Vertical cup/disc ratio in relation to optic disc size: its value in the assessment of the glaucoma suspect

A small disc size, a big challenge: effect of optic disc size on the correlation between peripapillary choroidal thickness, peripapillary retinal nerve fiber layer, and ganglion cell layer

PURPOSE

We aimed to analyze the effect of optic disc size on the correlation between the peripapillary choroid (PPC), peripapillary retinal nerve fiber layer (RNFL), and macular ganglion cell inner plexiform layer (MGCIPL) thicknesses in subjects with ocular hypertension (OHT) and primary open angle glaucoma (POAG).

METHODS

This study included 61 eyes with a disc area (DA) of ≤ 1.63 mm2, 92 eyes with a DA of 1.63-2.42 mm2, and 59 eyes with a DA of ≥ 2.42 mm2 in small disc, regular disc, and large disc groups, respectively. The swept-source optical coherence tomography scans of the PPC, RNFL, and MGCIPL were analyzed according to disc size.

RESULTS

The three groups did not significantly differ in RNFL or MGCIPL measurements, but the PPC measurement was statistically significantly higher in the small disc group and statistically significantly thinner in the large disc group. Most of the correlations observed between the RNFL and MGCIPL measurements and eye characteristics in the regular disc group were not detected in the small and large disc groups. While the RNFL and MGCIPL were well correlated in all disc size groups, the PPC did not correlate with the RNFL or MGCIPL in any of the groups.

CONCLUSIONS

Overall, the RNFL and MGCIPL measurements were consistent across all three disc sizes. While the PPC was thicker in small discs than in larger discs, it was not correlated with the RNFL or MGCIPL.

Calibrate the Inter-Observer Segmentation Uncertainty via Diagnosis-First Principle

Many of the tissues/lesions in the medical images may be ambiguous. Therefore, medical segmentation is typically annotated by a group of clinical experts to mitigate personal bias. A common solution to fuse different annotations is the majority vote, e.g., taking the average of multiple labels. However, such a strategy ignores the difference between the grader expertness. Inspired by the observation that medical image segmentation is usually used to assist the disease diagnosis in clinical practice, we propose the diagnosis-first principle, which is to take disease diagnosis as the criterion to calibrate the inter-observer segmentation uncertainty. Following this idea, a framework named Diagnosis-First segmentation Framework (DiFF) is proposed. Specifically, DiFF will first learn to fuse the multi-rater segmentation labels to a single ground-truth which could maximize the disease diagnosis performance. We dubbed the fused ground-truth as Diagnosis-First Ground-truth (DF-GT). Then, the Take and Give Model (T&G Model) to segment DF-GT from the raw image is proposed. With the T&G Model, DiFF can learn the segmentation with the calibrated uncertainty that facilitate the disease diagnosis. We verify the effectiveness of DiFF on three different medical segmentation tasks: optic-disc/optic-cup (OD/OC) segmentation on fundus images, thyroid nodule segmentation on ultrasound images, and skin lesion segmentation on dermoscopic images. Experimental results show that the proposed DiFF can effectively calibrate the segmentation uncertainty, and thus significantly facilitate the corresponding disease diagnosis, which outperforms previous state-of-the-art multi-rater learning methods.

Independent Effects of Blood Pressure on Intraocular Pressure and Retinal Ganglion Cell Degeneration: A Mendelian Randomization Study

Purpose

To investigate the causal effect of elevated blood pressure on primary open-angle glaucoma (POAG) and POAG endophenotypes.

Methods

Two-sample Mendelian randomization (MR) was performed to investigate the causal effect of elevated systolic blood pressure (SBP) (N = 757,601) and diastolic blood pressure (DBP) (N = 757,601) on intraocular pressure (IOP) (N = 139,555), macular retinal nerve fiber layer (mRNFL) thickness (N = 33,129), ganglion cell complex (GCC) thickness (N = 33,129), vertical cup-to-disc ratio (VCDR) (N = 111,724), and POAG liability (Ncases = 16,677, Ncontrols = 199,580). The primary analysis was conducted using the inverse-variance weighted approach. Sensitivity analyses were performed to investigate robustness to horizontal pleiotropy, winner's curse, and collider bias. Multivariable MR was performed to investigate whether any effect of blood pressure on retinal ganglion cell degeneration was mediated through increased IOP.

Results

Increased genetically predicted SBP and DBP associated with an increase in IOP (0.17 mm Hg [95% CI = 0.11 to 0.24] per 10 mm Hg higher SBP, P = 5.18 × 10-7, and 0.17 mm Hg [95% CI = 0.05 to 0.28 mm Hg] per 10 mm Hg higher DBP, P = 0.004). Increased genetically predicted SBP associated with a thinner GCC (0.04 µm [95% CI = -0.07 to -0.01 µm], P = 0.018) and a thinner mRNFL (0.04 µm [95% CI = -0.07 to -0.01 µm], P = 0.004), an effect that arises independently of IOP according to our mediation analysis. Neither SBP nor DBP associated with VCDR or POAG liability.

Conclusions

These findings support a causal effect of elevated blood pressure on retinal ganglion cell degeneration that does not require intermediary changes in IOP. Targeted blood pressure control may help preserve vision by lowering IOP and, independently, by preventing retinal ganglion cell degeneration, including in individuals with a normal IOP.

Relay learning: a physically secure framework for clinical multi-site deep learning

Big data serves as the cornerstone for constructing real-world deep learning systems across various domains. In medicine and healthcare, a single clinical site lacks sufficient data, thus necessitating the involvement of multiple sites. Unfortunately, concerns regarding data security and privacy hinder the sharing and reuse of data across sites. Existing approaches to multi-site clinical learning heavily depend on the security of the network firewall and system implementation. To address this issue, we propose Relay Learning, a secure deep-learning framework that physically isolates clinical data from external intruders while still leveraging the benefits of multi-site big data. We demonstrate the efficacy of Relay Learning in three medical tasks of different diseases and anatomical structures, including structure segmentation of retina fundus, mediastinum tumors diagnosis, and brain midline localization. We evaluate Relay Learning by comparing its performance to alternative solutions through multi-site validation and external validation. Incorporating a total of 41,038 medical images from 21 medical hosts, including 7 external hosts, with non-uniform distributions, we observe significant performance improvements with Relay Learning across all three tasks. Specifically, it achieves an average performance increase of 44.4%, 24.2%, and 36.7% for retinal fundus segmentation, mediastinum tumor diagnosis, and brain midline localization, respectively. Remarkably, Relay Learning even outperforms central learning on external test sets. In the meanwhile, Relay Learning keeps data sovereignty locally without cross-site network connections. We anticipate that Relay Learning will revolutionize clinical multi-site collaboration and reshape the landscape of healthcare in the future.

Value of disc topographic measurements for the discrimination of glaucoma suspects from early primary open angle glaucoma cases: A thin discriminative line

PURPOSE

To determine the diagnostic value of optic nerve head (ONH) topographic parameters measured by swept-source optical coherence tomography (SS-OCT) in differentiating between early glaucoma cases, glaucoma suspects (GSs), and healthy eyes.

METHODS

The files of GSs, those with an early primary open angle glaucoma (POAG) diagnosis, and healthy eyes were retrospectively screened. Demographic characteristics, retinal nerve fibre measurements, visual field examinations, and ONH topographic parameters, including rim area, disc area, vertical and horizontal cup-to-disc ratios (CDRs), and cup volume, were compared between the groups. A receiver operating characteristic (ROC) curve analysis was performed on the ONH parameters.

RESULTS

A total of 170 eyes from 85 GSs, 114 eyes from 57 patients with a diagnosis of early POAG, and 70 healthy eyes from 48 subjects were included in the study. The median age was 52.1 ± 0.9 years for the POAG group, 60.2 ± 1.1 years for the GS group, and 60.5 ± 1.6 for the controls. In the comparison of the ONH parameters between the groups, only rim area statistically significantly differed between the POAG and GS groups. However, all ONH parameters statistically significantly differed between the GS and control groups (p< 0.05). The mean rim areas of the GS, POAG, and control groups were 1.278 ± 0.055 mm2, 1.073 ± 0.065 mm2, and 1.446 ± 0.055 mm2, respectively, being statistically significantly higher in the controls and lower in the POAG group (p < 0.001). The highest area under the curve (AUC) value belonged to vertical CDR for the discrimination of the eyes with POAG from healthy eyes [0.806 (0.728-0.869, p < 0.001] and rim area for the discrimination of POAG cases and GSs [0.728 (0.650-0.797, p < 0.001].

CONCLUSION

There are significant differences in all ONH parameters between the eyes of GSs and healthy eyes. For the discrimination of early POAG and GS cases, topographic disc parameters may have value. In this study, rim area significantly differed between the GS and POAG groups, and it had the highest AUC value amongst the evaluated ONH parameters in the discrimination of early POAG cases S GSs.

