Oral Antioxidant and Lutein/Zeaxanthin Supplements Slow Geographic Atrophy Progression to the Fovea in Age-Related Macular Degeneration.

PURPOSE: The only treatments approved to slow geographic atrophy (GA) progression in age-related macular degeneration (AMD) require frequent intraocular injection and suffer from modest efficacy, important risks, and high costs. The purpose of this study was to determine whether oral supplements slow GA progression in AMD. DESIGN: Post hoc analysis of the Age-Related Eye Diseases Study (AREDS) and AREDS2, multi-center randomized placebo-controlled trials of oral micronutrient supplementation, each with 2x2 factorial design. PARTICIPANTS: 392 eyes (318 participants) with GA in AREDS; 1210 eyes (891 participants) with GA in AREDS2. METHODS: AREDS participants were randomly assigned to oral antioxidants (500 mg vitamin C; 400 IU vitamin E; 15 mg ß-carotene); 80 mg zinc; combination; or placebo. AREDS2 participants were randomly assigned to 10 mg lutein/2 mg zeaxanthin; 350 mg docosahexaenoic acid/650 mg eicosapentaenoic acid; combination; or placebo. Consenting AREDS2 participants were also randomly assigned to alternative AREDS formulations: original; no beta-carotene; 25 mg zinc instead of 80 mg; both. MAIN OUTCOME MEASURES: (1) Change in GA proximity to central macula over time, and (2) change in square root GA area over time, each measured from color fundus photographs at annual visits and analyzed by mixed-model regression according to randomized assignments. RESULTS: In AREDS eyes with non-central GA (n=208), proximity-based progression towards the central macula was significantly slower with randomization to antioxidants versus none, at 50.7 µm/year (95% CI 38.0-63.4 µm/year) versus 72.9 µm/year (95% CI 61.3-84.5 µm/year; p=0.012), respectively. In AREDS2 eyes with non-central GA, in participants assigned to AREDS antioxidants without ß-carotene (n=325 eyes), proximity-based progression was significantly slower with randomization to lutein/zeaxanthin versus none, at 80.1 µm/year (95% CI 60.9-99.3 µm/year) versus 114.4 µm/year (95% CI 96.2-132.7 µm/year; p=0.011), respectively. In AREDS eyes with any GA (n=392), area-based progression was not significantly different with randomization to antioxidants versus none (p=0.63). In AREDS2 eyes with any GA, in participants assigned to AREDS antioxidants without ß-carotene (n=505 eyes), area-based progression was not significantly different with randomization to lutein/zeaxanthin versus none (p=0.64). CONCLUSIONS: Oral micronutrient supplementation slowed GA progression towards the central macula, likely by augmenting the natural phenomenon of foveal sparing.

Foundation models in ophthalmology.

Foundation models represent a paradigm shift in artificial intelligence (AI), evolving from narrow models designed for specific tasks to versatile, generalisable models adaptable to a myriad of diverse applications. Ophthalmology as a specialty has the potential to act as an exemplar for other medical specialties, offering a blueprint for integrating foundation models broadly into clinical practice. This review hopes to serve as a roadmap for eyecare professionals seeking to better understand foundation models, while equipping readers with the tools to explore the use of foundation models in their own research and practice. We begin by outlining the key concepts and technological advances which have enabled the development of these models, providing an overview of novel training approaches and modern AI architectures. Next, we summarise existing literature on the topic of foundation models in ophthalmology, encompassing progress in vision foundation models, large language models and large multimodal models. Finally, we outline major challenges relating to privacy, bias and clinical validation, and propose key steps forward to maximise the benefit of this powerful technology.

Generative artificial intelligence in ophthalmology: current innovations, future applications and challenges.

The rapid advancements in generative artificial intelligence are set to significantly influence the medical sector, particularly ophthalmology. Generative adversarial networks and diffusion models enable the creation of synthetic images, aiding the development of deep learning models tailored for specific imaging tasks. Additionally, the advent of multimodal foundational models, capable of generating images, text and videos, presents a broad spectrum of applications within ophthalmology. These range from enhancing diagnostic accuracy to improving patient education and training healthcare professionals. Despite the promising potential, this area of technology is still in its infancy, and there are several challenges to be addressed, including data bias, safety concerns and the practical implementation of these technologies in clinical settings.

