Convolutional neural network-based classification of glaucoma using optic radiation tissue properties

BACKGROUND

Sensory changes due to aging or disease can impact brain tissue. This study aims to investigate the link between glaucoma, a leading cause of blindness, and alterations in brain connections.

METHODS

We analyzed diffusion MRI measurements of white matter tissue in a large group, consisting of 905 glaucoma patients (aged 49-80) and 5292 healthy individuals (aged 45-80) from the UK Biobank. Confounds due to group differences were mitigated by matching a sub-sample of controls to glaucoma subjects. We compared classification of glaucoma using convolutional neural networks (CNNs) focusing on the optic radiations, which are the primary visual connection to the cortex, against those analyzing non-visual brain connections. As a control, we evaluated the performance of regularized linear regression models.

RESULTS

We showed that CNNs using information from the optic radiations exhibited higher accuracy in classifying subjects with glaucoma when contrasted with CNNs relying on information from non-visual brain connections. Regularized linear regression models were also tested, and showed significantly weaker classification performance. Additionally, the CNN was unable to generalize to the classification of age-group or of age-related macular degeneration.

CONCLUSIONS

Our findings indicate a distinct and potentially non-linear signature of glaucoma in the tissue properties of optic radiations. This study enhances our understanding of how glaucoma affects brain tissue and opens avenues for further research into how diseases that affect sensory input may also affect brain aging.

Sight threatening diabetic retinopathy in patients with macular telangiectasia type 2

PURPOSE

Although diabetes is highly prevalent in patients with MacTel, progression to severe non-proliferative (NPDR) and proliferative diabetic retinopathy (PDR) is rarely reported. We report multimodal imaging features of sight-threatening diabetic retinopathy (STDR) in eyes with macular telangiectasia type 2 (MacTel).

METHODS

Retrospective case series of seven participants of the MacTel Study at the Moorfields Eye Hospital NHS Foundation Trust study site and one patient from the Institute of Retina and Vitreous of Londrina, Brazil. Sight threatening diabetic retinopathy was defined as severe NPDR, PDR or diabetic macular edema.

RESULTS

We report imaging features of 16 eyes of eight patients (7/8, 87.5% female) with diagnoses of MacTel and type 2 diabetes mellitus with STDR. Mean (SD) age was 56 (8.3) years. Patients were followed-up for a mean time of 9.1 (4.7) years. A total of 10/16 (62.5%) eyes showed PDR and 2/16 (12.5%) eyes presented a macular epiretinal neovascularization.

CONCLUSIONS

People with diabetes mellitus and MacTel may not be protected from STDR as previously reported. Although the two diseases rarely co-exist, regular monitoring for diabetic retinopathy progression is recommended according to baseline retinopathy severity grades in line with established international guidelines. The presence of MacTel may not modify extended screening intervals, but there is no current evidence. The limited case series in the literature support treatment for complications and should follow the standard of care for either condition. Due to dual pathology, reactivation may be difficult to diagnose on standard imaging and multimodal imaging is recommended.

Developing a Continuous Severity Scale for Macular Telangiectasia Type 2 Using Deep Learning and Implications for Disease Grading

PURPOSE

Deep learning (DL) models have achieved state-of-the-art medical diagnosis classification accuracy. Current models are limited by discrete diagnosis labels, but could yield more information with diagnosis in a continuous scale. We developed a novel continuous severity scaling system for macular telangiectasia (MacTel) type 2 by combining a DL classification model with uniform manifold approximation and projection (UMAP).

DESIGN

We used a DL network to learn a feature representation of MacTel severity from discrete severity labels and applied UMAP to embed this feature representation into 2 dimensions, thereby creating a continuous MacTel severity scale.

PARTICIPANTS

A total of 2003 OCT volumes were analyzed from 1089 MacTel Project participants.

METHODS

We trained a multiview DL classifier using multiple B-scans from OCT volumes to learn a previously published discrete 7-step MacTel severity scale. The classifiers' last feature layer was extracted as input for UMAP, which embedded these features into a continuous 2-dimensional manifold. The DL classifier was assessed in terms of test accuracy. Rank correlation for the continuous UMAP scale against the previously published scale was calculated. Additionally, the UMAP scale was assessed in the κ agreement against 5 clinical experts on 100 pairs of patient volumes. For each pair of patient volumes, clinical experts were asked to select the volume with more severe MacTel disease and to compare them against the UMAP scale.

MAIN OUTCOME MEASURES

Classification accuracy for the DL classifier and κ agreement versus clinical experts for UMAP.

RESULTS

The multiview DL classifier achieved top 1 accuracy of 63.3% (186/294) on held-out test OCT volumes. The UMAP metric showed a clear continuous gradation of MacTel severity with a Spearman rank correlation of 0.84 with the previously published scale. Furthermore, the continuous UMAP metric achieved κ agreements of 0.56 to 0.63 with 5 clinical experts, which was comparable with interobserver κ values.

CONCLUSIONS

Our UMAP embedding generated a continuous MacTel severity scale, without requiring continuous training labels. This technique can be applied to other diseases and may lead to more accurate diagnosis, improved understanding of disease progression, and key imaging features for pathologic characteristics.

FINANCIAL DISCLOSURE(S)

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

Human selection bias drives the linear nature of the more ground truth effect in explainable deep learning optical coherence tomography image segmentation

Supervised deep learning (DL) algorithms are highly dependent on training data for which human graders are assigned, for example, for optical coherence tomography (OCT) image annotation. Despite the tremendous success of DL, due to human judgment, these ground truth labels can be inaccurate and/or ambiguous and cause a human selection bias. We therefore investigated the impact of the size of the ground truth and variable numbers of graders on the predictive performance of the same DL architecture and repeated each experiment three times. The largest training dataset delivered a prediction performance close to that of human experts. All DL systems utilized were highly consistent. Nevertheless, the DL under-performers could not achieve any further autonomous improvement even after repeated training. Furthermore, a quantifiable linear relationship between ground truth ambiguity and the beneficial effect of having a larger amount of ground truth data was detected and marked as the more-ground-truth effect.

Retinal vessel volume reference database derived from volume-rendered optical coherence tomography angiography

Optical coherence tomography angiography (OCTA) enables three-dimensional reconstruction of the functional blood vessels in the retina. Therefore, it enables the quantification of 3D retinal vessel parameters such as surface area and vessel volume. In spite of the widespread use of OCTA, no representative volume-rendered vessel volume (VV) data are published to date. In this study, OCTA 3 × 3 mm macular cubes were processed with volume-rendering techniques to measure VV in 203 eyes from 107 healthy volunteers. Generalized linear models (GLM) were constructed to assess the impact of age, gender, visual acuity (VA), spherical equivalent (SE), and axial length (AL) on VV. Overall mean VV was 0.23 ± 0.05mm3. Age and axial length showed a negative correlation with VV. However, GLM model analysis found that AL exerted the most pronounced influence on VV. No statistically significant associations were identified between gender or between left and right eyes. This is the first study to assess 3D OCTA VV and its naturally occurring variations in a large series of healthy subjects. It offers novel insights into the characterization of normal retinal vascular anatomy in healthy individuals, contributing to a valuable reference for future research in this field.

Artificial intelligence-supported diabetic retinopathy screening in Tanzania: rationale and design of a randomised controlled trial

INTRODUCTION

Globally, diabetic retinopathy (DR) is a major cause of blindness. Sub-Saharan Africa is projected to see the largest proportionate increase in the number of people living with diabetes over the next two decades. Screening for DR is recommended to prevent sight loss; however, in many low and middle-income countries, because of a lack of specialist eye care staff, current screening services for DR are not optimal. The use of artificial intelligence (AI) for DR screening, which automates the grading of retinal photographs and provides a point-of-screening result, offers an innovative potential solution to improve DR screening in Tanzania.

METHODS AND ANALYSIS

We will test the hypothesis that AI-supported DR screening increases the proportion of persons with true referable DR who attend the central ophthalmology clinic following referral after screening in a single-masked, parallel group, individually randomised controlled trial. Participants (2364) will be randomised (1:1 ratio) to either AI-supported or the standard of care DR screening pathway. Participants allocated to the AI-supported screening pathway will receive their result followed by point-of-screening counselling immediately after retinal image capture. Participants in the standard of care arm will receive their result and counselling by phone once the retinal images have been graded in the usual way (typically after 2-4 weeks). The primary outcome is the proportion of persons with true referable DR attending the central ophthalmology clinic within 8 weeks of screening. Secondary outcomes, by trial arm, include the proportion of persons attending the central ophthalmology clinic out of all those referred, sensitivity and specificity, number of false positive referrals, acceptability and fidelity of AI-supported screening.

ETHICS AND DISSEMINATION

The London School of Hygiene & Tropical Medicine, Kilimanjaro Christian Medical Centre and Tanzanian National Institute of Medical Research ethics committees have approved the trial. The results will be submitted to peer-reviewed journals for publication.

TRIAL REGISTRATION NUMBER

ISRCTN18317152.