Characteristics of a large, labeled dataset for the training of artificial intelligence for glaucoma screening with fundus photographs

Purpose

Significant visual impairment due to glaucoma is largely caused by the disease being detected too late.

Objective

To build a labeled data set for training artificial intelligence (AI) algorithms for glaucoma screening by fundus photography, to assess the accuracy of the graders, and to characterize the features of all eyes with referable glaucoma (RG).

Design

Cross-sectional study.

Subjects

Color fundus photographs (CFPs) of 113 893 eyes of 60 357 individuals were obtained from EyePACS, California, United States, from a population screening program for diabetic retinopathy.

Methods

Carefully selected graders (ophthalmologists and optometrists) graded the images. To qualify, they had to pass the European Optic Disc Assessment Trial optic disc assessment with ≥ 85% accuracy and 92% specificity. Of 90 candidates, 30 passed. Each image of the EyePACS set was then scored by varying random pairs of graders as "RG," "no referable glaucoma (NRG)," or "ungradable (UG)." In case of disagreement, a glaucoma specialist made the final grading. Referable glaucoma was scored if visual field damage was expected. In case of RG, graders were instructed to mark up to 10 relevant glaucomatous features.

Main Outcome Measures

Qualitative features in eyes with RG.

Results

The performance of each grader was monitored; if the sensitivity and specificity dropped below 80% and 95%, respectively (the final grade served as reference), they exited the study and their gradings were redone by other graders. In all, 20 graders qualified; their mean sensitivity and specificity (standard deviation [SD]) were 85.6% (5.7) and 96.1% (2.8), respectively. The 2 graders agreed in 92.45% of the images (Gwet's AC2, expressing the inter-rater reliability, was 0.917). Of all gradings, the sensitivity and specificity (95% confidence interval) were 86.0 (85.2-86.7)% and 96.4 (96.3-96.5)%, respectively. Of all gradable eyes (n = 111 183; 97.62%) the prevalence of RG was 4.38%. The most common features of RG were the appearance of the neuroretinal rim (NRR) inferiorly and superiorly.

Conclusions

A large data set of CFPs was put together of sufficient quality to develop AI screening solutions for glaucoma. The most common features of RG were the appearance of the NRR inferiorly and superiorly. Disc hemorrhages were a rare feature of RG.

Financial Disclosures

Proprietary or commercial disclosure may be found after the references.

Multiple Preprocessing Hybrid Level Set Model for Optic Disc Segmentation in Fundus Images

The accurate segmentation of the optic disc (OD) in fundus images is a crucial step for the analysis of many retinal diseases. However, because of problems such as vascular occlusion, parapapillary atrophy (PPA), and low contrast, accurate OD segmentation is still a challenging task. Therefore, this paper proposes a multiple preprocessing hybrid level set model (HLSM) based on area and shape for OD segmentation. The area-based term represents the difference of average pixel values between the inside and outside of a contour, while the shape-based term measures the distance between a prior shape model and the contour. The average intersection over union (IoU) of the proposed method was 0.9275, and the average four-side evaluation (FSE) was 4.6426 on a public dataset with narrow-angle fundus images. The IoU was 0.8179 and the average FSE was 3.5946 on a wide-angle fundus image dataset compiled from a hospital. The results indicate that the proposed multiple preprocessing HLSM is effective in OD segmentation.

Particle Swarm Optimization-Based Approach for Optic Disc Segmentation

Fundus segmentation is an important step in the diagnosis of ophthalmic diseases, especially glaucoma. A modified particle swarm optimization algorithm for optic disc segmentation is proposed, considering the fact that the current public fundus datasets do not have enough images and are unevenly distributed. The particle swarm optimization algorithm has been proved to be a good tool to deal with various extreme value problems, which requires little data and does not require pre-training. In this paper, the segmentation problem is converted to a set of extreme value problems. The scheme performs data preprocessing based on the features of the fundus map, reduces noise on the picture, and simplifies the search space for particles. The search space is divided into multiple sub-search spaces according to the number of subgroups, and the particles inside the subgroups search for the optimal solution in their respective sub-search spaces. The gradient values are used to calculate the fitness of particles and contours. The entire group is divided into some subgroups. Every particle flies in their exploration for the best solution. During the iteration, particles are not only influenced by local and global optimal solutions but also additionally attracted by particles between adjacent subgroups. By collaboration and information sharing, the particles are capable of obtaining accurate disc segmentation. This method has been tested with the Drishti-GS and RIM-ONE V3 dataset. Compared to several state-of-the-art methods, the proposed method substantially improves the optic disc segmentation results on the tested datasets, which demonstrates the superiority of the proposed work.

Circular functional analysis of OCT data for precise identification of structural phenotypes in the eye

Progressive optic neuropathies such as glaucoma are major causes of blindness globally. Multiple sources of subjectivity and analytical challenges are often encountered by clinicians in the process of early diagnosis and clinical management of these diseases. In glaucoma, the structural damage is often characterized by neuroretinal rim (NRR) thinning of the optic nerve head, and other clinical parameters. Baseline structural heterogeneity in the eyes can play a key role in the progression of optic neuropathies, and present challenges to clinical decision-making. We generated a dataset of Optical Coherence Tomography (OCT) based high-resolution circular measurements on NRR phenotypes, along with other clinical covariates, of 3973 healthy eyes as part of an established clinical cohort of Asian Indian participants. We introduced CIFU, a new computational pipeline for CIrcular FUnctional data modeling and analysis. We demonstrated CIFU by unsupervised circular functional clustering of the OCT NRR data, followed by meta-clustering to characterize the clusters using clinical covariates, and presented a circular visualization of the results. Upon stratification by age, we identified a healthy NRR phenotype cluster in the age group 40-49 years with predictive potential for glaucoma. Our dataset also addresses the disparity of representation of this particular population in normative OCT databases.

Differences in clinical presentation of primary open-angle glaucoma between African and European populations

Purpose

Primary open‐angle glaucoma (POAG) has been reported to occur more frequently in Africans, and to follow a more severe course compared to Europeans. We aimed to describe characteristics of POAG presentation and treatment across three ethnic groups from Africa and one from Europe.

Methods

We ascertained 151 POAG patients from South African Coloured (SAC) and 94 South African Black (SAB) ethnicity from a university hospital in South Africa. In Tanzania, 310 patients were recruited from a university hospital and a referral hospital. In the Netherlands, 241 patients of European ancestry were included. All patients were over 35 years old and had undergone an extensive ophthalmic examination. Patients were diagnosed according to the ISGEO criteria. A biogeographic ancestry analysis was performed to estimate the proportion of genetic African ancestry (GAA).

Results

The biogeographic ancestry analysis showed that the median proportion of GAA was 97.6% in Tanzanian, 100% in SAB, 34.2% in SAC and 1.5% in Dutch participants. Clinical characteristics at presentation for Tanzanians, SAB, SAC and Dutch participants, respectively: mean age: 63, 57, 66, 70 years (p < 0.001); visual acuity in the worse eye: 1.78, 1.78, 0.3, 0.3 LogMAR (p < 0.001); maximum intraocular pressure of both eyes: 36, 34, 29, 29 mmHg (p_anova < 0.001); maximum vertical cup to disc ratio (VCDR) of both eyes: 0.90, 0.90, 0.84, 0.83 (p < 0.001); mean central corneal thickness: 506, 487, 511, 528 μm (p < 0.001). Fourteen percent of Tanzanian patients presented with blindness (<3/60 Snellen) in the better eye in contrast to only 1% in the Dutch.

Conclusion

In this multi‐ethnic comparative study, Sub‐Saharan Africans present at a younger age with lower visual acuity, higher IOP, larger VCDR, than SAC and Dutch participants. This indicates the more progressive and destructive course in Sub‐Saharan Africans.

Translaminar Pressure Difference and Ocular Perfusion Pressure in Glaucomatous Eyes with Different Optic Disc Sizes

Purpose

Intracranial pressure (ICP) and ocular perfusion pressure (OPP) are both involved with the pathogenesis of glaucoma. The orbital ICP determines a retrolaminar counter pressure that is antagonistic to the intraocular pressure (IOP). The purpose of this study is to evaluate whether the translaminar pressure difference (TLPD) and the OPP varies in glaucoma patients with different optic disc sizes.

Methods

In this university hospital-based, observational, cross-sectional clinical study, all patients underwent an ophthalmic evaluation. Blood pressure, height, weight, and the results of retinal nerve fiber layer examination with optical coherence tomography examination were recorded. TLPD and OPP were calculated for each patient using proxy algorithms to attain indirect surrogate parameter values. Patients' eyes were stratified into three quantiles according to optic disc sizes and the differences compared. Data from both eyes were used after using the appropriate correction for inter-eye dependency.

Results

The sample consisted of 140 eyes of 73 patients with primary open-angle glaucoma and suspects. Patients with large disc size presented with higher TLPD as compared to those with average and small-sized discs (2.4 ± 4.5, 2.8 ± 3.8, and 3.7 ± 4.7 mmHg for first, second, and third tertile, respectively (P< 0.000). OPP did not vary according to the optic disc size.