Study protocol: optical coherence tomography angiography for the detection of neovascular age-related macular degeneration: a comprehensive multicentre diagnostic accuracy study in the UK-the ATHENA study.

INTRODUCTION: The diagnosis of neovascular age-related macular degeneration (nAMD), the leading cause of visual impairment in the developed world, relies on the interpretation of various imaging tests of the retina. These include invasive angiographic methods, such as Fundus Fluorescein Angiography (FFA) and, on occasion, Indocyanine-Green Angiography (ICGA). Newer, non-invasive imaging modalities, predominately Optical Coherence Tomography (OCT) and Optical Coherence Tomography Angiography (OCTA), have drastically transformed the diagnostic approach to nAMD. The aim of this study is to undertake a comprehensive diagnostic accuracy assessment of the various imaging modalities used in clinical practice for the diagnosis of nAMD (OCT, OCTA, FFA and, when a variant of nAMD called Polypoidal Choroidal Vasculopathy is suspected, ICGA) both alone and in various combinations. METHODS AND ANALYSIS: This is a non-inferiority, prospective, randomised diagnostic accuracy study of 1067 participants. Participants are patients with clinical features consistent with nAMD who present to a National Health Service secondary care ophthalmology unit in the UK. Patients will undergo OCT as per standard practice and those with suspicious features of nAMD on OCT will be approached for participation in the study. Patients who agree to take part will also undergo both OCTA and FFA (and ICGA if indicated). Interpretation of the imaging tests will be undertaken by clinicians at recruitment sites. A randomised design was selected to avoid bias from consecutive review of all imaging tests by the same clinician. The primary outcome of the study will be the difference in sensitivity and specificity between OCT+OCTA and OCT+FFA (±ICGA) for nAMD detection as interpreted by clinicians at recruitment sites. ETHICS AND DISSEMINATION: The study has been approved by the South Central-Oxford B Research Ethics Committee with reference number 21/SC/0412.Dissemination of study results will involve peer-review publications, presentations at major national and international scientific conferences. TRIAL REGISTRATION NUMBER: ISRCTN18313457.

Vision-Language Models for Feature Detection of Macular Diseases on Optical Coherence Tomography.

Importance: Vision-language models (VLMs) are a novel artificial intelligence technology capable of processing image and text inputs. While demonstrating strong generalist capabilities, their performance in ophthalmology has not been extensively studied. Objective: To assess the performance of the Gemini Pro VLM in expert-level tasks for macular diseases from optical coherence tomography (OCT) scans. Design, Setting, and Participants: This was a cross-sectional diagnostic accuracy study evaluating a generalist VLM on ophthalmology-specific tasks using the open-source Optical Coherence Tomography Image Database. The dataset included OCT B-scans from 50 unique patients: healthy individuals and those with macular hole, diabetic macular edema, central serous chorioretinopathy, and age-related macular degeneration. Each OCT scan was labeled for 10 key pathological features, referral recommendations, and treatments. The images were captured using a Cirrus high definition OCT machine (Carl Zeiss Meditec) at Sankara Nethralaya Eye Hospital, Chennai, India, and the dataset was published in December 2018. Image acquisition dates were not specified. Exposures: Gemini Pro, using a standard prompt to extract structured responses on December 15, 2023. Main Outcomes and Measures: The primary outcome was model responses compared against expert labels, calculating F1 scores for each pathological feature. Secondary outcomes included accuracy in diagnosis, referral urgency, and treatment recommendation. The model's internal concordance was evaluated by measuring the alignment between referral and treatment recommendations, independent of diagnostic accuracy. Results: The mean F1 score was 10.7% (95% CI, 2.4-19.2). Measurable F1 scores were obtained for macular hole (36.4%; 95% CI, 0-71.4), pigment epithelial detachment (26.1%; 95% CI, 0-46.2), subretinal hyperreflective material (24.0%; 95% CI, 0-45.2), and subretinal fluid (20.0%; 95% CI, 0-45.5). A correct diagnosis was achieved in 17 of 50 cases (34%; 95% CI, 22-48). Referral recommendations varied: 28 of 50 were correct (56%; 95% CI, 42-70), 10 of 50 were overcautious (20%; 95% CI, 10-32), and 12 of 50 were undercautious (24%; 95% CI, 12-36). Referral and treatment concordance were very high, with 48 of 50 (96%; 95 % CI, 90-100) and 48 of 49 (98%; 95% CI, 94-100) correct answers, respectively. Conclusions and Relevance: In this study, a generalist VLM demonstrated limited vision capabilities for feature detection and management of macular disease. However, it showed low self-contradiction, suggesting strong language capabilities. As VLMs continue to improve, validating their performance on large benchmarking datasets will help ascertain their potential in ophthalmology.