What are the perceptions and concerns of people living with diabetes and National Health Service staff around the potential implementation of AI-assisted screening for diabetic eye disease? Development and validation of a survey for use in a secondary care screening setting

INTRODUCTION

The English National Health Service (NHS) Diabetic Eye Screening Programme (DESP) performs around 2.3 million eye screening appointments annually, generating approximately 13 million retinal images that are graded by humans for the presence or severity of diabetic retinopathy. Previous research has shown that automated retinal image analysis systems, including artificial intelligence (AI), can identify images with no disease from those with diabetic retinopathy as safely and effectively as human graders, and could significantly reduce the workload for human graders. Some algorithms can also determine the level of severity of the retinopathy with similar performance to humans. There is a need to examine perceptions and concerns surrounding AI-assisted eye-screening among people living with diabetes and NHS staff, if AI was to be introduced into the DESP, to identify factors that may influence acceptance of this technology.

METHODS AND ANALYSIS

People living with diabetes and staff from the North East London (NEL) NHS DESP were invited to participate in two respective focus groups to codesign two online surveys exploring their perceptions and concerns around the potential introduction of AI-assisted screening.Focus group participants were representative of the local population in terms of ages and ethnicity. Participants' feedback was taken into consideration to update surveys which were circulated for further feedback. Surveys will be piloted at the NEL DESP and followed by semistructured interviews to assess accessibility, usability and to validate the surveys.Validated surveys will be distributed by other NHS DESP sites, and also via patient groups on social media, relevant charities and the British Association of Retinal Screeners. Post-survey evaluative interviews will be undertaken among those who consent to participate in further research.

ETHICS AND DISSEMINATION

Ethical approval has been obtained by the NHS Research Ethics Committee (IRAS ID: 316631). Survey results will be shared and discussed with focus groups to facilitate preparation of findings for publication and to inform codesign of outreach activities to address concerns and perceptions identified.

Ethnic disparities in progression rates for sight-threatening diabetic retinopathy in diabetic eye screening: a population-based retrospective cohort study

INTRODUCTION

The English Diabetic Eye Screening Programme (DESP) offers people living with diabetes (PLD) annual eye screening. We examined incidence and determinants of sight-threatening diabetic retinopathy (STDR) in a sociodemographically diverse multi-ethnic population.

RESEARCH DESIGN AND METHODS

North East London DESP cohort data (January 2012 to December 2021) with 137 591 PLD with no retinopathy, or non-STDR at baseline in one/both eyes, were used to calculate STDR incidence rates by sociodemographic factors, diabetes type, and duration. HR from Cox models examined associations with STDR.

RESULTS

There were 16 388 incident STDR cases over a median of 5.4 years (IQR 2.8-8.2; STDR rate 2.214, 95% CI 2.214 to 2.215 per 100 person-years). People with no retinopathy at baseline had a lower risk of sight-threatening diabetic retinopathy (STDR) compared with those with non-STDR in one eye (HR 3.03, 95% CI 2.91 to 3.15, p<0.001) and both eyes (HR 7.88, 95% CI 7.59 to 8.18, p<0.001). Black and South Asian individuals had higher STDR hazards than white individuals (HR 1.57, 95% CI 1.50 to 1.64 and HR 1.36, 95% CI 1.31 to 1.42, respectively). Additionally, every 5-year increase in age at inclusion was associated with an 8% reduction in STDR hazards (p<0.001).

CONCLUSIONS

Ethnic disparities exist in a health system limited by capacity rather than patient economic circumstances. Diabetic retinopathy at first screen is a strong determinant of STDR development. By using basic demographic characteristics, screening programmes or clinical practices can stratify risk for sight-threatening diabetic retinopathy development.

Novel Outer Retinal Columnar Abnormalities (ORCA) and Non-Vasogenic Cystoid Macular Edema in Dense Deposit Disease

PURPOSE

To describe the occurrence of bilateral outer retinal columnar abnormalities, non-vasogenic cystoid macular edema, and drusen in the context of dense deposit disease.

METHODS

Case report.

PATIENT

An 18-year-old female with dense deposit disease was referred to our specialist center for diagnosis and management with findings consistent with bilateral non-vasogenic cystoid macular edema and drusen. She was followed up in our clinic for forty months and treated with acetazolamide and ketorolac drops.

RESULTS

Baseline examination revealed bilateral visual acuity (VA) reduction, and macular elevation with peripapillary drusen on fundus biomicroscopy. Optical coherence tomography revealed bilateral hyporeflective cystoid central macula changes, microcystoid changes with increased central subfield thickness (>450 microns), and outer retinal columnar abnormalities (ORCAs). Fluorescein angiography showed no evidence of macular leakage. Electrodiagnostic testing was within normal limits. Over the course of follow-up, she received treatment with acetazolamide 250mg BD PO and ketorolac 0.5% eye drops, with a partial reduction in her edema and improvement in VA.

CONCLUSION

Dense deposit disease is a rare disease secondary to complement cascade dysregulation, associated with drusen. To the best of our knowledge, this is the first report of bilateral non-vasogenic cystoid macular edema and ORCA in a young female patient with dense deposit disease, confirmed with multimodal imaging.

Ethnicity is not biology: retinal pigment score to evaluate biological variability from ophthalmic imaging using machine learning

Background

Few metrics exist to describe phenotypic diversity within ophthalmic imaging datasets, with researchers often using ethnicity as an inappropriate marker for biological variability.

Methods

We derived a continuous, measured metric, the retinal pigment score (RPS), that quantifies the degree of pigmentation from a colour fundus photograph of the eye. RPS was validated using two large epidemiological studies with demographic and genetic data (UK Biobank and EPIC-Norfolk Study).

Findings

A genome-wide association study (GWAS) of RPS from UK Biobank identified 20 loci with known associations with skin, iris and hair pigmentation, of which 8 were replicated in the EPIC-Norfolk cohort. There was a strong association between RPS and ethnicity, however, there was substantial overlap between each ethnicity and the respective distributions of RPS scores.

Interpretation

RPS serves to decouple traditional demographic variables, such as ethnicity, from clinical imaging characteristics. RPS may serve as a useful metric to quantify the diversity of the training, validation, and testing datasets used in the development of AI algorithms to ensure adequate inclusion and explainability of the model performance, critical in evaluating all currently deployed AI models. The code to derive RPS is publicly available at: https://github.com/uw-biomedical-ml/retinal-pigmentation-score.

Funding

The authors did not receive support from any organisation for the submitted work.

Divergent amino acid and sphingolipid metabolism in patients with inherited neuro-retinal disease

OBJECTIVES

The non-essential amino acids serine, glycine, and alanine, as well as diverse sphingolipid species, are implicated in inherited neuro-retinal disorders and are metabolically linked by serine palmitoyltransferase (SPT), a key enzyme in membrane lipid biogenesis. To gain insight into the pathophysiological mechanisms linking these pathways to neuro-retinal diseases we compared patients diagnosed with two metabolically intertwined diseases: macular telangiectasia type II (MacTel), hereditary sensory autonomic neuropathy type 1 (HSAN1), or both.

METHODS

We performed targeted metabolomic analyses of amino acids and broad sphingolipids in sera from a cohort of MacTel (205), HSAN1 (25) and Control (151) participants.

RESULTS

MacTel patients exhibited broad alterations of amino acids, including changes in serine, glycine, alanine, glutamate, and branched-chain amino acids reminiscent of diabetes. MacTel patients had elevated 1-deoxysphingolipids but reduced levels of complex sphingolipids in circulation. A mouse model of retinopathy indicates dietary serine and glycine restriction can drive this depletion in complex sphingolipids. HSAN1 patients exhibited elevated serine, lower alanine, and a reduction in canonical ceramides and sphingomyelins compared to controls. Those patients diagnosed with both HSAN1 and MacTel showed the most significant decrease in circulating sphingomyelins.

CONCLUSIONS

These results highlight metabolic distinctions between MacTel and HSAN1, emphasize the importance of membrane lipids in the progression of MacTel, and suggest distinct therapeutic approaches for these two neurodegenerative diseases.

Training Deep Learning Models to Work on Multiple Devices by Cross-Domain Learning with No Additional Annotations

PURPOSE

To create an unsupervised cross-domain segmentation algorithm for segmenting intraretinal fluid and retinal layers on normal and pathologic macular OCT images from different manufacturers and camera devices.

DESIGN

We sought to use generative adversarial networks (GANs) to generalize a segmentation model trained on one OCT device to segment B-scans obtained from a different OCT device manufacturer in a fully unsupervised approach without labeled data from the latter manufacturer.

PARTICIPANTS

A total of 732 OCT B-scans from 4 different OCT devices (Heidelberg Spectralis, Topcon 1000, Maestro2, and Zeiss Plex Elite 9000).

