3, respectively (P < .001). No significant difference was found in minimum, maximum, and average cell area between the stages (P > .05). A weak Spearman rank correlation of ECD, coefficient of variation, and frequency of hexagon cells was found between eyes with keratoconus stages 0 and 1 and stages 2 and 3 (r = -0.05, P = .65; r = -0.11, P = .37; r = 0.05, P = .67, respectively). No significant correlation was found in the number of cells analyzed and minimum, maximum, and average cell area between the stages (P > .05).

CONCLUSIONS

Endothelial cell parameters do not show any significant changes and correlation up to stage 3 of keratoconus in non-contact lens wearers.

Corneal endothelial cells changes in different stages of Keratoconus: a multi-Centre clinical study

Purpose

To assess the corneal endothelial cells morphology and count in keratoconus patients and their correlation with different stages of keratoconus.

Methods

Prospective non randomized multi-centric clinical study included 150 eyes of 150 keratoconus patients. Four centers in Egypt participated in this study included: Departments of Ophthalmology in Alexandria University, Tanta University and Port Said University and Alex I-Care hospital. Pentacam (Wavelight Oculyzer II) and specular microscopy (Tomey EM-3000) were done to all eyes. Keratoconic eyes were classified according to Amsler classification into stage 1, 2 and 3. Stage 1 included 99 eyes, stage 2 included 32 eyes & stage 3 included 19 eyes.

Results

The mean age of keratoconus patients was 24.07 ± 6.154 years. Forty five cases were males (30%) and 105 cases were females (70%). There was statistically significant difference in endothelial cell density (p < 0.001) and coefficient of variation (p = 0.012) between different stages of keratoconus eyes. Regarding cell surface area, there was statistically significant difference in cell surface area between different stages of keratoconus eyes (p < 0.001). In addition, for cell morphology, there was statistically significant difference between different stages of keratoconus eyes (p < 0.001).

Conclusions

Qualitative and quantitative structural changes were seen in endothelial cells of keratoconus eyes by using specular microscopy. For stages 1 and 2, keratoconus may not affect the corneal endothelim significantly. The endothelium in stage 3 shows significant changes regarding polymegathism and pleomorphism.

A contralateral eye study comparing characteristics of corneal endothelial cells in bilateral keratoconus patients with unilateral corneal Vogt's striae

Purpose

The aim of this study was to analyze and compare corneal endothelial cell morphology and characteristics in bilateral keratoconus (KCN) patients with unilateral Vogt's striae.

Methods

Fifty patients aged 20–38 years were recruited in this cross-sectional contralateral eye study. In this study, corneal endothelial cell parameters were evaluated in patients with bilateral KCN and unilateral Vogt's striae using the Topcon SP2000P specular microscope (Topcon, Tokyo, Japan).

Results

In the current study, there were no significant differences in corneal endothelial cell parameters including endothelial cell density (ECD), hexagonal cell ratio (HEX), and coefficient of variance of cell size (CV) between the KCN groups with and without Vogt's striae, [(2968.34 ± 276.65 vs. 2980.05 ± 253.30, P = 0.618), (51.88 ± 13.57 vs. 53.24 ± 9.31, P = 0.658), and (32.50 ± 5.40 vs. 32.97 ± 4.07, P = 0.467), respectively]. Also, among study groups with and without Vogt's striae, ECD did not correlate with anterior chamber depth (ACD) [(P = 0.564, r = 0.09), (P = 0.219, r = −0.18), respectively], maximum keratometry (Kmax) [(P = 0.215, r = 0.18), (P = 0.898, r = 0.02), respectively], and central corneal thickness (CCT) [(P = 0.989, r = −0.02), (P = 0.643, r = −0.07), respectively].

Our results showed significant differences in corrected and uncorrected distance visual acuity (UDVA), cycloplegic refractive error components (calculated by vectorial analysis), CCT, and Kmax between two study groups (all P < 0.05) except for J₄₅ (Jackson cross cylinder, axes at 45 and 135°) (P = 0.131).