Pegcetacoplan Treatment and Consensus Features of Geographic Atrophy Over 24 Months.

Importance: Despite widespread availability and consensus on its advantages for detailed imaging of geographic atrophy (GA), spectral-domain optical coherence tomography (SD-OCT) might benefit from automated quantitative OCT analyses in GA diagnosis, monitoring, and reporting of its landmark clinical trials. Objective: To analyze the association between pegcetacoplan and consensus GA SD-OCT end points. Design, Setting, and Participants: This was a post hoc analysis of 11 614 SD-OCT volumes from 936 of the 1258 participants in 2 parallel phase 3 studies, the Study to Compare the Efficacy and Safety of Intravitreal APL-2 Therapy With Sham Injections in Patients With Geographic Atrophy (GA) Secondary to Age-Related Macular Degeneration (OAKS) and Study to Compare the Efficacy and Safety of Intravitreal APL-2 Therapy With Sham Injections in Patients With Geographic Atrophy (GA) Secondary to Age-Related Macular Degeneration (DERBY). OAKS and DERBY were 24-month, multicenter, randomized, double-masked, sham-controlled studies conducted from August 2018 to July 2020 among adults with GA with total area 2.5 to 17.5 mm2 on fundus autofluorescence imaging (if multifocal, at least 1 lesion ≥1.25 mm2). This analysis was conducted from September to December 2023. Interventions: Study participants received pegcetacoplan, 15 mg per 0.1-mL intravitreal injection, monthly or every other month, or sham injection monthly or every other month. Main Outcomes and Measures: The primary end point was the least squares mean change from baseline in area of retinal pigment epithelium and outer retinal atrophy in each of the 3 treatment arms (pegcetacoplan monthly, pegcetacoplan every other month, and pooled sham [sham monthly and sham every other month]) at 24 months. Feature-specific area analysis was conducted by Early Treatment Diabetic Retinopathy Study (ETDRS) regions of interest (ie, foveal, parafoveal, and perifoveal). Results: Among 936 participants, the mean (SD) age was 78.5 (7.22) years, and 570 participants (60.9%) were female. Pegcetacoplan, but not sham treatment, was associated with reduced growth rates of SD-OCT biomarkers for GA for up to 24 months. Reductions vs sham in least squares mean (SE) change from baseline of retinal pigment epithelium and outer retinal atrophy area were detectable at every time point from 3 through 24 months (least squares mean difference vs pooled sham at month 24, pegcetacoplan monthly: -0.86 mm2; 95% CI, -1.15 to -0.57; P < .001; pegcetacoplan every other month: -0.69 mm2; 95% CI, -0.98 to -0.39; P < .001). This association was more pronounced with more frequent dosing (pegcetacoplan monthly vs pegcetacoplan every other month at month 24: -0.17 mm2; 95% CI, -0.43 to 0.08; P = .17). Stronger associations were observed in the parafoveal and perifoveal regions for both pegcetacoplan monthly and pegcetacoplan every other month. Conclusions and Relevance: These findings offer additional insight into the potential effects of pegcetacoplan on the development of GA, including potential effects on the retinal pigment epithelium and photoreceptors. Trial Registration: ClinicalTrials.gov Identifiers: NCT03525600 and NCT03525613.

Medical education with large language models in ophthalmology: custom instructions and enhanced retrieval capabilities.