METHODS

We developed an unsupervised GAN model, GANSeg, to segment 7 retinal layers and intraretinal fluid in Topcon 1000 OCT images (domain B) that had access only to labeled data on Heidelberg Spectralis images (domain A). GANSeg was unsupervised because it had access only to 110 Heidelberg labeled OCTs and 556 raw and unlabeled Topcon 1000 OCTs. To validate GANSeg segmentations, 3 masked graders manually segmented 60 OCTs from an external Topcon 1000 test dataset independently. To test the limits of GANSeg, graders also manually segmented 3 OCTs from Zeiss Plex Elite 9000 and Topcon Maestro2. A U-Net was trained on the same labeled Heidelberg images as baseline. The GANSeg repository with labeled annotations is at https://github.com/uw-biomedical-ml/ganseg.

MAIN OUTCOME MEASURES

Dice scores comparing segmentation results from GANSeg and the U-Net model with the manual segmented images.

RESULTS

Although GANSeg and U-Net achieved comparable Dice scores performance as human experts on the labeled Heidelberg test dataset, only GANSeg achieved comparable Dice scores with the best performance for the ganglion cell layer plus inner plexiform layer (90%; 95% confidence interval [CI], 68%-96%) and the worst performance for intraretinal fluid (58%; 95% CI, 18%-89%), which was statistically similar to human graders (79%; 95% CI, 43%-94%). GANSeg significantly outperformed the U-Net model. Moreover, GANSeg generalized to both Zeiss and Topcon Maestro2 swept-source OCT domains, which it had never encountered before.

CONCLUSIONS

GANSeg enables the transfer of supervised deep learning algorithms across OCT devices without labeled data, thereby greatly expanding the applicability of deep learning algorithms.

A Deep Learning Framework for the Detection and Quantification of Reticular Pseudodrusen and Drusen on Optical Coherence Tomography

Purpose

The purpose of this study was to develop and validate a deep learning (DL) framework for the detection and quantification of reticular pseudodrusen (RPD) and drusen on optical coherence tomography (OCT) scans.

Methods

A DL framework was developed consisting of a classification model and an out-of-distribution (OOD) detection model for the identification of ungradable scans; a classification model to identify scans with drusen or RPD; and an image segmentation model to independently segment lesions as RPD or drusen. Data were obtained from 1284 participants in the UK Biobank (UKBB) with a self-reported diagnosis of age-related macular degeneration (AMD) and 250 UKBB controls. Drusen and RPD were manually delineated by five retina specialists. The main outcome measures were sensitivity, specificity, area under the receiver operating characteristic (ROC) curve (AUC), kappa, accuracy, intraclass correlation coefficient (ICC), and free-response receiver operating characteristic (FROC) curves.

Results

The classification models performed strongly at their respective tasks (0.95, 0.93, and 0.99 AUC, respectively, for the ungradable scans classifier, the OOD model, and the drusen and RPD classification models). The mean ICC for the drusen and RPD area versus graders was 0.74 and 0.61, respectively, compared with 0.69 and 0.68 for intergrader agreement. FROC curves showed that the model's sensitivity was close to human performance.

Conclusions

The models achieved high classification and segmentation performance, similar to human performance.

Translational Relevance

Application of this robust framework will further our understanding of RPD as a separate entity from drusen in both research and clinical settings.

Quantification and Predictors of OCT-Based Macular Curvature and Dome-Shaped Configuration: Results From the UK Biobank

Purpose

To investigate macular curvature, including the evaluation of potential associations and the dome-shaped macular configuration, given the increasing myopia prevalence and expected associated macular malformations.

Methods

The study included a total of 65,440 subjects with a mean age (± SD) of 57.3 ± 8.11 years with spectral-domain optical coherence tomography (OCT) data from a unique contemporary resource for the study of health and disease that recruited more than half a million people in the United Kingdom (UK Biobank). A deep learning model was used to segment the retinal pigment epithelium. The macular curvature of the OCT scans was calculated by polynomial fit and evaluated. Further, associations with demographic, functional, ocular, and infancy factors were examined.

Results

The overall macular curvature values followed a Gaussian distribution with high inter-eye agreement. Although all of the investigated parameters, except maternal smoking, were associated with the curvature in a multilinear analysis, ethnicity and refractive error consistently revealed the most significant effect. The prevalence of a macular dome-shaped configuration was 4.8% overall, most commonly in Chinese subjects as well as hypermetropic eyes. An increasing frequency up to 22.0% was found toward high refractive error. Subretinal fluid was rarely found in these eyes.

Conclusions

Macular curvature revealed associations with demographic, functional, ocular, and infancy factors, as well as increasing prevalence of a dome-shaped macular configuration in high refractive error including high myopia and hypermetropia. These findings imply different pathophysiologic processes that lead to macular development and might open new fields to future myopia and macula research.

Description of a patient cohort with Hereditary Sensory Neuropathy Type 1 without retinal disease Macular Telangiectasia type 2 - implications for retinal screening in HSN1

BACKGROUND AND AIMS

Pathogenic variants in the genes encoding serine palmitoyl transferase (SPTLC1 or SPTLC2) are the most common causes of the rare peripheral nerve disorder Hereditary Sensory Neuropathy Type 1 (HSN1). Macular telangiectasia type 2 (MacTel), a retinal disorder associated with disordered serine-glycine metabolism and has been described in some patients with HSN1. This study aims to further investigate this association in a cohort of people with HSN1.

METHODS

Fourteen patients with a clinically and genetically confirmed diagnosis of HSN1 from the National Hospital for Neurology and Neurosurgery (NHNN, University College London Hospitals NHS Foundation Trust, London, United Kingdom) were recruited to the MacTel Registry, between July 2018 and April 2019. Two additional patients were identified from the dataset of the international clinical registry study (www.lmri.net). Ocular examination included fundus autofluorescence, blue light and infrared reflectance, macular pigment optical density mapping, and optical coherence tomography.

RESULTS

Twelve patients had a pathogenic variant in the SPTLC1 gene, with p.Cys133Trp in eleven cases (92%) and p.Cys133Tyr in one case (8%). Four patients had a variant in the SPTLC2 gene. None of the patients showed clinical evidence of MacTel.

INTERPRETATION

The link between HSN1 and MacTel seems more complex than can solely be explained by the genetic variants. An extension of the spectrum of SPTLC1/2-related disease with phenotypic pleiotropy is proposed. HSN1 patients should be screened for visual symptoms and referred for specialist retinal screening, but the association of the two diseases is likely to be variable and remains unexplained. This article is protected by copyright. All rights reserved.

Foveal Curvature and Its Associations in UK Biobank Participants

Purpose

To examine whether sociodemographic, and ocular factors relate to optical coherence tomography (OCT)-derived foveal curvature (FC) in healthy individuals.

Methods

We developed a deep learning model to quantify OCT-derived FC from 63,939 participants (age range, 39-70 years). Associations of FC with sociodemographic, and ocular factors were obtained using multilevel regression analysis (to allow for right and left eyes) adjusting for age, sex, ethnicity, height (model 1), visual acuity, spherical equivalent, corneal astigmatism, center point retinal thickness (CPRT), intraocular pressure (model 2), deprivation (Townsend index), higher education, annual income, and birth order (model 3). Fovea curvature was modeled as a z-score.

Results

Males had on average steeper FC (0.077; 95% confidence interval [CI] 0.077-0.078) than females (0.068; 95% CI 0.068-0.069). Compared with whites, non-white individuals showed flatter FC, particularly those of black ethnicity. In black males, -0.80 standard deviation (SD) change when compared with whites (95% CI -0.89, -0.71; P 5.2e10-68). In black females, -0.70 SD change when compared with whites (95% CI -0.77, -0.63; p 2.3e10-93). Ocular factors (visual acuity, refractive status, and CPRT) showed a graded inverse association with FC that persisted after adjustment. Macular curvature showed a positive association with FC. Income showed a linear trend increase in males (P for linear trend = 0.005).

Conclusions

We demonstrate marked differences in FC with ethnicity on the largest cohort studied for this purpose to date. Ocular factors showed a graded association with FC. Implementation of FC quantification in research and on the clinical setting can enhance the understanding of clinical macular phenotypes in health and disease.

Estimating excess visual loss from neovascular age-related macular degeneration in the UK during the COVID-19 pandemic: a retrospective clinical audit and simulation model

OBJECTIVES

To report the reduction in new neovascular age-related macular degeneration (nAMD) referrals during the COVID-19 pandemic and estimate the impact of delayed treatment on visual outcomes at 1 year.

DESIGN

Retrospective clinical audit and simulation model.

SETTING

Multiple UK National Health Service (NHS) ophthalmology centres.

PARTICIPANTS

Data on the reduction in new nAMD referrals were obtained from four NHS Trusts comparing April 2020 with April 2019. To estimate the potential impact on 1-year visual outcomes, a stratified bootstrap simulation model was developed drawing on an electronic medical records dataset of 20 825 nAMD eyes from 27 NHS Trusts.

MAIN OUTCOME MEASURES

Simulated mean visual acuity and proportions of eyes with vision ≤6/60, ≤6/24 and ≥6/12 at 1 year under four hypothetical scenarios: 0-month, 3-month, 6-month and 9-month treatment delays. Estimated additional number of eyes with vision ≤6/60 at 1 year nationally.