Conclusion

Glaucoma patients with larger optic discs have higher TLPD. The pathological significance of this finding warrants further investigation.

The Global Extent of Undetected Glaucoma in Adults: A Systematic Review and Meta-analysis

TOPIC

Glaucoma is the leading cause of irreversible blindness, despite having good prognosis with early treatment. We evaluated the global extent of undetected glaucoma and the factors associated with it in this systematic review and meta-analysis.

CLINICAL RELEVANCE

Undetected glaucoma increases the risk of vision impairment, which leads to detrimental effects on the quality-of-life and socioeconomic well-being of those affected. Detailed information on the extent and factors associated with undetected glaucoma aid in the development of public health interventions.

METHODS

We conducted a systematic review and meta-analysis of population-based studies published between January 1, 1990, and June 1, 2020. Article search was conducted in online databases (PubMED, Web-of-Science), grey literatures (OpenGrey), and nongovernment organization reports. Our outcome measure was the proportion of glaucoma cases that were undetected previously. Manifest glaucoma included any form of glaucoma reported in the original studies and may include primary open-angle glaucoma (POAG), primary angle-closure-glaucoma, secondary glaucoma, or a combination thereof. Undetected glaucoma was defined as glaucoma cases that were undetected prior to diagnosis in the respective study. Random-effect meta-analysis was used to estimate the pooled proportion of undetected glaucoma. We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses and the Meta-analysis of Observational Studies in Epidemiology guidelines in our study.

RESULTS

We identified 61 articles from 55 population-based studies (n = 189 359 participants; n = 6949 manifest glaucoma). Globally, more than half of all glaucoma cases were undetected previously on average in each geographical region. Africa (odds ratio [OR], 12.70; 95% confidence interval [CI], 4.91-32.86) and Asia (OR, 3.41; 95% CI, 1.63-7.16) showed higher odds of undetected glaucoma as compared with Europe. Countries with low Human Development Index (HDI; <0.55) showed a higher proportion of undetected manifest glaucoma as compared with countries of medium to very high HDI (≥0.55; all P < 0.001). In 2020, 43.78 million POAG cases were projected to be undetected, of which 76.7% were in Africa and Asia.

DISCUSSION

Undetected glaucoma is highly prevalent across diverse communities worldwide and more common in Africa and Asia. Strategies to improve detection are needed to prevent excess visual disability and blindness resulting from glaucoma.

DoFE: Domain-Oriented Feature Embedding for Generalizable Fundus Image Segmentation on Unseen Datasets

Deep convolutional neural networks have significantly boosted the performance of fundus image segmentation when test datasets have the same distribution as the training datasets. However, in clinical practice, medical images often exhibit variations in appearance for various reasons, e.g., different scanner vendors and image quality. These distribution discrepancies could lead the deep networks to over-fit on the training datasets and lack generalization ability on the unseen test datasets. To alleviate this issue, we present a novel Domain-oriented Feature Embedding (DoFE) framework to improve the generalization ability of CNNs on unseen target domains by exploring the knowledge from multiple source domains. Our DoFE framework dynamically enriches the image features with additional domain prior knowledge learned from multi-source domains to make the semantic features more discriminative. Specifically, we introduce a Domain Knowledge Pool to learn and memorize the prior information extracted from multi-source domains. Then the original image features are augmented with domain-oriented aggregated features, which are induced from the knowledge pool based on the similarity between the input image and multi-source domain images. We further design a novel domain code prediction branch to infer this similarity and employ an attention-guided mechanism to dynamically combine the aggregated features with the semantic features. We comprehensively evaluate our DoFE framework on two fundus image segmentation tasks, including the optic cup and disc segmentation and vessel segmentation. Our DoFE framework generates satisfying segmentation results on unseen datasets and surpasses other domain generalization and network regularization methods.

Extraction of Retinal Layers Through Convolution Neural Network (CNN) in an OCT Image for Glaucoma Diagnosis

Glaucoma is a progressive and deteriorating optic neuropathy that leads to visual field defects. The damage occurs as glaucoma is irreversible, so early and timely diagnosis is of significant importance. The proposed system employs the convolution neural network (CNN) for automatic segmentation of the retinal layers. The inner limiting membrane (ILM) and retinal pigmented epithelium (RPE) are used to calculate cup-to-disc ratio (CDR) for glaucoma diagnosis. The proposed system uses structure tensors to extract candidate layer pixels, and a patch across each candidate layer pixel is extracted, which is classified using CNN. The proposed framework is based upon VGG-16 architecture for feature extraction and classification of retinal layer pixels. The output feature map is merged into SoftMax layer for classification and produces probability map for central pixel of each patch and decides whether it is ILM, RPE, or background pixels. Graph search theory refines the extracted layers by interpolating the missing points, and these extracted ILM and RPE are finally used to compute CDR value and diagnose glaucoma. The proposed system is validated using a local dataset of optical coherence tomography images from 196 patients, including normal and glaucoma subjects. The dataset contains manually annotated ILM and RPE layers; manually extracted patches for ILM, RPE, and background pixels; CDR values; and eventually final finding related to glaucoma. The proposed system is able to extract ILM and RPE with a small absolute mean error of 6.03 and 5.56, respectively, and it finds CDR value within average range of ± 0.09 as compared with glaucoma expert. The proposed system achieves average sensitivity, specificity, and accuracies of 94.6, 94.07, and 94.68, respectively.

Insights into the regulatory molecules involved in glaucoma pathogenesis

Glaucoma is an important cause of irreversible blindness, characterized by optic nerve anomalies. Increased intraocular pressure (IOP) and aging are major risk factors. Retinal ganglion cells and trabecular meshwork cells are certainly involved in the etiology of glaucoma. Glaucoma is usually a complex disease, and various genes and functions may contribute to its etiology. Among these may be genes that encode regulatory molecules. In this review, regulatory molecules including 18 transcription factors (TFs), 195 microRNAs (miRNAs), 106 long noncoding RNAs (lncRNAs), and two circular RNAs (circRNAs) that are reasonable candidates for having roles in glaucoma pathogenesis are described. The targets of the regulators are reported. Glaucoma-related features including apoptosis, stress responses, immune functions, ECM properties, IOP, and eye development are affected by the targeted genes. The targeted genes that are frequently targeted by multiple regulators most often affect apoptosis and the related features of cell death and cell survival. BCL2, CDKN1A, and TP53 are among the frequent targets of three types of glaucoma-relevant regulators, TFs, miRNAs, and lncRNAs. TP53 was itself identified as a glaucoma-relevant TF. Several of the glaucoma-relevant TFs are themselves among frequent targets of regulatory molecules, which is consistent with existence of a complex network involved in glaucoma pathogenesis.

Asian-specific vertical cup-to-disc ratio cut-off for glaucoma screening: An evidence-based recommendation from a multi-ethnic Asian population

IMPORTANCE

Evidence-based guidelines are essential for glaucoma screening to work effectively.

BACKGROUND

To derive a vertical cup-to-disc ratio (VCDR) cut-off for glaucoma screening in a multi-ethnic Asian population.

DESIGN

The Singapore Epidemiology of Eye Diseases (SEED) study is a population-based study conducted from 2004 to 2011 in a single tertiary care research institute.

PARTICIPANTS

SEED comprised of 10 033 Chinese, Malay and Indian adults aged ≥40 (response rate 75.6%). After excluding participants with a history of glaucoma medication or surgery, 9673 participants were included for analysis.

METHODS

A systematic eye examination, which included applanation tonometry, visual field testing, gonioscopy and dilated fundus examination was conducted.

MAIN OUTCOME MEASURE

Diagnosis of glaucoma.

RESULTS

The distribution of VCDR and VCDR asymmetry were relatively homogenous in this multi-ethnic Asian population, with a 97.5th percentile value of 0.67 and 0.17, respectively. In the absence of more definite signs of glaucoma, VCDR ≥0.60 and VCDR asymmetry ≥0.20 provided the best balance between sensitivity (95.1%) and specificity (90.9%) in detecting glaucoma. For larger optic disc (≥2.0 mm), VCDR ≥0.65 with VCDR asymmetry ≥0.20 provided the best balance between sensitivity (84.8%) and specificity (93.2%).

CONCLUSION AND RELEVANCE

Overall, VCDR ≥0.60 with VCDR ≥0.20 asymmetry provides a good balance between sensitivity and specificity in detecting glaucoma. For larger optic disc, VCDR ≥0.65 should be considered instead to mitigate against false-referrals due to larger physiological disc cupping. Our findings may act as a reference to populations with similar VCDR distribution.

The Distribution of Vertical Cup-to-Disc Ratio and its Determinants in the Iranian Adult Population

Purpose:

To determine the distribution of vertical cup-to-disc ratio (VCDR) and its relationship with ocular biometric indices.