Foundation models are the next generation of artificial intelligence that has the potential to provide novel use cases for healthcare. Large language models (LLMs), a type of foundation model, are capable of language comprehension and the ability to generate human-like text. Researchers and developers have been tuning LLMs to optimise their performance in specific tasks, such as medical challenge problems. Until recently, tuning required technical programming expertise, but the release of custom generative pre-trained transformers (GPTs) by OpenAI has allowed users to tune their own GPTs with natural language. This has the potential to democratise access to high-quality bespoke LLMs globally. In this review, we provide an overview of LLMs, how they are tuned and how custom GPTs work. We provide three use cases of custom GPTs in ophthalmology to demonstrate the versatility and effectiveness of these tools. First, we present 'EyeTeacher', an educational aid that generates questions from clinical guidelines to facilitate learning. Second, we built 'EyeAssistant', a clinical support tool that is tuned with clinical guidelines to respond to various physician queries. Lastly, we design 'The GPT for GA', which offers clinicians a comprehensive summary of emerging management strategies for geographic atrophy by analysing peer-reviewed documents. The review underscores the significance of custom instructions and information retrieval in tuning GPTs for specific tasks in ophthalmology. We also discuss the evaluation of LLM responses and address critical aspects such as privacy and accountability in their clinical application. Finally, we discuss their potential in ophthalmic education and clinical practice.

Comparing generative and retrieval-based chatbots in answering patient questions regarding age-related macular degeneration and diabetic retinopathy.

BACKGROUND/AIMS: To compare the performance of generative versus retrieval-based chatbots in answering patient inquiries regarding age-related macular degeneration (AMD) and diabetic retinopathy (DR). METHODS: We evaluated four chatbots: generative models (ChatGPT-4, ChatGPT-3.5 and Google Bard) and a retrieval-based model (OcularBERT) in a cross-sectional study. Their response accuracy to 45 questions (15 AMD, 15 DR and 15 others) was evaluated and compared. Three masked retinal specialists graded the responses using a three-point Likert scale: either 2 (good, error-free), 1 (borderline) or 0 (poor with significant inaccuracies). The scores were aggregated, ranging from 0 to 6. Based on majority consensus among the graders, the responses were also classified as 'Good', 'Borderline' or 'Poor' quality. RESULTS: Overall, ChatGPT-4 and ChatGPT-3.5 outperformed the other chatbots, both achieving median scores (IQR) of 6 (1), compared with 4.5 (2) in Google Bard, and 2 (1) in OcularBERT (all p ≤8.4×10-3). Based on the consensus approach, 83.3% of ChatGPT-4's responses and 86.7% of ChatGPT-3.5's were rated as 'Good', surpassing Google Bard (50%) and OcularBERT (10%) (all p ≤1.4×10-2). ChatGPT-4 and ChatGPT-3.5 had no 'Poor' rated responses. Google Bard produced 6.7% Poor responses, and OcularBERT produced 20%. Across question types, ChatGPT-4 outperformed Google Bard only for AMD, and ChatGPT-3.5 outperformed Google Bard for DR and others. CONCLUSION: ChatGPT-4 and ChatGPT-3.5 demonstrated superior performance, followed by Google Bard and OcularBERT. Generative chatbots are potentially capable of answering domain-specific questions outside their original training. Further validation studies are still required prior to real-world implementation.

Periodontitis and Outer Retinal Thickness: a Cross-Sectional Analysis of the United Kingdom Biobank Cohort.