RESULTS

The number of nAMD referrals dropped on average by 72% (range 65%-87%). Simulated 1-year visual outcomes for 1000 nAMD eyes with a 3-month treatment delay suggested an increase in the proportion of eyes with vision ≤6/60 from 15.5% (13.2%-17.9%) to 23.3% (20.7%-25.9%), and a decrease in the proportion of eyes with vision ≥6/12 (driving vision) from 35.1% (32.1%-38.1%) to 26.4% (23.8%-29.2%). Outcomes worsened incrementally with longer modelled delays. Assuming nAMD referrals are reduced to this level for 1 month nationally, these simulated results suggest an additional 186-365 eyes with vision ≤6/60 at 1 year.

CONCLUSIONS

We report a large decrease in nAMD referrals during the COVID-19 lockdown and provide an important public health message regarding the risk of delayed treatment. As a conservative estimate, a treatment delay of 3 months could lead to a >50% relative increase in the number of eyes with vision ≤6/60 and 25% relative decrease in the number of eyes with driving vision at 1 year.

Elongin C (ELOC/TCEB1)-associated von Hippel-Lindau disease

Around 95% of patients with clinical features that meet the diagnostic criteria for von Hippel-Lindau disease (VHL) have a detectable inactivating germline variant in VHL. The VHL protein (pVHL) functions as part of the E3 ubiquitin ligase complex comprising pVHL, elongin C, elongin B, cullin 2 and ring box 1 (VCB-CR complex), which plays a key role in oxygen sensing and degradation of hypoxia-inducible factors. To date, only variants in VHL have been shown to cause VHL disease. We undertook trio analysis by whole-exome sequencing in a proband with VHL disease but without a detectable VHL mutation. Molecular studies were also performed on paired DNA extracted from the proband's kidney tumour and blood and bioinformatics analysis of sporadic renal cell carcinoma (RCC) dataset was undertaken. A de novo pathogenic variant in ELOC NM_005648.4(ELOC):c.236A>G (p.Tyr79Cys) gene was identified in the proband. ELOC encodes elongin C, a key component [C] of the VCB-CR complex. The p.Tyr79Cys substitution is a mutational hotspot in sporadic VHL-competent RCC and has previously been shown to mimic the effects of pVHL deficiency on hypoxic signalling. Analysis of an RCC from the proband showed similar findings to that in somatically ELOC-mutated RCC (expression of hypoxia-responsive proteins, no somatic VHL variants and chromosome 8 loss). These findings are consistent with pathogenic ELOC variants being a novel cause for VHL disease and suggest that genetic testing for ELOC variants should be performed in individuals with suspected VHL disease with no detectable VHL variant.

Reference database of total retinal vessel surface area derived from volume-rendered optical coherence tomography angiography

Optical coherence tomography angiography (OCTA) enables three-dimensional, high-resolution, depth-resolved flow to be distinguished from non-vessel tissue signals in the retina. Thus, it enables the quantification of the 3D surface area of the retinal vessel signal. Despite the widespread use of OCTA, no representative spatially rendered reference vessel surface area data are published. In this study, the OCTA vessel surface areas in 203 eyes of 107 healthy participants were measured in the 3D domain. A Generalized Linear Model (GLM) model analysis was performed to investigate the effects of sex, age, spherical equivalent, axial length, and visual acuity on the OCTA vessel surface area. The mean overall vessel surface area was 54.53 mm² (range from 27.03 to 88.7 mm²). OCTA vessel surface area was slightly negatively correlated with age. However, the GLM model analysis identified axial length as having the strongest effect on OCTA vessel surface area. No significant correlations were found for sex or between left and right eyes. This is the first study to characterize three-dimensional vascular parameters in a population based on OCTA with respect to the vessel surface area.

Feasibility and tolerability of ophthalmic virtual reality as a medical communication tool in children and young people

PURPOSE

Virtual reality (VR) can be useful in explaining diseases and complications that affect children in order to improve medical communications with this vulnerable patient group. So far, children and young people's responses to high-end medical VR environments have never been assessed.

METHODS

An unprecedented number of 320 children and young people were given the opportunity to interact with a VR application displaying original ophthalmic volume data via a commercially available tethered head-mounted display (HMD). Participants completed three surveys: demographics and experience with VR, usability and perceived utility of this technology and the Simulator Sickness Questionnaire. The second survey also probed participants for suggestions on improvements and whether this system could be useful for increasing engagement in science.

RESULTS

A total of 206 sets of surveys were received. 165 children and young people (84 female) aged 12-18 years (mean, 15 years) completed surveys that could be used for analysis. 69 participants (47.59%) were VR-naïve, and 76 (52.41%) reported that they had previous VR experience. Results show that VR facilitated understanding of ophthalmological complications and was reasonably tolerated. Lastly, exposure to VR raised children and young people's awareness and interest in science.

CONCLUSIONS

The VR platform used was successfully utilized and was well accepted in children to display and interact with volume-rendered 3D ophthalmological data. Virtual reality (VR) is suitable as a novel image display platform in ophthalmology to engage children and young people.

Dynamic volume-rendered optical coherence tomography pupillometry

PURPOSE

To assess intrapupillary space (IPS) changes in healthy subjects with regard to decreased iris motility in patients with pseudoexfoliation glaucoma (PEXG) or non-arteritic anterior ischaemic optic neuropathy (NAION) in a feasibility study in a clinical environment.

METHODS

Scotopic and photopic IPS measurements using three-dimensionally rendered swept-source optical coherence tomography (SS-OCT) data were obtained and compared for all subjects. Intrapupillary space (IPS) parameters were evaluated such as absolute volumetric differences, relative light response for volumetric ratios and pupillary ejection fraction (PEF) for functional contraction measurements.

RESULTS

From a total of 122 IPS from 66 subjects, 106 IPS were eligible for comparison providing values for 72 normal, 30 PEXG and 4 NAION eyes. In healthy, PEXG and NAION subjects, scotopic overall mean IPS was 8.90, 3.45 and 4.16 mm³ , and photopic overall mean IPS was 0.87, 0.74 and 1.13 mm³ , respectively. Three-dimensional contractility showed a mean absolute difference of 8.03 mm³ for normals (defined as 100% contractility), 2.72 mm³ for PEXG (33.88% of normal) and 3.03 mm³ for NAION (38.50% of normal) with a relative light response ratio between scotopic and photopic volumes of 10.26 (100%), 4.69 (45.70%) and 3.67 (35.78%), respectively. Pupillary ejection fraction (PEF) showed a contractile pupillary emptying of 88.11% for normals, 76.92% for PEXG and 70.91% for NAION patients.

CONCLUSION

This 3D pupillometry OCT assessment allows for quantitative measurements of pupil function, contractility and response to light. More specifically, PEF is presented as a potential (neuro)-pupillary outcome measure that could be useful in the monitoring of ophthalmic disorders that affect pupillary function.

Multicentre study of 4626 patients assesses the effectiveness, safety and burden of two categories of treatments for central retinal vein occlusion: intravitreal anti-vascular endothelial growth factor injections and intravitreal Ozurdex injections

BACKGROUND/AIMS

To assess the effectiveness, burden and safety of two categories of treatment for central retinal vein occlusion (CRVO): intravitreal injections of anti-vascular endothelial growth factor (anti-VEGF) and dexamethasone (Ozurdex).

METHODS

A retrospective analysis of Medisoft electronic medical record (EMR) data from 27 National Health Service sites in the UK identified 4626 treatment-naive patients with a single mode of treatment for macular oedema secondary to CRVO. Statistics describing the overall CRVO patient cohort and individual patient subpopulations stratified by treatment type were generated. Mean age at baseline, gender, ethnicity, social deprivation and visual acuity (VA) follow-up was reported. Absolute and change in VA using ETDRS are used to describe treatment effectiveness, the number of injections and visits used to describe treatment burden and endophthalmitis rates as a marker of treatment safety.

RESULTS

Mean VA was 47.9 and 45.3 EDTRS letters in the anti-VEGF and Ozurdex groups, respectively. This changed to 57.9/53.7 at 12 months, 58.3/46.9 at 18 months and 59.4/51.0 at 36 months. Mean number of injections were 5.6/1.6 at 12 months, 6.0/1.7 at 18 months and 7.0/1.8 at 36 months. Endophthalmitis rates were 0.003% (n=4) for the anti-VEGF group and 0.09% (n=1) for the Ozurdex group.

CONCLUSIONS

VA improvements were greater and more sustained with anti-VEGF treatment. Lower starting acuity resulted in bigger gains in both groups, while higher starting acuity resulted in higher VA at 36 months. Although treatment burden was greater with anti-VEGF, Ozurdex was associated with higher rates of endophthalmitis.