Methods:

This study was conducted in 4737 individuals aged 45-69 years living in Shahroud who participated in the second phase of Shahroud Eye Cohort Study in 2014. All participants underwent eye examinations including the measurement of visual acuity and refraction, slit-lamp biomicroscopy, retinal examination, and fundoscopy. Normality index was used to describe data distribution, and a multiple beta regression, with adjustment for the effect of cluster sampling, was applied to explore the relationship between VCDR and the study variables.

Results:

The mean [95% confidence interval (CI)] VCDR was 0.297 (0.293-0.301) in all participants; 0.296 (0.291-0.302) in men and 0.297 (0.292-0.302) in women. The highest mean VCDR was seen in the age group 55-59 years (0.299, 95% CI: 0.292-0.307). The 97.5^th percentile was 0.600. According to multiple beta regression analysis, VCDR had a positive association with the female sex (P = 0.028), spherical equivalent (P < 0.001), cigarette smoking (P = 0.020), and axial length (P < 0.001), and had a negative association with hypertension (P = 0.001), best corrected visual acuity (P < 0.001), hyperlipidemia (P = 0.029) and anterior chamber depth (P = 0.001).

Conclusions:

The mean VCDR and the 97.5^th percentile were lower than most other studies. Although ethnicity and race may play a role in this difference, this difference should be considered in clinical decisions in the current population.

A Retrospective Comparison of Deep Learning to Manual Annotations for Optic Disc and Optic Cup Segmentation in Fundus Photographs

Purpose

Optic disc (OD) and optic cup (OC) segmentation are fundamental for fundus image analysis. Manual annotation is time consuming, expensive, and highly subjective, whereas an automated system is invaluable to the medical community. The aim of this study is to develop a deep learning system to segment OD and OC in fundus photographs, and evaluate how the algorithm compares against manual annotations.

Methods

A total of 1200 fundus photographs with 120 glaucoma cases were collected. The OD and OC annotations were labeled by seven licensed ophthalmologists, and glaucoma diagnoses were based on comprehensive evaluations of the subject medical records. A deep learning system for OD and OC segmentation was developed. The performances of segmentation and glaucoma discriminating based on the cup-to-disc ratio (CDR) of automated model were compared against the manual annotations.

Results

The algorithm achieved an OD dice of 0.938 (95% confidence interval [CI] = 0.934–0.941), OC dice of 0.801 (95% CI = 0.793–0.809), and CDR mean absolute error (MAE) of 0.077 (95% CI = 0.073 mean absolute error (MAE)0.082). For glaucoma discriminating based on CDR calculations, the algorithm obtained an area under receiver operator characteristic curve (AUC) of 0.948 (95% CI = 0.920 mean absolute error (MAE)0.973), with a sensitivity of 0.850 (95% CI = 0.794–0.923) and specificity of 0.853 (95% CI = 0.798–0.918).

Conclusions

We demonstrated the potential of the deep learning system to assist ophthalmologists in analyzing OD and OC segmentation and discriminating glaucoma from nonglaucoma subjects based on CDR calculations.

Translational Relevance

We investigate the segmentation of OD and OC by deep learning system compared against the manual annotations.

Deep learning assisted detection of glaucomatous optic neuropathy and potential designs for a generalizable model

Purpose

To evaluate ways to improve the generalizability of a deep learning algorithm for identifying glaucomatous optic neuropathy (GON) using a limited number of fundus photographs, as well as the key features being used for classification.

Methods

A total of 944 fundus images from Taipei Veterans General Hospital (TVGH) were retrospectively collected. Clinical and demographic characteristics, including structural and functional measurements of the images with GON, were recorded. Transfer learning based on VGGNet was used to construct a convolutional neural network (CNN) to identify GON. To avoid missing cases with advanced GON, an ensemble model was adopted in which a support vector machine classifier would make final classification based on cup-to-disc ratio if the CNN classifier had low-confidence score. The CNN classifier was first established using TVGH dataset, and then fine-tuned by combining the training images of TVGH and Drishti-GS datasets. Class activation map (CAM) was used to identify key features used for CNN classification. Performance of each classifier was determined through area under receiver operating characteristic curve (AUC) and compared with the ensemble model by diagnostic accuracy.

Results

In 187 TVGH test images, the accuracy, sensitivity, and specificity of the CNN classifier were 95.0%, 95.7%, and 94.2%, respectively, and the AUC was 0.992 compared to the 92.8% accuracy rate of the ensemble model. For the Drishti-GS test images, the accuracy of the CNN, the fine-tuned CNN and ensemble model was 33.3%, 80.3%, and 80.3%, respectively. The CNN classifier did not misclassify images with moderate to severe diseases. Class-discriminative regions revealed by CAM co-localized with known characteristics of GON.

Conclusions

The ensemble model or a fine-tuned CNN classifier may be potential designs to build a generalizable deep learning model for glaucoma detection when large image databases are not available.

Automatic optic nerve head localization and cup-to-disc ratio detection using state-of-the-art deep-learning architectures

Computer vision has greatly advanced recently. Since AlexNet was first introduced, many modified deep learning architectures have been developed and they are still evolving. However, there are few studies comparing these architectures in the field of ophthalmology. This study compared the performance of various state-of-the-art deep-learning architectures for detecting the optic nerve head and vertical cup-to-disc ratio in fundus images. Three different architectures were compared: YOLO V3, ResNet, and DenseNet. We compared various aspects of performance, which were not confined to the accuracy of detection but included, as well, the processing time, diagnostic performance, effect of the graphic processing unit (GPU), and image resolution. In general, as the input image resolution increased, the classification accuracy, localization error, and diagnostic performance all improved, but the optimal architecture differed depending on the resolution. The processing time was significantly accelerated with GPU assistance; even at the high resolution of 832 × 832, it was approximately 170 ms, which was at least 26 times slower without GPU. The choice of architecture may depend on the researcher's purpose when balancing between speed and accuracy. This study provides a guideline to determine deep learning architecture, optimal image resolution, and the appropriate hardware.

Identification of genes involved in glaucoma pathogenesis using combined network analysis and empirical studies

Glaucoma is a leading cause of blindness. We aimed in this study to identify genes that may make subtle and cumulative contributions to glaucoma pathogenesis. To this end, we identified molecular interactions and pathways that include transcription factors (TFs) FOXC1, PITX2, PAX6 and NFKB1 and various microRNAs including miR-204 known to have relevance to trabecular meshwork (TM) functions and/or glaucoma. TM tissue is involved in glaucoma pathogenesis. In-house microarray transcriptome results and data sources were used to identify target genes of the regulatory molecules. Bioinformatics analyses were done to filter TM and glaucoma relevant genes. These were submitted to network-creating softwares to define interactions, pathways and a network that would include the genes. The network was stringently scrutinized and minimized, then expanded by addition of microarray data and data on TF and microRNA-binding sites. Selected features of the network were confirmed by empirical studies such as dual luciferase assays, real-time PCR and western blot experiments and apoptosis assays. MYOC, WDR36, LTPBP2, RHOA, CYP1B1, OPA1, SPARC, MEIS2, PLEKHG5, RGS5, BBS5, ALDH1A1, NOMO2, CXCL6, FMNL2, ADAMTS5, CLOCK and DKK1 were among the genes included in the final network. Pathways identified included those that affect ECM properties, IOP, ciliary body functions, retinal ganglion cell viability, apoptosis, focal adhesion and oxidative stress response. The identification of many genes potentially involved in glaucoma pathology is consistent with its being a complex disease. The inclusion of several known glaucoma-related genes validates the approach used.

Normal tension glaucoma-like degeneration of the visual system in aged marmosets

The common marmoset (Callithrix jacchus) is a non-human primate that provides valuable models for neuroscience and aging research due to its anatomical similarities to humans and relatively short lifespan. This study was carried out to examine whether aged marmosets develop glaucoma, as seen in humans. We found that 11% of the aged marmosets presented with glaucoma-like characteristics; this incident rate is very similar to that in humans. Magnetic resonance imaging showed a significant volume loss in the visual cortex, and histological analyses confirmed the degeneration of the lateral geniculate nuclei and visual cortex in the affected marmosets. These marmosets did not have elevated intraocular pressure, but showed an increased oxidative stress level, low cerebrospinal fluid (CSF) pressure, and low brain-derived neurotrophic factor (BDNF) and TrkB expression in the retina, optic nerve head and CSF. Our findings suggest that marmosets have potential to provide useful information for the research of eye and the visual system.

Association of LOXL1 Gene Polymorphisms with Exfoliation Glaucoma Patients

BACKGROUND

Pseudoexfoliation syndrome (XFS) is pathogenetically related to exfoliative glaucoma (XFG), which is the most common type of secondary glaucoma. We aimed to investigate the relationship between LOXL1 SNPs (rs1048661, rs3825942) and XFS and/or XFG in a cohort of Iranian subjects.