Purpose: Periodontitis, a ubiquitous severe gum disease affecting the teeth and surrounding alveolar bone, can heighten systemic inflammation. We investigated the association between very severe periodontitis and early biomarkers of age-related macular degeneration (AMD), in individuals with no eye disease. Design: Cross-sectional analysis of the prospective community-based cohort United Kingdom (UK) Biobank. Participants: Sixty-seven thousand three hundred eleven UK residents aged 40 to 70 years recruited between 2006 and 2010 underwent retinal imaging. Methods: Macular-centered OCT images acquired at the baseline visit were segmented for retinal sublayer thicknesses. Very severe periodontitis was ascertained through a touchscreen questionnaire. Linear mixed effects regression modeled the association between very severe periodontitis and retinal sublayer thicknesses, adjusting for age, sex, ethnicity, socioeconomic status, alcohol consumption, smoking status, diabetes mellitus, hypertension, refractive error, and previous cataract surgery. Main Outcome Measures: Photoreceptor layer (PRL) and retinal pigment epithelium-Bruch's membrane (RPE-BM) thicknesses. Results: Among 36 897 participants included in the analysis, 1571 (4.3%) reported very severe periodontitis. Affected individuals were older, lived in areas of greater socioeconomic deprivation, and were more likely to be hypertensive, diabetic, and current smokers (all P < 0.001). On average, those with very severe periodontitis were hyperopic (0.05 ± 2.27 diopters) while those unaffected were myopic (-0.29 ± 2.40 diopters, P < 0.001). Following adjusted analysis, very severe periodontitis was associated with thinner PRL (-0.55 µm, 95% confidence interval [CI], -0.97 to -0.12; P = 0.022) but there was no difference in RPE-BM thickness (0.00 µm, 95% CI, -0.12 to 0.13; P = 0.97). The association between PRL thickness and very severe periodontitis was modified by age (P < 0.001). Stratifying individuals by age, thinner PRL was seen among those aged 60 to 69 years with disease (-1.19 µm, 95% CI, -1.85 to -0.53; P < 0.001) but not among those aged < 60 years. Conclusions: Among those with no known eye disease, very severe periodontitis is statistically associated with a thinner PRL, consistent with incipient AMD. Optimizing oral hygiene may hold additional relevance for people at risk of degenerative retinal disease. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

AI as a Medical Device for Ophthalmic Imaging in Europe, Australia, and the United States: Protocol for a Systematic Scoping Review of Regulated Devices.

BACKGROUND: Artificial intelligence as a medical device (AIaMD) has the potential to transform many aspects of ophthalmic care, such as improving accuracy and speed of diagnosis, addressing capacity issues in high-volume areas such as screening, and detecting novel biomarkers of systemic disease in the eye (oculomics). In order to ensure that such tools are safe for the target population and achieve their intended purpose, it is important that these AIaMD have adequate clinical evaluation to support any regulatory decision. Currently, the evidential requirements for regulatory approval are less clear for AIaMD compared to more established interventions such as drugs or medical devices. There is therefore value in understanding the level of evidence that underpins AIaMD currently on the market, as a step toward identifying what the best practices might be in this area. In this systematic scoping review, we will focus on AIaMD that contributes to clinical decision-making (relating to screening, diagnosis, prognosis, and treatment) in the context of ophthalmic imaging. OBJECTIVE: This study aims to identify regulator-approved AIaMD for ophthalmic imaging in Europe, Australia, and the United States; report the characteristics of these devices and their regulatory approvals; and report the available evidence underpinning these AIaMD. METHODS: The Food and Drug Administration (United States), the Australian Register of Therapeutic Goods (Australia), the Medicines and Healthcare products Regulatory Agency (United Kingdom), and the European Database on Medical Devices (European Union) regulatory databases will be searched for ophthalmic imaging AIaMD through a snowballing approach. PubMed and clinical trial registries will be systematically searched, and manufacturers will be directly contacted for studies investigating the effectiveness of eligible AIaMD. Preliminary regulatory database searches, evidence searches, screening, data extraction, and methodological quality assessment will be undertaken by 2 independent review authors and arbitrated by a third at each stage of the process. RESULTS: Preliminary searches were conducted in February 2023. Data extraction, data synthesis, and assessment of methodological quality commenced in October 2023. The review is on track to be completed and submitted for peer review by April 2024. CONCLUSIONS: This systematic review will provide greater clarity on ophthalmic imaging AIaMD that have achieved regulatory approval as well as the evidence that underpins them. This should help adopters understand the range of tools available and whether they can be safely incorporated into their clinical workflow, and it should also support developers in navigating regulatory approval more efficiently. INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/52602.

Diagnostic decisions of specialist optometrists exposed to ambiguous deep-learning outputs.