Automated Detection of Glaucoma With Interpretable Machine Learning Using Clinical Data and Multimodal Retinal Images

PURPOSE

To develop a multimodal model to automate glaucoma detection DESIGN: Development of a machine-learning glaucoma detection model METHODS: We selected a study cohort from the UK Biobank data set with 1193 eyes of 863 healthy subjects and 1283 eyes of 771 subjects with glaucoma. We trained a multimodal model that combines multiple deep neural nets, trained on macular optical coherence tomography volumes and color fundus photographs, with demographic and clinical data. We performed an interpretability analysis to identify features the model relied on to detect glaucoma. We determined the importance of different features in detecting glaucoma using interpretable machine learning methods. We also evaluated the model on subjects who did not have a diagnosis of glaucoma on the day of imaging but were later diagnosed (progress-to-glaucoma [PTG]).

RESULTS

Results show that a multimodal model that combines imaging with demographic and clinical features is highly accurate (area under the curve 0.97). Interpretation of this model highlights biological features known to be related to the disease, such as age, intraocular pressure, and optic disc morphology. Our model also points to previously unknown or disputed features, such as pulmonary function and retinal outer layers. Accurate prediction in PTG highlights variables that change with progression to glaucoma-age and pulmonary function.

CONCLUSIONS

The accuracy of our model suggests distinct sources of information in each imaging modality and in the different clinical and demographic variables. Interpretable machine learning methods elucidate subject-level prediction and help uncover the factors that lead to accurate predictions, pointing to potential disease mechanisms or variables related to the disease.

Uncovering of intraspecies macular heterogeneity in cynomolgus monkeys using hybrid machine learning optical coherence tomography image segmentation

The fovea is a depression in the center of the macula and is the site of the highest visual acuity. Optical coherence tomography (OCT) has contributed considerably in elucidating the pathologic changes in the fovea and is now being considered as an accompanying imaging method in drug development, such as antivascular endothelial growth factor and its safety profiling. Because animal numbers are limited in preclinical studies and automatized image evaluation tools have not yet been routinely employed, essential reference data describing the morphologic variations in macular thickness in laboratory cynomolgus monkeys are sparse to nonexistent. A hybrid machine learning algorithm was applied for automated OCT image processing and measurements of central retina thickness and surface area values. Morphological variations and the effects of sex and geographical origin were determined. Based on our findings, the fovea parameters are specific to the geographic origin. Despite morphological similarities among cynomolgus monkeys, considerable variations in the foveolar contour, even within the same species but from different geographic origins, were found. The results of the reference database show that not only the entire retinal thickness, but also the macular subfields, should be considered when designing preclinical studies and in the interpretation of foveal data.

Effect of ethnicity and other sociodemographic factors on attendance at diabetic eye screening: a 12-month retrospective cohort study

OBJECTIVES

To examine the association of sociodemographic characteristics with attendance at diabetic eye screening in a large ethnically diverse urban population.

DESIGN

Retrospective cohort study.

SETTING

Screening visits in the North East London Diabetic Eye Screening Programme (NELDESP).

PARTICIPANTS

84 449 people with diabetes aged 12 years or older registered in the NELDESP and scheduled for screening between 1 April 2017 and 31 March 2018.

MAIN OUTCOME MEASURE

Attendance at diabetic eye screening appointments.

RESULTS

The mean age of people with diabetes was 60 years (SD 14.2 years), 53.4% were men, 41% South Asian, 29% White British and 17% Black; 83.4% attended screening. Black people with diabetes had similar levels of attendance compared with White British people. However, South Asian, Chinese and 'Any other Asian' background ethnicities showed greater odds of attendance compared with White British. When compared with their respective reference group, high levels of deprivation, younger age, longer duration of diabetes and worse visual acuity, were all associated with non-attendance. There was a higher likelihood of attendance per quintile improvement in deprivation (OR, 1.06; 95% CI, 1.03 to 1.08), with increasing age (OR per decade, 1.17; 95% CI, 1.15 to 1.19), with better visual acuity (OR per Bailey-Lovie chart line 1.12; 95% CI, 1.11 to 1.14) and with longer time of NELDESP registration (OR per year, 1.02; 95% CI, 1.01 to 1.03).

CONCLUSION

Ethnic differences in diabetic eye screening uptake, though small, are evident. Despite preconceptions, a higher likelihood of screening attendance was observed among Asian ethnic groups when compared with the White ethnic group. Poorer socioeconomic profile was associated with higher likelihood of non-attendance for screening. Further work is needed to understand how to target individuals at risk of non-attendance and reduce inequalities.

Evolving Treatment Patterns and Outcomes of Neovascular Age-Related Macular Degeneration Over a Decade

PURPOSE

Management of neovascular age-related macular degeneration (nAMD) has evolved over the last decade with several treatment regimens and medications. This study describes the treatment patterns and visual outcomes over 10 years in a large cohort of patients.

DESIGN

Retrospective analysis of electronic health records from 27 National Health Service secondary care healthcare providers in the UK.

PARTICIPANTS

Treatment-naïve patients receiving at least 3 intravitreal anti-vascular endothelial growth factor (VEGF) injections for nAMD in their first 6 months of follow-up were included. Patients with missing data for age or gender and those aged less than 55 years were excluded.

METHODS

Eyes with at least 3 years of follow-up were grouped by years of treatment initiation, and 3-year outcomes were compared between the groups. Data were generated during routine clinical care between September 2008 and December 2018.

MAIN OUTCOME MEASURES

Visual acuity (VA), number of injections, and number of visits.

RESULTS

A total of 15 810 eyes of 13 705 patients receiving 195 104 injections were included. Visual acuity improved from baseline during the first year, but decreased thereafter, resulting in loss of visual gains. This trend remained consistent throughout the past decade. Although an increasing proportion of eyes remained in the driving standard, this was driven by better presenting VA over the decade. The number of injections decreased substantially between the first and subsequent years, from a mean of 6.25 in year 1 to 3 in year 2 and 2.5 in year 3, without improvement over the decade. In a multivariable regression analysis, final VA improved by 0.24 letters for each year since 2008, and younger age and baseline VA were significantly associated with VA at 3 years.

CONCLUSIONS

Our findings show that despite improvement in functional VA over the years, primarily driven by improving baseline VA, patients continue to lose vision after the first year of treatment, with only marginal change over the past decade. The data suggest these results may be related to suboptimal treatment patterns, which have not improved over the years. Rethinking treatment strategies may be warranted, possibly on a national level or through the introduction of longer-acting therapies.

Foveal structure and visual function in nanophthalmos and posterior microphthalmos

BACKGROUND/AIMS

The reason for visual impairment in patients with nanophthalmos and posterior microphthalmos is not completely understood. Therefore, this study aims to investigate foveal structure, and the impact of demographic, clinical and imaging parameters on best-corrected visual acuity (BCVA) in these conditions.

METHODS

Sixty-two eyes of 33 patients with nanophthalmos (n=40) or posterior microphthalmos (n=22), and 114 eyes of healthy controls with high-resolution retinal imaging including spectral-domain or swept-source optical coherence tomography images were included in this cross-sectional case-control study. Foveal retinal layer thickness was determined by two independent readers. A mixed-effect model was used to perform structure-function correlations and predict the BCVA based on subject-specific variables.

RESULTS

Most patients (28/33) had altered foveal structure associated with loss of foveal avascular zone and impaired BCVA. However, widening of outer nuclear layer, lengthening of photoreceptor outer segments, normal distribution of macular pigment and presence of Henle fibres were consistently found. Apart from the presence of choroidal effusion, which had significant impact on BCVA, the features age, refractive error, axial length and retinal layer thickness at the foveal centre explained 61.7% of the variability of BCVA.

CONCLUSION

This study demonstrates that choroidal effusion, age, refractive error, axial length and retinal layer thickness are responsible for the majority of interindividual variability of BCVA as well as the morphological foveal heterogeneity in patients with nanophthalmos or posterior microphthalmos. This might give further insights into the physiology of foveal development and the process of emmetropisation, and support clinicians in the assessment of these disease entities.

Real world evidence on 5661 patients treated for macular oedema secondary to branch retinal vein occlusion with intravitreal anti-vascular endothelial growth factor, intravitreal dexamethasone or macular laser

BACKGROUND/AIMS

Clinical trials suggest anti-vascular endothelial growth factor is more effective than intravitreal dexamethasone as treatment for macular oedema secondary to branch retinal vein occlusion. This study asks if 'real world' data from a larger and more diverse population, followed for a longer period, also support this conclusion.

METHODS

Data collected to support routine care at 27 NHS (National Health Service) Trusts between February 2002 and September 2017 contained 5661 treatment-naive patients with a single mode of treatment for macular oedema secondary to branch retinal vein occlusion and no history of cataract surgery either during or recently preceding the treatment. Number of treatment visits and change in visual acuity from baseline was plotted for three treatment groups (anti-vascular endothelial growth factor (anti-VEGF), intravitreal dexamethasone, macular laser) for up to 3 years.