METHODS

This cross-sectional study investigated possible association between LOXL1 gene polymorphisms and exfoliative glaucoma in Northeastern part of Iran between May 2014 and May 2015. Sixty unrelated XFS/XFG patients, as well as 40 control subjects, were studied by direct sequencing. In fifteen senile cataract patients without glaucoma and fifteen patients with coexisting XFG and cataract, capsulorhexis specimen of the anterior lens capsule was used to evaluate LOXL1 gene transcripts by Real-Time PCR technique. We analyzed the results for allele frequencies and haplotype association and investigated the relative gene expression.

RESULTS

Significant associations between the rs382594 SNP and XFG and between rs1048661 SNP and XFG were observed (P<0.05 for both). The frequency of the G allele in the exonic SNP (rs1048661) appeared to be higher in XFS or XFG patients compared to control subjects (P= 0.0497). Moreover, in the rs3825942 SNP, the G allele was more frequent in XFS/XFG patients compared to control subjects (P=0.0016). The highest cumulative frequency was for the GG haplotype. GG haplotype was associated with increased risk of XFG compared to the rs1048661 G/T and rs3825942 G/A haplotypes. LOXL1 mRNA expression was not statistically significantly different between XFS/XFG and control subjects.

CONCLUSION

We reported the LOXL1 gene polymorphism in an Iranian XFS/XFG cohort. Similar to many other ethnic groups and geographic regions, our results confirmed an association between LOXL1 gene variants and XFG in Iran.

Comparison of Fundus Biomicroscopy Examination of the Optic Nerve Head with and without Mydriasis

BACKGROUND

Notwithstanding the significant advances in automated imaging techniques in the past 2 decades, subjective evaluation of the optic disc still remains an important part of glaucoma propaedeutic. In places with limited resources and a high demand for ophthalmic care, anatomical evaluation of glaucoma cases often relies solely on slit-lamp-based fundus biomicroscopic examination, which is frequently performed without mydriasis.

OBJECTIVE

The aim of this study was to compare metrics related to fundus biomicroscopy examination of the optic nerve head and peripapillary retinal nerve fiber layer (pRNFL) with and without mydriasis.

MATERIAL AND METHODS

Healthy individuals, patients with early glaucoma, and glaucoma suspects were prospectively enrolled. Patients were examined before and after mydriasis by three glaucoma specialists, who estimated patients' vertical cup-to-disc ratio (CDR) and evaluated the presence of glaucomatous signs: laminar dot sign, disc hemorrhage, disc saucering, disc notching, peripapillary atrophy, localized pRNFL defect, and loss of the ISNT pattern. Main outcome measures were the intra-observer comparison, the inter-observer agreement, and the abilities to identify glaucomatous signs before and after mydriasis.

RESULTS

Thirty patients (60 eyes) were enrolled (mean age, 62.3 ± 11.6 years). Considering the evaluation of the three examiners, the mean vertical CDR increased from 0.41 to 0.44 (p = 0.02), and the median of the coefficient of variation of the measures was reduced from 0.24 to 0.11 (p = 0.01) after mydriasis. Regarding the inter-observer agreement evaluation, the kappa coefficient values ranged from 0.64 to 0.72 before mydriasis and from 0.71 to 0.77 after mydriasis. Dot sign and disc notching were better identified through fundoscopic examination with mydriasis compared to the nonmydriatic examination (p < 0.01).

CONCLUSION

Our results suggest that fundus biomicroscopy should be performed with mydriasis whenever possible, as it yells a better intra- and inter-observer agreement and improves the detection of glaucomatous signs. Moreover, examiners seem to underestimate CDR values without mydriasis. Further investigation is warranted to validate these findings by general ophthalmologists and in different populations.

Border Tissue Morphology Is Spatially Associated with Focal Lamina Cribrosa Defect and Deep-Layer Microvasculature Dropout in Open-Angle Glaucoma

PURPOSE

To investigate the topographic relationship among focal lamina cribrosa (LC) defect, microvasculature dropout (MvD) and border tissue morphology in open angle glaucoma (OAG) eyes using spectral-domain (SD) optical coherence tomography (OCT) and OCT angiography.

DESIGN

Cross-sectional study.

METHODS

One hundred twenty-six OAG eyes and 97 normal eyes were included. The maximum externally oblique border tissue (EOBT) length was measured by using enhanced depth imaging SD-OCT as well as focal LC defect size. Circumferential MvD width and height ratio were measured using OCT angiography.

RESULTS

Significant correlations were found among the locations of focal LC defect, MvD and maximum EOBT length. The mean absolute locational difference was 29.1° (95% CI, -47.6 to 105.7) between focal LC defect and MvD, 10.0° (95% CI, -79.4 to 99.4) between focal LC defect and maximum EOBT length, and 10.6° (95% CI, -71.1 to 92.3) between MvD and maximum EOBT length. In multivariate logistic regression analysis, a worse VF defect was significantly associated with the presence of focal LC defects and MvDs (P < .002; P = .002, respectively). MvD circumferential width was associated with glaucoma severity (R = -0.66, P < .001), whereas focal LC defect size and MvD height ratio were associated with maximum EOBT length (R = 0.48, P < .001; R = 0.65, P < .001, respectively) and AL (R = 0.53, P < .001; R = 0.52, P < .001, respectively).

CONCLUSIONS

There was a topographical correlation among the locations of focal LC defect, MvD and maximum border length. In addition, the presence of focal LC defect and MvD were also strongly associated with glaucoma severity. Thus, it is thought that focal LC defect and MvD may be biomarkers that reflect glaucoma severity especially at the location of maximum border tissue elongation.

Robust optic disc and cup segmentation with deep learning for glaucoma detection

Glaucoma is rated as the leading cause of irreversible vision loss worldwide. Early detection of glaucoma is important for providing timely treatment and minimizing the vision loss. In this paper, we developed a robust segmentation method for optic disc and cup segmentation using a modified U-Net architecture, which combines the widely adopted pre-trained ResNet-34 model as encoding layers with classical U-Net decoding layers. The model was trained on the newly available RIGA dataset, and achieved an average dice value of 97.31% for disc segmentation and 87.61% for cup segmentation, comparable to that of the experts' performance for optic disc/cup segmentation and Cup-Disc-Ratio (CDR) calculation on a reserved RIGA dataset. When tested on DRISHTI-GS and RIM-ONE dataset without re-training or fine-tuning, the model achieved comparable performance to that of the state-of-the-art in literature. We have also fine-tuned the model on two databases, which achieves an average disc dice value of 97.38% and cup dice value of 88.77% for DRISHTI-GS test set, disc dice of 96.10% and cup dice of 84.45% for RIM-ONE database, which is the state-of-the-art performance on both databases in terms of cup dice and disc dice value. The advantage of the proposed method is the combination of the pre-trained ResNet and U-Net, which avoids training the network from scratch, thereby enabling fast network training with less epochs, thus further avoids over-fitting and achieves robust performance.

The impact of SIGN glaucoma guidelines on false-positive referrals from community optometrists in Central Scotland

Background

Since the introduction of National Institute for Health and Care Excellence glaucoma guidelines 2009, the number of referrals from community optometrists to hospital eye services has increased across the UK, resulting in increase in first visit discharge rates (FVDRs).

Aim

To assess the impact of Scottish Intercollegiate Guidelines Network (SIGN) 144 on quality of referrals from community optometrists.

Methodology

A retrospective study of patient records who attended as new adult glaucoma referrals to clinics in Princess Alexandra Eye Pavilion, Edinburgh, and in Greater Glasgow and Clyde, was carried out across October–November 2014 (group 1) and September–October 2016 (group 2), before and after the introduction of SIGN 144. The primary outcome of this study is FVDRs. A secondary outcome is the extent of compliance to referral recommendations by SIGN guidelines.

Results

Three hundred and twelve and 325 patients were included in groups 1 and 2, respectively. There was a significant decline in FVDRs between these two periods from 29.2% to 19.2%. (p=0.004) (OR 0.58 (95%CI 0.40 to 0.84)). Post-SIGN guidelines, 87% of referrals were compliant to SIGN referral criteria while 13% remained non-compliant. The main reasons for non-compliance were no repeatable visual field defects (42.0%) and referrals due to high intraocular pressure were either not repeated or not interpreted in the context of age and central corneal thickness (36.8%).

Conclusion

Patients referred after the introduction of SIGN guidelines were 33.5% less likely to be discharged at the first visit. Although compliance to most recommendations in SIGN guidelines has improved, there is still a need to improve adherence to referral criteria

Application of Bayes' to the prediction of referral decisions made by specialist optometrists in relation to chronic open angle glaucoma

PURPOSE

To determine the accuracy of a Bayesian learning scheme (Bayes') applied to the prediction of clinical decisions made by specialist optometrists in relation to the referral refinement of chronic open angle glaucoma.