Conclusions

We were not able to find the statistically significant differences in ECD, HEX, and CV between KCN eyes with and without Vogt's striae. Despite clinical and tomographic results, it seems that Vogt's striae cannot cause deterioration in the corneal endothelial morphology.

Corneal imaging and densitometry measurements in keratoconus patients to monitor disease progression and treatment outcomes after contact lens or Intacs treatment

Aim

The aim of this study was to compare the pre- and posttreatment corneal densitometry and corneal thickness value of keratoconus (KCN) patients managed via contact lenses (CLs) or by both intrastromal corneal rings and contact lenses.

Patients and methods

This prospective study was performed at the Manchester Royal Eye Hospital, UK. Patients were recruited before treatment and followed up for 12 months. Data of corneal densitometry and corneal thickness were collected using the Oculus Pentacam at the pretreatment visit and posttreatment visit at 12 months.

Results

Corneal clarity significantly differs between both groups at pre treatment at zone 0-2 mm for the anterior layer (P=0.002). The same diversity is present at zone 2-6 mm for the anterior layer (P=0.003) and posterior layer (P=0.008). The corneal clarity diversity found was not statistically significant at 12 months post treatment (P>0.05). Corneal thickness was found to be statistically significantly different between pre treatment and post treatment for the CL group for central corneal thickness (CCT) and thinnest area (P=0.01 and P=0.02), respectively.

Discussion

This study shows that KCN management with Intacs was found to be effective in maintaining corneal clarity for a longer time than that with CL alone. On the other hand, corneal clarity reduces with disease progression in cases managed with CLs only. Analysis of Oculus Pentacam images provides an objective evaluation to monitor the corneal status after these different pathways of management.

Endothelial alterations in 712 keratoconus patients

PURPOSE

To investigate the effect of the severity of keratoconus on the corneal endothelium using specular microscopy.

METHODS

Seven hundred and twelve eyes from the Homburg Keratoconus Center (HKC) database were included in this retrospective study. Corneal endothelium was evaluated using the Tomey EM-3000 specular microscope. Keratoconus-related topographic and tomographic data were obtained from Scheimpflug-based tomography (Oculus Pentacam^® HR). Eyes were classified into stages 0 (healthy) to 4 (severe keratoconus) according to the Topographic Keratoconus Classification (TKC). Subgroups were analysed based on contact lens (CL) type (none/rigid/soft).

RESULTS

The frequencies of keratoconus stages 0/1/2/3/4 according to TKC were 169/94/206/166/77. The endothelial cell density (ECD) for the endothelial cell area for TKC 0/1/2/3/4 was 2611/2624/2557/2487/2401 cells per mm² and the coefficient of variation (CV) was 40.9/40.0/41.6/46.2/49.0%, respectively. The more severe the keratoconus stage, the lower the endothelial cell count (p < 0.001) and the higher the CV (p < 0.001). No contact lens wearing was noted in 207 eyes (NoCL), rigid CL in 200 (RCL) and soft CL in 54 (SCL). CD for NoCL/RCL/SCL was 2523/2533/2644 per mm² and CV was 41.8/54.1/43.1%, respectively. A significant difference in CV was found between NoCL and RCL (p = 0.02), and no significant difference in CV was found between NoCL and SCL (p = 0.07). Endothelial cell density (ECD) did not differ significantly between NoCL and RCL or SCL.

CONCLUSION

Endothelial cell density (ECD) decreases and CV increases significantly with increasing tomographic severity of keratoconus. In patients with RCL compared to eyes without CL wear, we found a statistically significantly higher CV in the endothelial cell size.

Correlation of Corneal Endothelial Cell Density with Corneal Tomographic Parameters in Eyes with Keratoconus

Objectives:

To examine changes in corneal endothelial cell density (ECD) in different stages of keratoconus and evaluate its correlation with corneal tomographic parameters.

Materials and Methods:

Two hundred six patients with keratoconus were enrolled in the study. Corneal topography was performed by Sirius (CSO, Italy), which has a rotating Scheimpflug camera and a Placido disc topographer. Automatic endothelial analysis was done with the non-contact endothelial microscope (20x probe) of Confoscan-4 (NIDEK, Japan). The eyes were classified into stages based on steepest keratometric value as follows: mild <45 D; moderate 45-52 D; severe >52 D and according to thinnest cornea thickness (TCT) as <400 μm, 400-450 μm, and >450 μm. Tomographic and endothelial cell parameters were compared among the groups using Kruskal-Wallis test and the correlations between them were analyzed using Spearman correlation.