Artificial intelligence (AI) has great potential in ophthalmology. We investigated how ambiguous outputs from an AI diagnostic support system (AI-DSS) affected diagnostic responses from optometrists when assessing cases of suspected retinal disease. Thirty optometrists (15 more experienced, 15 less) assessed 30 clinical cases. For ten, participants saw an optical coherence tomography (OCT) scan, basic clinical information and retinal photography ('no AI'). For another ten, they were also given AI-generated OCT-based probabilistic diagnoses ('AI diagnosis'); and for ten, both AI-diagnosis and AI-generated OCT segmentations ('AI diagnosis + segmentation') were provided. Cases were matched across the three types of presentation and were selected to include 40% ambiguous and 20% incorrect AI outputs. Optometrist diagnostic agreement with the predefined reference standard was lowest for 'AI diagnosis + segmentation' (204/300, 68%) compared to 'AI diagnosis' (224/300, 75% p = 0.010), and 'no Al' (242/300, 81%, p = < 0.001). Agreement with AI diagnosis consistent with the reference standard decreased (174/210 vs 199/210, p = 0.003), but participants trusted the AI more (p = 0.029) with segmentations. Practitioner experience did not affect diagnostic responses (p = 0.24). More experienced participants were more confident (p = 0.012) and trusted the AI less (p = 0.038). Our findings also highlight issues around reference standard definition.

Leading in the development, standardised evaluation, and adoption of artificial intelligence in clinical practice: regional anaesthesia as an example.

A recent study by Suissa and colleagues explored the clinical relevance of a medical image segmentation metric (Dice metric) commonly used in the field of artificial intelligence (AI). They showed that pixel-wise agreement for physician identification of structures on ultrasound images is variable, and a relatively low Dice metric (0.34) correlated to a substantial agreement on subjective clinical assessment. We highlight the need to bring structure and clinical perspective to the evaluation of medical AI, which clinicians are best placed to direct.

Concordance of randomised controlled trials for artificial intelligence interventions with the CONSORT-AI reporting guidelines.

The Consolidated Standards of Reporting Trials extension for Artificial Intelligence interventions (CONSORT-AI) was published in September 2020. Since its publication, several randomised controlled trials (RCTs) of AI interventions have been published but their completeness and transparency of reporting is unknown. This systematic review assesses the completeness of reporting of AI RCTs following publication of CONSORT-AI and provides a comprehensive summary of RCTs published in recent years. 65 RCTs were identified, mostly conducted in China (37%) and USA (18%). Median concordance with CONSORT-AI reporting was 90% (IQR 77-94%), although only 10 RCTs explicitly reported its use. Several items were consistently under-reported, including algorithm version, accessibility of the AI intervention or code, and references to a study protocol. Only 3 of 52 included journals explicitly endorsed or mandated CONSORT-AI. Despite a generally high concordance amongst recent AI RCTs, some AI-specific considerations remain systematically poorly reported. Further encouragement of CONSORT-AI adoption by journals and funders may enable more complete adoption of the full CONSORT-AI guidelines.

Cross-modality Labeling Enables Noninvasive Capillary Quantification as a Sensitive Biomarker for Assessing Cardiovascular Risk.

Purpose: We aim to use fundus fluorescein angiography (FFA) to label the capillaries on color fundus (CF) photographs and train a deep learning model to quantify retinal capillaries noninvasively from CF and apply it to cardiovascular disease (CVD) risk assessment. Design: Cross-sectional and longitudinal study. Participants: A total of 90732 pairs of CF-FFA images from 3893 participants for segmentation model development, and 49229 participants in the UK Biobank for association analysis. Methods: We matched the vessels extracted from FFA and CF, and used vessels from FFA as labels to train a deep learning model (RMHAS-FA) to segment retinal capillaries using CF. We tested the model's accuracy on a manually labeled internal test set (FundusCapi). For external validation, we tested the segmentation model on 7 vessel segmentation datasets, and investigated the clinical value of the segmented vessels in predicting CVD events in the UK Biobank. Main Outcome Measures: Area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity for segmentation. Hazard ratio (HR; 95% confidence interval [CI]) for Cox regression analysis. Results: On the FundusCapi dataset, the segmentation performance was AUC = 0.95, accuracy = 0.94, sensitivity = 0.90, and specificity = 0.93. Smaller vessel skeleton density had a stronger correlation with CVD risk factors and incidence (P < 0.01). Reduced density of small vessel skeletons was strongly associated with an increased risk of CVD incidence and mortality for women (HR [95% CI] = 0.91 [0.84-0.98] and 0.68 [0.54-0.86], respectively). Conclusions: Using paired CF-FFA images, we automated the laborious manual labeling process and enabled noninvasive capillary quantification from CF, supporting its potential as a sensitive screening method for identifying individuals at high risk of future CVD events. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

CF-Loss: Clinically-relevant feature optimised loss function for retinal multi-class vessel segmentation and vascular feature measurement.