METHODS

This cross-sectional observational study involved collection of data from the worst affected or right eyes of a consecutive sample of cases (n = 1,006) referred into the West Kent Clinical Commissioning Group Community Ophthalmology Team (COT) by high street optometrists. Multilevel classification of each case was based on race, sex, age, family history of chronic open angle glaucoma, reason for referral, Goldmann Applanation Tonometry (intraocular pressure and interocular asymmetry), optic nerve head assessment (vertical size, cup disc ratio and interocular asymmetry), central corneal thickness and visual field analysis (Hodapp-Parrish-Anderson classification). Randomised stratified tenfold cross-validation was applied to determine the accuracy of Bayes' by comparing its output to the clinical decisions of three COT specialist optometrists; namely, the decision to discharge, follow-up or refer each case.

RESULTS

Outcomes of cross-validation, expressed as means and standard deviations, showed that the accuracy of Bayes' was high (95%, 2.0%) but that it falsely discharged (3.4%, 1.6%) or referred (3.1%, 1.5%) some cases.

CONCLUSIONS

The results indicate that Bayes' has the potential to augment the decisions of specialist optometrists.

The relationship between HDAC6, CXCR3, and SIRT1 genes expression levels with progression of primary open-angle glaucoma

BACKGROUND

Primary open-angle glaucoma (POAG) belongs to neurodegenerative diseases. Its etiology is not fully understood. However, a lot of reports have indicated that many biochemical molecules are involved in the retinal ganglion cell damage. Therefore, the purpose of this study was to evaluate a relationship between HDAC6, CXCR3, and SIRT1 genes expression levels with the occurrence risk of POAG and its progression.

MATERIALS AND METHODS

The study included 34 glaucoma patients and 32 subjects without glaucoma symptoms. RNA was isolated from peripheral blood lymphocytes. Level of mRNA expression was determined by real-time PCR method.

RESULTS

Our results have shown significant association of the HDAC6 and SIRT1 expression levels with progression of POAG according to rim area (RA) value, p = 0.041; p = 0.012. Moreover, the analysis of the CXCR3 expression level showed a correlation with progression of POAG based on RA and cup disc ratio (c/d) value, p = 0.006 and p = 0.012, respectively.

CONCLUSIONS

The expression level of HDAC6, CXCR3, and SIRT1 genes may be involved in the progression of POAG.

Relationship between corneal biomechanical properties and structural biomarkers in patients with normal-tension glaucoma: a retrospective study

Background

We evaluated the relationships between corneal biomechanical properties and structural parameters in patients with newly diagnosed, untreated normal-tension glaucoma (NTG).

Methods

All subjects were evaluated using an Ocular Response Analyzer (ORA) measuring corneal hysteresis (CH) and the corneal resistance factor (CRF). Central corneal thickness (CCT), Goldmann applanation tonometric (GAT) data, axial length, and the spherical equivalent (SE), were also measured. Confocal scanning laser ophthalmoscopy was performed with the aid of a Heidelberg retina tomograph (HRT III). We sought correlations between HRT parameters and different variables including CCT, CH, and the CRF. Multiple linear regression analysis was performed to identify significant associations between corneal biomechanical properties and optic nerve head parameters.

Results

We enrolled 95 eyes of 95 NTG patients and 93 eyes of 93 normal subjects. CH and the CRF were significantly lower in more advanced glaucomatous eyes (P = 0.001, P = 0.008, respectively). The rim area, rim volume, linear cup-to-disc ratio (LCDR), and mean retinal nerve fiber layer (RNFL) thickness were significantly worse in more advanced glaucomatous eyes (P < 0.001, P < 0.001, P < 0.001, and P = 0.001). CH was directly associated with rim area, rim volume, and mean RNFL thickness (P = 0.012, P = 0.028, and P = 0.043) and inversely associated with LCDR (P = 0.015), after adjusting for age, axial length, CCT, disc area, GAT data, and SE. However, in normal subjects, there were no significant associations between corneal biomechanical properties and HRT parameters.

Conclusions

A lower CH is significantly associated with a smaller rim area and volume, a thinner RNFL, and a larger LCDR, independent of disc size, corneal thickness, intraocular pressure, and age.

The Association Between Clinical Features Seen on Fundus Photographs and Glaucomatous Damage Detected on Visual Fields and Optical Coherence Tomography Scans

PURPOSE

To classify the appearance of the optic disc seen on fundus photographs of healthy subjects and patients with or suspected glaucoma whose diagnosis was based upon visual fields (VFs) and spectral-domain optical coherence tomography (sdOCT) results.

PATIENTS AND METHODS

One eye of 100 patients with or suspected glaucoma and 62 healthy subjects were prospectively tested with 24-2 and 10-2 VF and macular and disc sdOCT cube scans. All eyes with or suspected glaucoma had a 24-2 mean deviation better than -6.0 dB and an abnormal appearing disc on stereophotographs. The retinal ganglion cell plus inner plexiform layer (RGC+) from the macular scans and the retinal nerve fiber layer (RNFL) from the macular and disc scans were segmented and converted to probabilities plots. An eye was considered "glaucoma" if the sdOCT probability plots showed an abnormality in a region that corresponded to a defect seen on the 24-2 and/or 10-2 VF total deviation plot. Similarly, an eye was considered "suspect" only if both the sdOCT and VF plots were normal. Healthy subjects (normal VFs and sdOCT) were classified as "controls" and used as reference for comparisons. Glaucoma specialists reviewed the stereophotographs and classified eyes based on the presence of signs suggestive of glaucomatous optic neuropathy.

RESULTS

The pattern of clinical signs of glaucomatous optic neuropathy seen on stereophotographs was statistically different between glaucoma (P<0.001) and suspects (P<0.001) vs. controls and explained up to 68% of the total variance of the diagnosis based upon sdOCT and VFs. Vertical cup-to-disc>0.6, focal neuroretinal rim thinning, focal RNFL loss, and violation of the ISNT rule had the best performance to differentiate glaucoma and suspects from controls. Compared with the suspect group, glaucoma eyes (abnormal sdOCT and VF tests) were more likely to have vertical cup-to-disc>0.6 (92% vs. 69%, P=0.003), diffuse rim (53% vs. 9%, P<0.001) and RNFL (61% vs. 26%, P<0.001) thinning, and β-zone parapapillary atrophy (68% vs. 17%, P<0.001).

CONCLUSIONS

Focal and diffuse signs of glaucoma damage seen on stereophotographs often match damage shown on VFs and sdOCT. In addition, damage shown on VFs and sdOCT is often missed during clinical evaluation. Longitudinal studies ought to differentiate focal signs of glaucoma damage seen on stereophotography from false-positives or very early loss.

An Automatic Image Processing System for Glaucoma Screening

Horizontal and vertical cup to disc ratios are the most crucial parameters used clinically to detect glaucoma or monitor its progress and are manually evaluated from retinal fundus images of the optic nerve head. Due to the rarity of the glaucoma experts as well as the increasing in glaucoma's population, an automatically calculated horizontal and vertical cup to disc ratios (HCDR and VCDR, resp.) can be useful for glaucoma screening. We report on two algorithms to calculate the HCDR and VCDR. In the algorithms, level set and inpainting techniques were developed for segmenting the disc, while thresholding using Type-II fuzzy approach was developed for segmenting the cup. The results from the algorithms were verified using the manual markings of images from a dataset of glaucomatous images (retinal fundus images for glaucoma analysis (RIGA dataset)) by six ophthalmologists. The algorithm's accuracy for HCDR and VCDR combined was 74.2%. Only the accuracy of manual markings by one ophthalmologist was higher than the algorithm's accuracy. The algorithm's best agreement was with markings by ophthalmologist number 1 in 230 images (41.8%) of the total tested images.

Morphometric parameters of the optic disc in normal and glaucomatous eyes based on time-domain optical coherence tomography image analysis

BACKGROUND AND OBJECTIVE

Assessment of optic disc morphology is essential in diagnosis and management of visual impairment. The aim of this study was to evaluate associations between optic disc morphometric parameters, i.e., size and shape, and age, gender, and ocular axial length in normal and glaucomatous eyes based on time-domain optical coherence tomography image analysis.

MATERIALS AND METHODS

It was a case-control study of 998 normal and 394 eyes with primary open angle glaucoma that underwent an ophthalmological examination and time-domain optical coherence topography scanning. Areas and shapes of the disc, cup, and neuroretinal rim were analyzed.

RESULTS

The shape of the optic disc did not differ between the study groups, i.e., normal and glaucomatous case groups, but the disc area of the primary open angle glaucoma group was significantly larger. The shape of the small disc was significantly different, but the shape of the medium and the large disc did not differ between the study groups. The central area of the disc, i.e., cup area was significantly larger in the case group and its shape was significantly different between the study groups. No significant differences in the area of the cup and its shape, nerve fibers on the edge of the disc, i.e., neuroretinal rim area, were found between the study groups of the small discs. There were significant associations between age, gender, and ocular axial length and morphometric parameters of the optic disc.