RESULTS

Mean baseline visual acuity was 57.1/53.1/62.3 letters in the anti-VEGF/dexamethasone/macular laser groups, respectively. This changed to 66.72 (+9.6)/57.6 (+4.5)/63.2 (+0.9) at 12 months. Adequate numbers allowed analysis at 18 months for all groups (66.6 (+9.5)/56.1 (+3.0)/60.8 (-1.5)) and for anti-VEGF at 36 months (68.0, +10.9) Mean number of treatments were 5.1/1.5/1.2 at 12 months, 5.9/1.7/1.2 at 18 months for all three groups and 10.3 at 36 months for anti-VEGF.

CONCLUSIONS

Visual acuity improvements were higher and more sustained with anti-VEGF. Higher treatment burden occurred with anti-VEGF but this reduced over 36 months. Patients with better vision at baseline than those in the clinical trials maintained high levels of vision with both anti-VEGF and dexamethasone.

Contextualizing single-arm trials with real-world data: An emulated target trial comparing therapies for neovascular age-related macular degeneration

Abstract

One‐in‐four ophthalmology trials are single‐armed, which poses challenges to their interpretation. We demonstrate how real‐world cohorts used as external/synthetic control arms can contextualize such trials. We herein emulated a target trial on the intention‐to‐treat efficacy of off‐label bevacizumab (q6w) pro re nata relative to fixed‐interval aflibercept (q8w) for improving week 54 visual acuity of eyes affected by neovascular age‐related macular degeneration. The bevacizumab arm (n = 65) was taken from the ABC randomized controlled trial. A total of 4,471 aflibercept‐treated eyes aligning with the ABC trial eligibility were identified from electronic health records and synthetic control arms were created by emulating randomization conditional on age, sex, and baseline visual read via exact matching and propensity score methods. We undertook an inferiority analysis on mean difference at 54 weeks; outcomes regression on achieving a change in visual acuity of greater than or equal to 15, greater than or equal to 10, and less than or equal to −15 Early Treatment Diabetic Retinopathy (ETDRS) letters at week 54; and a time‐to‐event analysis on achieving a change in visual acuity of greater than or equal to 15, greater than or equal to 10, and less than or equal to −15 ETDRS letters by week 54. The findings suggest off‐label bevacizumab to be neither inferior nor superior to licensed aflibercept. Our study highlights how real‐world cohorts representing the counterfactual intervention could aid the interpretation of single‐armed trials when analyzed in accord to the target trial framework.

Unraveling the deep learning gearbox in optical coherence tomography image segmentation towards explainable artificial intelligence

Machine learning has greatly facilitated the analysis of medical data, while the internal operations usually remain intransparent. To better comprehend these opaque procedures, a convolutional neural network for optical coherence tomography image segmentation was enhanced with a Traceable Relevance Explainability (T-REX) technique. The proposed application was based on three components: ground truth generation by multiple graders, calculation of Hamming distances among graders and the machine learning algorithm, as well as a smart data visualization ('neural recording'). An overall average variability of 1.75% between the human graders and the algorithm was found, slightly minor to 2.02% among human graders. The ambiguity in ground truth had noteworthy impact on machine learning results, which could be visualized. The convolutional neural network balanced between graders and allowed for modifiable predictions dependent on the compartment. Using the proposed T-REX setup, machine learning processes could be rendered more transparent and understandable, possibly leading to optimized applications.

Intravitreal ranibizumab for the management of serous maculopathy secondary to optic disc coloboma-associated choroidal neovascularisation

We report the case of a 19-year-old patient with symptomatic unilateral serous maculopathy associated with an optic nerve coloboma. Fluorescein angiography detected a focal late leak at the temporal edge of the coloboma which was later found to correspond with an area of choroidal neovascularisation on optical coherence tomography angiography. A course of intravitreal ranibizumab achieved good clinical and structural response. This report contributes to the evidence that maculopathies associated with cavitary optic nerve anomalies may in some instances result from choroidal neovascularisation. It also highlights the importance of angiography to identify potential choroidal neovascular membranes, particularly in the absence of haemorrhages and neovascular membranes on fundus examination and conventional optical coherence tomography.

A virtual-clinic pathway for patients referred from a national diabetes eye screening programme reduces service demands whilst maintaining quality of care

Aim

To evaluate the potential of an integrated virtual medical retina clinic in secondary care for diabetic patients screened and referred by the UK National Diabetic Eye Screening Program (DESP).

Methods

This retrospective cohort study included diabetic patients referred by the DESP to either a virtual or a traditional doctor’s appointment (face-to-face, F2F) at the Moorfields Eye Hospital NHS Foundation Trust (London, UK) between January 2015 and December 2018. The primary outcome was the proportion of patients that qualified for a virtual-clinic appointment according to hospital guidance. Secondary outcomes included the rate of attendance, mean time from DESP referral to initial hospital appointment, mean time-to-discharge and -to-treatment of either panretinal photocoagulation or intravitreal injection of anti-vascular endothelial growth factor.

Results

We included 12,563 patients in this study. While 8833 patients (70.7%) would have qualified for a virtual appointment according to local triage guidance, only 2306 (18.4%) were referred to a virtual consultation due to capacity constraints. For routine referrals, mean time to the first hospital appointment was 66.9 days with a standard deviation of ±35.9 and 80.9 ± 44.4 days for a virtual and a F2F consultation, respectively. The mean time from referral to discharge to community was 71.7 ± 30.8 and 86.3 ± 37.0 days for a virtual and a F2F consultation, respectively. We did not observe a statistically significant difference in the mean time-to-treatment in the sub-cohort that required intravitreal therapy for maculopathy (virtual clinics: 220.7 ± 84.8; F2F: 178.0 days ± 80.7; p value > 0.05). Moreover, we observed a non-inferior attendance rate in virtual as compared to F2F clinics.

Conclusion

A significant proportion of diabetic patients referred to a F2F clinic could initially be managed in a virtual clinic. Increasing the adoption of virtual clinics in the management of diabetic patients that do not need long-term management or monitoring in secondary services may help alleviate service demands without diminishing quality of clinical care. Collectively, our analyses suggest that virtual consultations are a faster and clinically appropriate alternative for a substantial proportion of diabetic patients.

Longitudinal Assessment of Remnant Foveal Cone Structure in a Case Series of Early Macular Telangiectasia Type 2

Purpose

To determine the extent of remnant cone structure within early foveal ellipsoid zone (EZ) lesions in macular telangiectasia type 2 longitudinally using both confocal and split detector adaptive optics scanning light ophthalmoscopy (AOSLO).

Methods

Spectral domain optical coherence tomography (SDOCT), confocal and split detector AOSLO were acquired from seven patients (10 eyes) with small (early) EZ lesions on SDOCT secondary to macular telangiectasia type 2 at baseline, 6 months, and 12 months. The presence of cone structure on AOSLO in areas of EZ loss as well as cones at 1° eccentricity, and their change over time were quantified.

Results

By split detector AOSLO, remnant cone structure was identified within and on the borders of all foveal EZ lesions. Within the extent of these lesions, cone spacing ranged from 4.97 to 9.95 µm at baseline, 5.30 to 6.10 µm at 6 months, and 4.99 to 7.12 µm at 12 months. Four eyes with significantly smaller EZ lesions showed evidence of recovery of EZ reflectivity on SDOCT B-scans. Remnant cone structure was identified in some areas where EZ reflectivity recovered at the following time point. Eyes that showed recovery of EZ reflectivity had a continuous external limiting membrane.

Conclusions

Remnant cone structure can persist within small SDOCT-defined EZ lesions, which can wax and wane in appearance over time. AOSLO can help to inform the interpretation of SDOCT imaging.

Translational Relevance

The absence of EZ in early macular telangiectasia type 2 and other retinal conditions needs careful interpretation because it does not always indicate an absence of underlying cone structure. The integrity of the external limiting membrane may better predict the presence of remnant cone structure and recovery of EZ reflectivity.

Validation of virtual reality orbitometry bridges digital and physical worlds

Clinical science and medical imaging technology are traditionally displayed in two dimensions (2D) on a computer monitor. In contrast, three-dimensional (3D) virtual reality (VR) expands the realm of 2D image visualization, enabling an immersive VR experience with unhindered spatial interaction by the user. Thus far, analysis of data extracted from VR applications was mainly qualitative. In this study, we enhance VR and provide evidence for quantitative VR research by validating digital VR display of computed tomography (CT) data of the orbit. Volumetric CT data were transferred and rendered into a VR environment. Subsequently, seven graders performed repeated and blinded diameter measurements. The intergrader variability of the measurements in VR was much lower compared to measurements in the physical world and measurements were reasonably consistent with their corresponding elements in the real context. The overall VR measurements were 5.49% higher. As such, this study attests the ability of VR to provide similar quantitative data alongside the added benefit of VR interfaces. VR entails a lot of potential for the future research in ophthalmology and beyond in any scientific field that uses three-dimensional data.