Characterising clinically-relevant vascular features, such as vessel density and fractal dimension, can benefit biomarker discovery and disease diagnosis for both ophthalmic and systemic diseases. In this work, we explicitly encode vascular features into an end-to-end loss function for multi-class vessel segmentation, categorising pixels into artery, vein, uncertain pixels, and background. This clinically-relevant feature optimised loss function (CF-Loss) regulates networks to segment accurate multi-class vessel maps that produce precise vascular features. Our experiments first verify that CF-Loss significantly improves both multi-class vessel segmentation and vascular feature estimation, with two standard segmentation networks, on three publicly available datasets. We reveal that pixel-based segmentation performance is not always positively correlated with accuracy of vascular features, thus highlighting the importance of optimising vascular features directly via CF-Loss. Finally, we show that improved vascular features from CF-Loss, as biomarkers, can yield quantitative improvements in the prediction of ischaemic stroke, a real-world clinical downstream task. The code is available at https://github.com/rmaphoh/feature-loss.

Adalimumab vs placebo as add-on to Standard Therapy for autoimmune Uveitis: Tolerability, Effectiveness and cost-effectiveness-a protocol for a randomised controlled trial (ASTUTE trial).

INTRODUCTION: Adalimumab is an effective treatment for autoimmune non-infectious uveitis (ANIU), but it is currently only funded for a minority of patients with ANIU in the UK as it is restricted by the National Institute for Health and Care Excellence guidance. Ophthalmologists believe that adalimumab may be effective in a wider range of patients. The Adalimumab vs placebo as add-on to Standard Therapy for autoimmune Uveitis: Tolerability, Effectiveness and cost-effectiveness (ASTUTE) trial will recruit patients with ANIU who do and do not meet funding criteria and will evaluate the effectiveness and cost-effectiveness of adalimumab versus placebo as an add-on therapy to standard care. METHODS AND ANALYSIS: The ASTUTE trial is a multicentre, parallel-group, placebo-controlled, pragmatic randomised controlled trial with a 16-week treatment run-in (TRI). At the end of the TRI, only responders will be randomised (1:1) to 40 mg adalimumab or placebo (both are the study investigational medicinal product) self-administered fortnightly by subcutaneous injection. The target sample size is 174 randomised participants. The primary outcome is time to treatment failure (TF), a composite of signs indicative of active ANIU. Secondary outcomes include individual TF components, retinal morphology, adverse events, health-related quality of life, patient-reported side effects and visual function, best-corrected visual acuity, employment status and resource use. In the event of TF, open-label drug treatment will be restarted as per TRI for 16 weeks, and if a participant responds again, allocation will be switched without unmasking and treatment with investigational medicinal product restarted. ETHICS AND DISSEMINATION: The trial received Research Ethics Committee (REC) approval from South Central - Oxford B REC in June 2020. The findings will be presented at international meetings, by peer-reviewed publications and through patient organisations and newsletters to patients, where available. TRIAL REGISTRATION: ISRCTN31474800. Registered 14 April 2020.

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.

Validation of a deep learning system for the detection of diabetic retinopathy in Indigenous Australians.