CONCLUSIONS

Informative results with regard to the size and shape due to various ocular characteristics between the healthy control group and patients suffering with primary open angle glaucoma were obtained. Both study groups were significant in size, which makes the findings interesting and important contribution in the field.

Pattern noise (PANO): a new automated functional glaucoma test

PURPOSE

To present a newly developed visual field device (pattern noise: PANO) designed to be sensitive to glaucoma defects, cost-effective, material-practical and easy to repair and therefore particularly suited for low-income countries, where glaucoma can be highly prevalent (e.g. sub-Saharan Africa).

METHODS

This is primarily a descriptive paper, but it also includes a prospective matched case-control pilot study. Hardware, stimulus, target configuration, testing strategy and result sheet are described. The main outcome measure is the contrast level (range 2-64). Targets are composed of bright/dark pixels flickering with 18 Hz and have a size of 5°. Pixel size is approximated to the hill of vision. Average luminance of targets is constant and equals background luminance.The study was performed in the West Region in Cameroon. Twenty eyes of 20 newly presenting patients with glaucomatous optic disc cupping on funduscopy were compared with 20 eyes of 20 normal patients matched in age and laterality of eye.

RESULTS

Mean age was 32.9 ± 18.8 years for glaucoma patients and 32.2 ± 15.6 years for healthy subjects. Mean contrast threshold was significantly higher in eyes with abnormal disc (16.2 ± 14.3 vs. 4.4 ± 0.8, P = 0.002). Correlation of mean contrast thresholds and cup-to-disc ratio was significant (r = 0.59; P = 0.006). Average examination time was significantly longer for glaucoma eyes compared to healthy eyes (8.2 vs. 6.1 min, P < 0.001), whereas error rate did not differ (4.8 ± 2.5% vs. 4.1 ± 1.8%, P = 0.33).

CONCLUSION

PANO demonstrated visual field defects in patients with glaucomatous optic disc. Defects correlated significantly with glaucomatous optic nerve head morphological alterations. Healthy eyes obtained normal results. More studies are needed to establish PANO.

Retinal vasculature in glaucoma: a review

Despite the critical impact of glaucoma on global blindness, its aetiology is not fully characterised. Elevated intraocular pressure is highly associated with glaucomatous optic neuropathy. However, visual field loss still progresses in some patients with normal or even low intraocular pressure. Vascular factors have been suggested to play a role in glaucoma development, based on numerous studies showing associations of glaucoma with blood pressure, ocular perfusion pressure, vasospasm, cardiovascular disease and ocular blood flow. As the retinal vasculature is the only part of the human circulation that readily allows non-invasive visualisation of the microcirculation, a number of quantitative retinal vascular parameters measured from retinal photographs using computer software (eg, calibre, fractal dimension, tortuosity and branching angle) are currently being explored for any association with glaucoma and its progression. Several population-based and clinical studies have reported that changes in retinal vasculature (eg, retinal arteriolar narrowing and decreased fractal dimension) are associated with optic nerve damage and glaucoma, supporting the vascular theory of glaucoma pathogenesis. This review summarises recent findings on the relationships between quantitatively measured structural retinal vascular changes with glaucoma and other markers of optic nerve head damage, including retinal nerve fibre layer thickness. Clinical implications, recent new advances in retinal vascular imaging (eg, optical coherence tomography angiography) and future research directions are also discussed.

Quantitative Assessment of Optic Nerve Changes in Patients With Diabetic Macular Edema Treated With Fluocinolone Acetonide Vitreous Implants

BACKGROUND AND OBJECTIVE

To evaluate glaucomatous changes in patients with diabetic macular edema (DME) treated with intravitreal implants releasing 0.2 µg/day or 0.5 µg/day fluocinolone acetonide (FAc) (Iluvien 0.2 µg/day; Alimera Sciences, Alpharetta, GA) or sham control.

PATIENTS AND METHODS

Fundus photographs were assessed to determine clinically significant changes in glaucomatous indicators.

RESULTS

The mean cup-to-disc ratio (CDR) change was similar with all three treatments. Compared with sham control, a significantly greater proportion of patients treated with 0.5 µg/day but not 0.2 µg/day FAc experienced a CDR increase of greater than 0.1. There was no significant increase in the proportion of patients experiencing a CDR increase of greater than 0.2 with either dose of implant versus sham control. Other indicators of glaucomatous change did not differ significantly with treatment. Subgroup analyses showed no differences in cupping based on ocular or baseline characteristics.

CONCLUSION

Treatment with FAc for 36 months was not associated with significant glaucomatous optic nerve head changes in patients with DME with or without increased intraocular pressure. [Ophthalmic Surg Lasers Imaging Retina. 2016;47:418-425.].

Influence of Optic Disc Size on Identifying Glaucomatous Optic Neuropathy

PURPOSE

The aim of this study was to use a large group of observers to test prior research findings that suggest optic disc size, clinical evaluation of the neuroretinal rim (ISNT rule), and practitioner characteristics influence the accuracy of differentiating normal from glaucomatous optic nerves.

METHODS

Participant observers were optometrists, optometry students, and vision scientists/researchers attending the 2013 American Academy of Optometry Annual Meeting. Each observer viewed and judged six sets of stereoscopic photographs of normal and clinically confirmed glaucomatous optic nerves of different sizes presented in random order. Observers were queried on whether each nerve was glaucomatous or normal, whether the nerve followed the ISNT rule, and whether further evaluation with advanced imaging techniques was indicated.

RESULTS

Of the 261 observers who participated, 59% were practicing optometrists, 7% were vision scientists, and 34% were residents or students. Of practicing optometrists and vision scientists, half (49%) had more than 15 years of experience, whereas 11% had less than 2 years of experience. Diagnostic accuracy differed based on optic nerve size: average-sized nerves were correctly identified by 90% of subjects, whereas small nerves and large nerves were correctly identified by 42% and 62%, respectively. Notably, only 9% of subjects correctly identified the small glaucomatous nerve, and only 34% correctly identified the large normal nerve. No practitioner characteristics were associated with diagnostic accuracy.

CONCLUSIONS

Accurate identification of glaucomatous optic neuropathy was significantly influenced by optic disc size. This was particularly evident for the large normal nerve and the small glaucomatous nerve. The ISNT rule provided value for differentiating normal from glaucomatous nerves, but its subjective interpretation resulted in considerable intergrader variability. These findings agree with other studies utilizing smaller numbers of observers but larger numbers of optic nerve presentations that disc size and the ISNT rule have value for enhancing accuracy of optic nerve assessment.

Neuroprotection for treatment of glaucoma in adults

BACKGROUND

Glaucoma is a heterogeneous group of conditions involving progressive damage to the optic nerve, deterioration of retinal ganglion cells, and ultimately visual field loss. It is a leading cause of blindness worldwide. Open angle glaucoma (OAG), the most common form of glaucoma, is a chronic condition that may or may not present with increased intraocular pressure (IOP). Neuroprotection for glaucoma refers to any intervention intended to prevent optic nerve damage or cell death.

OBJECTIVES

The objective of this review was to systematically examine the evidence regarding the effectiveness of neuroprotective agents for slowing the progression of OAG in adults compared with no neuroprotective agent, placebo, or other glaucoma treatment.

SEARCH METHODS

We searched CENTRAL (which contains the Cochrane Eyes and Vision Trials Register) (2016, Issue 7), Ovid MEDLINE, Epub Ahead of Print, In-Process & Other Non-Indexed Citations, Ovid MEDLINE Daily (January 1946 to August 2016), Embase (January 1980 to August 2016), Latin American and Caribbean Literature on Health Sciences (LILACS) (January 1982 to August 2016), the ISRCTN registry (www.isrctn.com/editAdvancedSearch), ClinicalTrials.gov (www.clinicaltrials.gov), and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp/search/en). We did not use any date or language restrictions in the electronic searches for trials. We last searched the electronic databases on 16 August 2016.

SELECTION CRITERIA

We included randomised controlled trials (RCTs) in which topical or oral treatments were used for neuroprotection in adults with OAG. Minimum follow-up time was four years.

DATA COLLECTION AND ANALYSIS

Two review authors independently reviewed titles and abstracts from the literature searches. We obtained full-text copies of potentially relevant studies and re-evaluated for inclusion. Two review authors independently extracted data related to study characteristics, risk of bias, and outcomes. We identified one trial for this review, thus we performed no meta-analysis. Two studies comparing memantine to placebo are currently awaiting classification until study investigators provide additional study details. We documented reasons for excluding studies from the review.