Prospective evaluation of an artificial intelligence-enabled algorithm for automated diabetic retinopathy screening of 30 000 patients

Background/aims

Human grading of digital images from diabetic retinopathy (DR) screening programmes represents a significant challenge, due to the increasing prevalence of diabetes. We evaluate the performance of an automated artificial intelligence (AI) algorithm to triage retinal images from the English Diabetic Eye Screening Programme (DESP) into test-positive/technical failure versus test-negative, using human grading following a standard national protocol as the reference standard.

Methods

Retinal images from 30 405 consecutive screening episodes from three English DESPs were manually graded following a standard national protocol and by an automated process with machine learning enabled software, EyeArt v2.1. Screening performance (sensitivity, specificity) and diagnostic accuracy (95% CIs) were determined using human grades as the reference standard.

Results

Sensitivity (95% CIs) of EyeArt was 95.7% (94.8% to 96.5%) for referable retinopathy (human graded ungradable, referable maculopathy, moderate-to-severe non-proliferative or proliferative). This comprises sensitivities of 98.3% (97.3% to 98.9%) for mild-to-moderate non-proliferative retinopathy with referable maculopathy, 100% (98.7%,100%) for moderate-to-severe non-proliferative retinopathy and 100% (97.9%,100%) for proliferative disease. EyeArt agreed with the human grade of no retinopathy (specificity) in 68% (67% to 69%), with a specificity of 54.0% (53.4% to 54.5%) when combined with non-referable retinopathy.

Conclusion

The algorithm demonstrated safe levels of sensitivity for high-risk retinopathy in a real-world screening service, with specificity that could halve the workload for human graders. AI machine learning and deep learning algorithms such as this can provide clinically equivalent, rapid detection of retinopathy, particularly in settings where a trained workforce is unavailable or where large-scale and rapid results are needed.

Feasibility and Safety of a Coaxial Dual-Wavelength Optical Coherence Tomography Apparatus

PURPOSE

To evaluate the feasibility and safety of a coaxial dual-wavelength optical coherence tomography (OCT) device (marked as Hydra-OCT).

METHODS

Healthy participants without ocular pathology underwent retinal imaging using the Hydra-OCT allowing for simultaneous measurement of retinal scanning of 840 and 1,072 nm wavelength. Before and after measurement, best-corrected visual acuity and patients' comfort were assessed. Representative OCT images from both wavelengths were compared by 5 independent graders using a subjective grading scheme.

RESULTS

A total of 30 eyes of 30 participants (8 females and 22 males) with a mean age of 26.5 years (range from 19 to 55 years) were included. Dual-wavelength image acquisition was made possible in each subject. The participant's effort and comfort assessment using the Hydra-OCT imaging revealed an equivalent value as compared to the commercially available OCT machine. No adverse events were reported, and visual acuity was not altered by the Hydra-OCT. Imaging between the systems was comparable.

CONCLUSIONS

This study provides evidence for the feasibility and safety of a coaxial dual-wavelength OCT imaging method under real-life conditions. The novel Hydra-OCT imaging device may offer additional insights into the pathology of retinal and choroidal diseases.

Diagnostic accuracy of diabetic retinopathy grading by an artificial intelligence-enabled algorithm compared with a human standard for wide-field true-colour confocal scanning and standard digital retinal images

BACKGROUND

Photographic diabetic retinopathy screening requires labour-intensive grading of retinal images by humans. Automated retinal image analysis software (ARIAS) could provide an alternative to human grading. We compare the performance of an ARIAS using true-colour, wide-field confocal scanning images and standard fundus images in the English National Diabetic Eye Screening Programme (NDESP) against human grading.

METHODS

Cross-sectional study with consecutive recruitment of patients attending annual diabetic eye screening. Imaging with mydriasis was performed (two-field protocol) with the EIDON platform (CenterVue, Padua, Italy) and standard NDESP cameras. Human grading was carried out according to NDESP protocol. Images were processed by EyeArt V.2.1.0 (Eyenuk Inc, Woodland Hills, California). The reference standard for analysis was the human grade of standard NDESP images.

RESULTS

We included 1257 patients. Sensitivity estimates for retinopathy grades were: EIDON images; 92.27% (95% CI: 88.43% to 94.69%) for any retinopathy, 99% (95% CI: 95.35% to 100%) for vision-threatening retinopathy and 100% (95% CI: 61% to 100%) for proliferative retinopathy. For NDESP images: 92.26% (95% CI: 88.37% to 94.69%) for any retinopathy, 100% (95% CI: 99.53% to 100%) for vision-threatening retinopathy and 100% (95% CI: 61% to 100%) for proliferative retinopathy. One case of vision-threatening retinopathy (R1M1) was missed by the EyeArt when analysing the EIDON images, but identified by the human graders. The EyeArt identified all cases of vision-threatening retinopathy in the standard images.

CONCLUSION

EyeArt identified diabetic retinopathy in EIDON images with similar sensitivity to standard images in a large-scale screening programme, exceeding the sensitivity threshold recommended for a screening test. Further work to optimise the identification of 'no retinopathy' and to understand the differential lesion detection in the two imaging systems would enhance the use of these two innovative technologies in a diabetic retinopathy screening setting.

Progression of Retinopathy Secondary to Maternally Inherited Diabetes and Deafness - Evaluation of Predicting Parameters

PURPOSE

To investigate the prognostic value of demographic, functional, and imaging parameters on retinal pigment epithelium (RPE) atrophy progression secondary to maternally inherited diabetes and deafness (MIDD) and to evaluate the application of these factors in clinical trial design.

DESIGN

Retrospective observational case series.

METHODS

Thirty-five eyes of 20 patients (age range, 24.9-75.9 years) with genetically proven MIDD and demarcated RPE atrophy on serial fundus autofluorescence (AF) images were included. Lesion size and shape-descriptive parameters were longitudinally determined by 2 independent readers. A linear mixed-effect model was used to predict the lesion enlargement rate based on baseline variables. Sample size calculations were performed to model the power in a simulated interventional study.

RESULTS

The mean follow-up time was 4.27 years. The mean progression rate of RPE atrophy was 2.33 mm²/year, revealing a dependence on baseline lesion size (+0.04 [0.02-0.07] mm²/year/mm², P < .001), which was absent after square root transformation. The fovea was preserved in the majority of patients during the observation time. In the case of foveal involvement, the loss of visual acuity lagged behind central RPE atrophy in AF images. Sex, age, and number of atrophic foci predicted future progression rates with a cross-validated mean absolute error of 0.13 mm/year and to reduce the required sample size for simulated interventional trials.

CONCLUSIONS

Progressive RPE atrophy could be traced in all eyes using AF imaging. Shape-descriptive factors and patients' baseline characteristics had significant prognostic value, guiding appropriate subject selection and sample size in future interventional trial design.

The Usefulness of Serum Biomarkers in the Early Stages of Diabetic Retinopathy: Results of the EUROCONDOR Clinical Trial

The main aim of this study was to evaluate the ability of serum biomarkers to predict the worsening of retinal neurodysfunction in subjects with type 2 diabetes. For this purpose, we measured selected molecules (N-epsilon-carboxy methyl lysine (CML), laminin P1 (Lam-P1), and asymmetric dimethylarginine (ADMA)) in the serum of 341 participants of the EUROCONDOR study at baseline, 24, and 48 weeks. Retinal neurodysfunction was assessed by measuring implicit time (IT) using multifocal electroretinography, and structural changes were examined by spectral domain–optical coherence tomography. The values of IT at baseline were directly correlated with baseline serum concentrations of CML (r = 0.135, p = 0.013). Furthermore, in the placebo group, increase in CML concentration throughout follow-up correlated with the IT (r = 0.20; p = 0.03). Baseline serum levels of CML also correlated with macular retinal thickness (RT) (r = 0.231; p < 0.001). Baseline Lam-P1 levels correlated with the increase of the RT at the end of follow-up in the placebo group (r = 0.22; p = 0.016). We provide evidence that CML may be a biomarker of both retinal neurodysfunction and RT, whereas Lam-P1 was associated with RT only. Therefore, circulating levels of these molecules could provide a complementary tool for monitoring the early changes of diabetic retinopathy (DR).

Comparison of true-colour wide-field confocal scanner imaging with standard fundus photography for diabetic retinopathy screening

BACKGROUND

Screening of diabetic retinopathy (DR) reduces blindness by early identification of retinopathy. This study compares DR grades derived from a two-field imaging protocol from two imaging platforms, one providing a single 60-degree horizontal field of view (FOV) and the other, a standard 45-degree FOV.

METHODS

Cross-sectional study which included 1257 diabetic patients aged ≥18 years attending their DR screening visit in the English National Diabetic Eye Screening Programme (NDESP). Patients with maculopathy (M1), preproliferative (R2) or proliferative DR (R3) were referred to an ophthalmologist. Patients with ungradable images (U) are examined in a slit-lamp biomicroscopy clinic. Image acquisition under mydriasis of two images per eye was carried out with the EIDON and with standard fundus cameras. Evaluation was performed by masked graders.