Results:

The study included 391 eyes of 100 male (24.29±7.7 years, range 11-47 years) and 106 female (26.26±7.5 years, range 13-45 years) patients (p=0.07). Mean ECD values were 2628±262 cells/mm², 2541.9±260.4 cells/mm², and 2414.6±384.3 cells/mm² in mild, moderate, and severe keratoconus, respectively (p<0.001) and 2592.3±277 cells/mm², 2502±307 cells/mm² and 2348±296 cells/mm² in corneas with TCT values >450 µm, 400-450 µm, and <400 µm, respectively (p<0.001). ECD showed significant negative correlation with keratometric and elevation parameters and positive correlation with pachymetric parameters (p<0.05).

Conclusion:

As endothelial cell numbers seem to decrease with the progression of keratoconus, specular/confocal microscopy screening should be carried out, especially in eyes with advanced stages and corneas with TCT <400 µm.

Correlation of corneal endothelial changes with different stages of keratoconus

PURPOSE

The aim was to study the corneal endothelial count and morphology in patients with keratoconus by specular microscopy and to correlate them to the stage of keratoconus.

METHODS

Forty eyes of 29 patients with keratoconus were enrolled in this cross-sectional cohort study. Corneal endothelium was evaluated using specular microscopy, and corneal topography and thickness data were obtained from Scheimpflug-based corneal tomography. Eyes were classified into stages 1 through 4 of keratoconus according to Amsler classification, using keratometry and pachymetry readings obtained from corneal tomography.

RESULTS

Eleven eyes (27.5%) had stage 1, 17 eyes (42.5%) had stage 2, and 12 eyes (30%) had stage 3. Specular microscopy was not possible in stage 4. There was no statistically significant correlation between the stage of keratoconus and the endothelial cell density (ECD) (r = 0.018, P = 0.91), coefficient of variation (CV) (r = -0.011, P = 0.94), or percentage of hexagonality (6A) (r = -0.112, P = 0.51). When mild-to-moderate keratoconus (stages 1 and 2) was compared with severe keratoconus (stage 3), the difference was not significant regarding ECD (P = 0.1), CV (P = 0.3), or 6A (P = 0.4). However, there was a trend toward lower ECD and percentage of hexagonality, and a higher CV with advancing disease.

CONCLUSIONS

Up to stage 3, keratoconus does not significantly affect the corneal endothelium, as measured by specular microscopy. Eyes with stage 4 could not be studied by specular microscopy and may require other imaging methods such as confocal microscopy.

Polymorphism of the transferrin gene in eye diseases: keratoconus and Fuchs endothelial corneal dystrophy

Oxidative stress may play a role in the pathogenesis of keratoconus (KC) and Fuchs endothelial corneal dystrophy (FECD). Iron may promote the stress by the Fenton reaction, so its homeostasis should be strictly controlled. Transferrin is essential for iron homeostasis because it transports iron from plasma into cells. The malfunction of transferrin, which may be caused by variation in its gene (TF) variation, may contribute to oxidative stress and change KC and FECD risk. To verify this hypothesis we investigated the association between three polymorphisms of the TF gene, g.3296G>A (rs8177178), g.3481A>G (rs8177179), and c.-2G>A (rs1130459), and KC and FECD occurrence. Genotyping was performed in blood lymphocytes in 216 patients with KC, 130 patients with FECD and 228 controls by PCR-RFLP. We studied also the influence of other risk factors. The A/A genotype and the A allele of the g.3296G>A polymorphism were associated with KC occurrence, while the G allele was negatively correlated with it. We observed a decrease in KC occurrence associated with the A/G genotype of the g.3481A>G polymorphism. We did not find any association between the c.-2G>A polymorphism and KC. No association was found between all three polymorphisms and FECD occurrence.