MAIN RESULTS

We included one multicenter RCT of adults with low-pressure glaucoma (Low-pressure Glaucoma Treatment Study, LoGTS) conducted in the USA. The primary outcome was progression of visual field loss after four years of treatment with either brimonidine or timolol. Of the 190 adults enrolled in the study, the investigators excluded 12 (6.3%) after randomization; 77 participants (40.5%) did not complete four years of follow-up. The rate of attrition was unbalanced between groups with more participants dropping out of the brimonidine group (55%) than the timolol group (29%).Of those remaining in the study at four years, participants assigned to brimonidine showed less progression of visual field loss than participants assigned to timolol (risk ratio (RR) 0.35, 95% confidence interval (CI) 0.14 to 0.86; 101 participants). Because of high risk of attrition bias and potential selective outcome reporting, we graded the certainty of evidence for this outcome as very low. At the four-year follow-up, the mean IOP was similar in both groups among those for whom data were available (mean difference 0.20 mmHg, 95% CI -0.73 to 1.13; 91 participants; very low-certainty evidence). The study authors did not report analyzable data for visual acuity or any data related to vertical cup-disc ratio, quality of life, or economic outcomes. The most frequent adverse event was ocular allergy to the study drug, which affected more participants in the brimonidine group than the timolol group (RR 5.32, 95% CI 1.64 to 17.26; 178 participants; very low-certainty evidence).

AUTHORS' CONCLUSIONS

Although the only trial we included in this review found less visual field loss in the brimonidine-treated group, the evidence was of such low certainty that we can draw no conclusions from this finding. Further clinical research is needed to determine whether neuroprotective agents may be beneficial for individuals with OAG. Such research should focus on outcomes important to patients, such as preservation of vision, and how these outcomes relate to cell death and optic nerve damage. As OAG is a chronic, progressive disease with variability in symptoms, RCTs designed to measure the effectiveness of neuroprotective agents require a long-term follow-up of five years or longer to detect clinically meaningful effects.

The Evonik-Mainz Eye Care-Study (EMECS): Development of an Expert System for Glaucoma Risk Detection in a Working Population

PURPOSE

To develop an expert system for glaucoma screening in a working population based on a human expert procedure using images of optic nerve head (ONH), visual field (frequency doubling technology, FDT) and intraocular pressure (IOP).

METHODS

4167 of 13037 (32%) employees between 40 and 65 years of Evonik Industries were screened. An experienced glaucoma expert (JW) assessed papilla parameters and evaluated all individual screening results. His classification into "no glaucoma", "possible glaucoma" and "probable glaucoma" was defined as "gold standard". A screening model was developed which was tested versus the gold-standard. This model took into account the assessment of the ONH. Values and relationships of CDR and IOP and the FDT were considered additionally and a glaucoma score was generated. The structure of the screening model was specified a priori whereas values of the parameters were chosen post-hoc to optimize sensitivity and specificity of the algorithm. Simple screening models based on IOP and / or FDT were investigated for comparison.

RESULTS

111 persons (2.66%) were classified as glaucoma suspects, thereof 13 (0.31%) as probable and 98 (2.35%) as possible glaucoma suspects by the expert. Re-evaluation by the screening model revealed a sensitivity of 83.8% and a specificity of 99.6% for all glaucoma suspects. The positive predictive value of the model was 80.2%, the negative predictive value 99.6%. Simple screening models showed insufficient diagnostic accuracy.

CONCLUSION

Adjustment of ONH and symmetry parameters with respect to excavation and IOP in an expert system produced sufficiently satisfying diagnostic accuracy. This screening model seems to be applicable in such a working population with relatively low age and low glaucoma prevalence. Different experts should validate the model in different populations.

Glaucoma Structural and Functional Progression in American and Korean Cohorts

PURPOSE

To compare the rate of glaucoma structural and functional progression in American and Korean cohorts.

DESIGN

Retrospective longitudinal study.

PARTICIPANTS

Three hundred thirteen eyes from 189 glaucoma and glaucoma suspects, followed up for an average of 38 months.

METHODS

All subjects were examined semiannually with visual field (VF) testing and spectral-domain optical coherence tomography. All subjects had 5 or more reliable visits.

MAIN OUTCOME MEASUREMENTS

The rates of change of retinal nerve fiber layer (RNFL) thickness, cup-to-disc (C/D) ratios, and VF mean deviation (MD) were compared between the cohorts. Variables affecting the rate of change for each parameter were determined, including ethnicity, refraction, baseline age and disease severity, disease subtype (high- vs. normal-tension glaucoma), clinical diagnosis (glaucoma vs. glaucoma suspect), and the interactions between variables.

RESULTS

The Korean cohort predominantly demonstrated normal-tension glaucoma, whereas the American cohort predominantly demonstrated high-tension glaucoma. Cohorts had similar VF parameters at baseline, but the Korean eyes had significantly thicker mean RNFL and larger cups. Korean glaucoma eyes showed a faster thinning of mean RNFL (mean, -0.71 μm/year vs. -0.24 μm/year; P < 0.01). There were no detectable differences in the rate of change between the glaucoma cohorts for C/D ratios and VF MD and for all parameters in glaucoma suspect eyes. Different combinations of the tested variables significantly impacted the rate of change.

CONCLUSIONS

Ethnicity, baseline disease severity, disease subtype, and clinical diagnosis should be considered when comparing glaucoma progression studies.

A common variant near TGFBR3 is associated with primary open angle glaucoma

Primary open angle glaucoma (POAG), a major cause of blindness worldwide, is a complex disease with a significant genetic contribution. We performed Exome Array (Illumina) analysis on 3504 POAG cases and 9746 controls with replication of the most significant findings in 9173 POAG cases and 26 780 controls across 18 collections of Asian, African and European descent. Apart from confirming strong evidence of association at CDKN2B-AS1 (rs2157719 [G], odds ratio [OR] = 0.71, P = 2.81 × 10(-33)), we observed one SNP showing significant association to POAG (CDC7-TGFBR3 rs1192415, ORG-allele = 1.13, Pmeta = 1.60 × 10(-8)). This particular SNP has previously been shown to be strongly associated with optic disc area and vertical cup-to-disc ratio, which are regarded as glaucoma-related quantitative traits. Our study now extends this by directly implicating it in POAG disease pathogenesis.

Megalopapilla in children: a spectral domain optical coherence tomography analysis

PURPOSE

To compare various optic nerve head (ONH) parameters and peripapillary retinal nerve fibre layer (pRNFL) thickness between megalopapilla cases and normal control using spectral domain optical coherence tomography (SD-OCT) in children.

METHODS

Fifty eyes with megalopapilla and 80 normal control eyes of totally 130 children between the ages of 6 and 15 were examined using SD-OCT. Optic nerve head parameters including disc diameter, cup-to-disc ratio, disc area, cup area, rim area, cup volume and rim volume, and pRNFL thickness were analysed and compared between two groups.

RESULTS

Megalopapilla group presented larger optic disc diameter, disc and cup area, cup-to-disc and cup-to-disc area ratio, when compared with normal control group. However, there is no difference in rim area (p = 0.25) and rim volume (p = 0.48) between the two groups. Average pRNFL thickness was higher in the megalopapilla group (p < 0.001).

CONCLUSIONS

Rim area, rim volume and pRNFL, which reflect the number of retinal nerve fibres, were preserved in children with megalopapilla. Various ONH parameters obtained with SD-OCT may be useful in the differential diagnosis of megalopapilla in children.

Corneal biomechanical properties and glaucoma-related quantitative traits in the EPIC-Norfolk Eye Study

PURPOSE

We examined the association of corneal hysteresis (CH) with Heidelberg retina tomograph (HRT)- and Glaucoma Detection with Variable Corneal Compensation scanning laser polarimeter (GDxVCC)-derived measures in a British population.

METHODS

The EPIC-Norfolk Eye Study is nested within a multicenter cohort study--the European Prospective Investigation of Cancer. Ocular response analyzer (ORA), HRT3, and GDxVCC measurements were taken at the research clinic. Three ORA measurements were taken per eye, and the single best value used. Participants meeting predefined criteria were referred for a second examination, including Goldmann applanation tonometry (GAT) and central corneal thickness (CCT) measurement. Generalized estimating equation models were used to examine the associations of CH with HRT and GDxVCC parameters, adjusted for disc area. The GDxVCC analyses were adjusted further for typical scan score to handle atypical retardation.

RESULTS

There were complete research clinic data from 5134 participants. Corneal hysteresis was associated positively with HRT rim area (P < 0.001), and GDxVCC retinal nerve fiber layer (RNFL) average thickness (P = 0.006) and modulation (P = 0.003), and associated negatively with HRT linear cup-to-disc ratio (LCDR, P < 0.001), after adjustment for Goldmann-correlated IOP and other possible confounders. In the 602 participants undergoing the second examination, CH was associated negatively with LCDR (P = 0.008) after adjustment for GAT, CCT, and other possible confounders.

CONCLUSIONS

Lower CH was associated with HRT and GDxVCC parameters in a direction that is seen in glaucoma and with ageing. Further research is required to establish if this is a causal relationship, or due to residual confounding by age, IOP, or CCT.