RESULTS

Agreement after consensus with kappa statistic was 0.89 (quadratic weights (95% CI 0.87 to 0.92)) for NDESP severity grade, 0.88 (quadratic weights (95% CI 0.82 to 0.94)) for referable disease and 0.92 (linear weights (95% CI 0.88 to 0.95)) for maculopathy. The EIDON detected clinically relevant DR features outside the 45-degree fields in two patients (0.16%): one with intraretinal microvascular abnormalities (IRMAs) and one with neovascularisation. In eight patients (0.64%), the EIDON allowed DR feature visualisation inside the 45-degree fields that were not identified in the NDESP images: three patients (0.24%) with IRMA and five patients (0.40%) with maculopathy. The rates of ungradable encounters were 12 (0.95%) and 13 (1.03%) with the EIDON and NDESP images, respectively.

CONCLUSION

The EIDON identifies a small number of additional patients with referable disease which are not detected with standard imaging. This is due to the EIDON finding disease outside the standard FOV and greater clarity finding disease within the standard FOV.

Binocular Inhibition of Reading in Macular Telangiectasia Type 2

Purpose

To assess the presence of binocular gain in macular telangiectasia type 2 (MacTel) and its correlation to paracentral scotomas.

Methods

Sixty-eight patients with MacTel were consecutively recruited for a cross-sectional analysis. Best-corrected visual acuity (BCVA), reading acuity, and reading speed were tested monocularly and binocularly. Macular retinal sensitivity was examined with fundus-controlled perimetry (microperimetry). Scotomas were quantified by their size, their depth, and their proximity to the fovea.

Results

Binocular reading speed and acuity were lower than monocular reading speed and acuity in the functionally better eye (142 vs. 159 words per minute and 0.43 vs. 0.28 log reading acuity determination, P < 0.001). Magnitude of binocular inhibition of reading speed was correlated to the degree of interocular functional difference (R2 = 0.61, P < 0.001). This correlation was not found for reading acuity or BCVA (R2 < 0.03). Binocular reading speed was negatively correlated to size of right and left eye scotomas, with bigger effect size for left eye scotomas. The magnitude of binocular inhibition was correlated to size of left eye scotomas, but not of right eye scotomas. When both eyes had similar scotoma characteristics, the right eye was more frequently the better reading eye.

Conclusions

We provide evidence for the presence of binocular inhibition of reading performance in MacTel, likely due to binocular rivalry. This may result from the characteristic paracentral scotomas in noncorresponding retinal fields and, in particular, a disruptive projection of scotomas in reading direction arising from the left eyes. Patients may benefit from occluding one eye while reading.

Progression characteristics of ellipsoid zone loss in macular telangiectasia type 2

PURPOSE

To investigate the progression characteristics of ellipsoid zone (EZ) loss in eyes with macular telangiectasia type 2 (MacTel) as reflected by area and linear measurements, and their relevance for visual acuity.

METHODS

Participants were selected from the MacTel Study cohort. Linear and area measurements of EZ loss were performed in Spectral-Domain Optical Coherence Tomograph (SD-OCT) volume scans. Progression characteristics and correlations between linear and area measurements were analysed using linear mixed effects models.

RESULTS

A total of 134 eyes of 70 patients were included (85 eyes with follow-up, mean 4.7 years, range: 1.4-8 years). Ellipsoid zone (EZ) loss significantly progressed at a mean annual increment of 0.057 mm² (p = 0.005). The progression rate was non-linear and interacted significantly with initial EZ lesion size indicating an exponential growth before reaching a plateau. There was a strong heterogeneity in area sizes between fellow eyes. EZ break length had a significant linear effect on EZ break area (b = 1.06, p < 0.001) and could predict it. The location of the EZ break had a significant impact on visual acuity.

CONCLUSION

Ellipsoid zone (EZ) loss in MacTel has a non-linear progression characteristic, and its rate depends on area size at baseline, which must be taken into account at sample selection in clinical trials. Our results show a good correlation of linear and area measures of EZ loss and a segregation of best-corrected visual acuity by EZ location, which may help routine clinical practice.

Serine and Lipid Metabolism in Macular Disease and Peripheral Neuropathy

BACKGROUND

Identifying mechanisms of diseases with complex inheritance patterns, such as macular telangiectasia type 2, is challenging. A link between macular telangiectasia type 2 and altered serine metabolism has been established previously.

METHODS

Through exome sequence analysis of a patient with macular telangiectasia type 2 and his family members, we identified a variant in SPTLC1 encoding a subunit of serine palmitoyltransferase (SPT). Because mutations affecting SPT are known to cause hereditary sensory and autonomic neuropathy type 1 (HSAN1), we examined 10 additional persons with HSAN1 for ophthalmologic disease. We assayed serum amino acid and sphingoid base levels, including levels of deoxysphingolipids, in patients who had macular telangiectasia type 2 but did not have HSAN1 or pathogenic variants affecting SPT. We characterized mice with low serine levels and tested the effects of deoxysphingolipids on human retinal organoids.

RESULTS

Two variants known to cause HSAN1 were identified as causal for macular telangiectasia type 2: of 11 patients with HSAN1, 9 also had macular telangiectasia type 2. Circulating deoxysphingolipid levels were 84.2% higher among 125 patients with macular telangiectasia type 2 who did not have pathogenic variants affecting SPT than among 94 unaffected controls. Deoxysphingolipid levels were negatively correlated with serine levels, which were 20.6% lower than among controls. Reduction of serine levels in mice led to increases in levels of retinal deoxysphingolipids and compromised visual function. Deoxysphingolipids caused photoreceptor-cell death in retinal organoids, but not in the presence of regulators of lipid metabolism.

CONCLUSIONS

Elevated levels of atypical deoxysphingolipids, caused by variant SPTLC1 or SPTLC2 or by low serine levels, were risk factors for macular telangiectasia type 2, as well as for peripheral neuropathy. (Funded by the Lowy Medical Research Institute and others.).

Validation of automated artificial intelligence segmentation of optical coherence tomography images

PURPOSE

To benchmark the human and machine performance of spectral-domain (SD) and swept-source (SS) optical coherence tomography (OCT) image segmentation, i.e., pixel-wise classification, for the compartments vitreous, retina, choroid, sclera.

METHODS

A convolutional neural network (CNN) was trained on OCT B-scan images annotated by a senior ground truth expert retina specialist to segment the posterior eye compartments. Independent benchmark data sets (30 SDOCT and 30 SSOCT) were manually segmented by three classes of graders with varying levels of ophthalmic proficiencies. Nine graders contributed to benchmark an additional 60 images in three consecutive runs. Inter-human and intra-human class agreement was measured and compared to the CNN results.

RESULTS

The CNN training data consisted of a total of 6210 manually segmented images derived from 2070 B-scans (1046 SDOCT and 1024 SSOCT; 630 C-Scans). The CNN segmentation revealed a high agreement with all grader groups. For all compartments and groups, the mean Intersection over Union (IOU) score of CNN compartmentalization versus group graders' compartmentalization was higher than the mean score for intra-grader group comparison.

CONCLUSION

The proposed deep learning segmentation algorithm (CNN) for automated eye compartment segmentation in OCT B-scans (SDOCT and SSOCT) is on par with manual segmentations by human graders.

Feasibility of support vector machine learning in age-related macular degeneration using small sample yielding sparse optical coherence tomography data

PURPOSE

A retrospective pilot study is conducted to demonstrate the utility of a novel support vector machine learning (SVML) algorithm in a small three-dimensional (3D) sample yielding sparse optical coherence tomography (spOCT) data for the automatic monitoring of neovascular (wet) age-related macular degeneration (wAMD).

METHODS

From the anti-vascular endothelial growth factor injection database, 588 consecutive pairs of OCT volumes (57.624 B-scans) were selected in 70 randomly chosen wAMD patients treated with ranibizumab. The SVML algorithm was applied to 183 OCT volume pairs (17.934 B-scans) in 30 patients. Four independent, diagnosis-blinded retina specialists indicated whether wAMD activity was present between 100 pairs of consecutive OCT volumes (9800 B-scans) in the remaining 40 patients for comparison with the SVML algorithm and a non-complex baseline algorithm using only retinal thickness. The SVML algorithm was assessed using inter-observer variability and receiver operating characteristic (ROC) analyses.

RESULTS

The retina specialists showed an average Cohen's κ of 0.57 ± 0.13 (minimum: 0.41, maximum: 0.83). The average κ between the proposed algorithm and the retina specialists was 0.62 ± 0.05 and 0.43 ± 0.14 between the baseline algorithm and the retina specialists. Using each of the four retina specialists as the reference, the proposed method showed a superior area under the ROC curve of 0.91 ± 0.03 compared to the ROC 0.81 ± 0.05 shown by the baseline algorithm.

CONCLUSION

The SVML algorithm was as effective as the retina specialists were in detecting activity in wAMD. Support vector machine learning (SVML) may be a useful monitoring tool in wAMD suited for small samples that yield sparse OCT data possibly derived from self-measuring OCT-robots.