BACKGROUND/AIMS: Deep learning systems (DLSs) for diabetic retinopathy (DR) detection show promising results but can underperform in racial and ethnic minority groups, therefore external validation within these populations is critical for health equity. This study evaluates the performance of a DLS for DR detection among Indigenous Australians, an understudied ethnic group who suffer disproportionately from DR-related blindness. METHODS: We performed a retrospective external validation study comparing the performance of a DLS against a retinal specialist for the detection of more-than-mild DR (mtmDR), vision-threatening DR (vtDR) and all-cause referable DR. The validation set consisted of 1682 consecutive, single-field, macula-centred retinal photographs from 864 patients with diabetes (mean age 54.9 years, 52.4% women) at an Indigenous primary care service in Perth, Australia. Three-person adjudication by a panel of specialists served as the reference standard. RESULTS: For mtmDR detection, sensitivity of the DLS was superior to the retina specialist (98.0% (95% CI, 96.5 to 99.4) vs 87.1% (95% CI, 83.6 to 90.6), McNemar's test p<0.001) with a small reduction in specificity (95.1% (95% CI, 93.6 to 96.4) vs 97.0% (95% CI, 95.9 to 98.0), p=0.006). For vtDR, the DLS's sensitivity was again superior to the human grader (96.2% (95% CI, 93.4 to 98.6) vs 84.4% (95% CI, 79.7 to 89.2), p<0.001) with a slight drop in specificity (95.8% (95% CI, 94.6 to 96.9) vs 97.8% (95% CI, 96.9 to 98.6), p=0.002). For all-cause referable DR, there was a substantial increase in sensitivity (93.7% (95% CI, 91.8 to 95.5) vs 74.4% (95% CI, 71.1 to 77.5), p<0.001) and a smaller reduction in specificity (91.7% (95% CI, 90.0 to 93.3) vs 96.3% (95% CI, 95.2 to 97.4), p<0.001). CONCLUSION: The DLS showed improved sensitivity and similar specificity compared with a retina specialist for DR detection. This demonstrates its potential to support DR screening among Indigenous Australians, an underserved population with a high burden of diabetic eye disease.

Image quality assessment of retinal fundus photographs for diabetic retinopathy in the machine learning era: a review.

This study aimed to evaluate the image quality assessment (IQA) and quality criteria employed in publicly available datasets for diabetic retinopathy (DR). A literature search strategy was used to identify relevant datasets, and 20 datasets were included in the analysis. Out of these, 12 datasets mentioned performing IQA, but only eight specified the quality criteria used. The reported quality criteria varied widely across datasets, and accessing the information was often challenging. The findings highlight the importance of IQA for AI model development while emphasizing the need for clear and accessible reporting of IQA information. The study suggests that automated quality assessments can be a valid alternative to manual labeling and emphasizes the importance of establishing quality standards based on population characteristics, clinical use, and research purposes. In conclusion, image quality assessment is important for AI model development; however, strict data quality standards must not limit data sharing. Given the importance of IQA for developing, validating, and implementing deep learning (DL) algorithms, it's recommended that this information be reported in a clear, specific, and accessible way whenever possible. Automated quality assessments are a valid alternative to the traditional manual labeling process, and quality standards should be determined according to population characteristics, clinical use, and research purpose.

Determinants of non-attendance at face-to-face and telemedicine ophthalmic consultations.

BACKGROUND/AIMS: Evaluation of telemedicine care models has highlighted its potential for exacerbating healthcare inequalities. This study seeks to identify and characterise factors associated with non-attendance across face-to-face and telemedicine outpatient appointments. METHODS: A retrospective cohort study at a tertiary-level ophthalmic institution in the UK, between 1 January 2019 and 31 October 2021. Logistic regression modelled non-attendance against sociodemographic, clinical and operational exposure variables for all new patient registrations across five delivery modes: asynchronous, synchronous telephone, synchronous audiovisual and face to face prior to the pandemic and face to face during the pandemic. RESULTS: A total of 85 924 patients (median age 55 years, 54.4% female) were newly registered. Non-attendance differed significantly by delivery mode: (9.0% face to face prepandemic, 10.5% face to face during the pandemic, 11.7% asynchronous and 7.8%, synchronous during pandemic). Male sex, greater levels of deprivation, a previously cancelled appointment and not self-reporting ethnicity were strongly associated with non-attendance across all delivery modes. Individuals identifying as black ethnicity had worse attendance in synchronous audiovisual clinics (adjusted OR 4.24, 95% CI 1.59 to 11.28) but not asynchronous. Those not self-reporting their ethnicity were from more deprived backgrounds, had worse broadband access and had significantly higher non-attendance across all modes (all p<0.001). CONCLUSION: Persistent non-attendance among underserved populations attending telemedicine appointments highlights the challenge digital transformation faces for reducing healthcare inequalities. Implementation of new programmes should be accompanied by investigation into the differential health outcomes of vulnerable populations.