Search

Abstract Background The large language model ChatGPT can now accept image input with the GPT4-vision (GPT4V) version. We aimed to compare the performance of GPT4V to pretrained U-Net and vision transformer (ViT) models for the identification of the progression of multiple sclerosis (MS) on magnetic resonance imaging (MRI). Methods Paired coregistered MR images with and without progression were provided as input to ChatGPT4V in a zero-shot experiment to identify radiologic progression. Its performance was compared to pretrained U-Net and ViT models. Accuracy was the primary evaluation metric and 95% confidence interval (CIs) were calculated by bootstrapping. We included 170 patients with MS (50 males, 120 females), aged 21–74 years (mean 42.3), imaged at a single institution from 2019 to 2021, each with 2–5 MRI studies (496 in total). Results One hundred seventy patients were included, 110 for training, 30 for tuning, and 30 for testing; 100 unseen paired images were randomly selected from the test set for evaluation. Both U-Net and ViT had 94% (95% CI: 89–98%) accuracy while GPT4V had 85% (77–91%). GPT4V gave cautious nonanswers in six cases. GPT4V had precision (specificity), recall (sensitivity), and F1 score of 89% (75–93%), 92% (82–98%), 91 (82–97%) compared to 100% (100–100%), 88 (78–96%), and 0.94 (88–98%) for U-Net and 94% (87–100%), 94 (88–100%), and 94 (89–98%) for ViT. Conclusion The performance of GPT4V combined with its accessibility suggests has the potential to impact AI radiology research. However, misclassified cases and overly cautious non-answers confirm that it is not yet ready for clinical use. Relevance statement GPT4V can identify the radiologic progression of MS in a simplified experimental setting. However, GPT4V is not a medical device, and its widespread availability highlights the need for caution and education for lay users, especially those with limited access to expert healthcare. Key Points Without fine-tuning or the need for prior coding experience, GPT4V can perform a zero-shot radiologic change detection task with reasonable accuracy. However, in absolute terms, in a simplified “spot the difference” medical imaging task, GPT4V was inferior to state-of-the-art computer vision methods. GPT4V’s performance metrics were more similar to the ViT than the U-net. This is an exploratory experimental study and GPT4V is not intended for use as a medical device. Graphical Abstract

Abstract Few metrics exist to describe phenotypic diversity within ophthalmic imaging datasets, with researchers often using ethnicity as a surrogate marker for biological variability. 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) and reproduced in a Tanzanian, an Australian, and a Chinese dataset. 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 eight 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. RPS decouples traditional demographic variables 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 .

Abstract Background Neovascular age-related macular degeneration (nAMD) is one of the largest single-disease contributors to hospital outpatient appointments. Challenges in finding the clinical capacity to meet this demand can lead to sight-threatening delays in the macular services that provide treatment. Clinical artificial intelligence (AI) technologies pose one opportunity to rebalance demand and capacity in macular services. However, there is a lack of evidence to guide early-adopters seeking to use AI as a solution to demand-capacity imbalance. This study aims to provide guidance for these early adopters on how AI-enabled macular services may best be implemented by exploring what will influence the outcome of AI implementation and why. Methods Thirty-six semi-structured interviews were conducted with participants. Data were analysed with the Nonadoption, Abandonment, Scale-up, Spread and Sustainability (NASSS) framework to identify factors likely to influence implementation outcomes. These factors and the primary data then underwent a secondary analysis using the Fit between Individuals, Technology and Task (FITT) framework to propose an actionable intervention. Results nAMD treatment should be initiated at face-to-face appointments with clinicians who recommend year-long periods of AI-enabled scheduling of treatments. This aims to maintain or enhance the quality of patient communication, whilst reducing consultation frequency. Appropriately trained photographers should take on the additional roles of inputting retinal imaging into the AI device and overseeing its communication to clinical colleagues, while ophthalmologists assume clinical oversight and consultation roles. Interoperability to facilitate this intervention would best be served by imaging equipment that can send images to the cloud securely for analysis by AI tools. Picture Archiving and Communication Software (PACS) should have the capability to output directly into electronic medical records (EMR) familiar to clinical and administrative staff. Conclusion There are many enablers to implementation and few of the remaining barriers relate directly to the AI technology itself. The proposed intervention requires local tailoring and prospective evaluation but can support early adopters in optimising the chances of success from initial efforts to implement AI-enabled macular services. Protocol registration Hogg HDJ, Brittain K, Teare D, Talks J, Balaskas K, Keane P, Maniatopoulos G. Safety and efficacy of an artificial intelligence-enabled decision tool for treatment decisions in neovascular age-related macular degeneration and an exploration of clinical pathway integration and implementation: protocol for a multi-methods validation study. BMJ Open. 2023 Feb 1;13(2):e069443. https://doi.org/10.1136/bmjopen-2022-069443 . PMID: 36725098; PMCID: PMC9896175.

Abstract Oestradiol and progesterone levels are higher in menstruating women than men of the same age, and their receptors are present in their neurosensory retina and retinal pigment epithelium. However, the impact of this hormonal environment on retinal physiology in women remains unclear. Using self-reported menstrual cycle phases as a surrogate for fluctuating hormonal levels, we investigated associations with retinovascular indices on colour fundus photograph and retinal thickness in optical coherence tomography across regularly menstruating women in the UK Biobank. We found no differences in retinal thickness across the cycle; however, vessel density, arteriolar and venular, and fractal dimension were higher in the luteal phase than follicular. The calibre of the central retinal vessels did not differ. This study suggests that the menstrual cycle phase might be associated with retinal microvasculature density in non-invasive imaging. It raises awareness for this understudied area, providing insights into neuroscience fields and epidemiological studies.

Abstract Introduction This study reports our experiences with systematic retinal screening in Denmark through optometrists with access to tele-ophthalmological services before, during, and after the COVID-19 pandemic. Methods We evaluated an optometrist-based retinal screening system with a referral option for tele-ophthalmological service by a consultant ophthalmologist within the time period of August 1, 2018 to September 30, 2023. The optometrist collected patient history, refraction, best-corrected visual acuity, intraocular pressure, basic slit-lamp examination, 4-in-1 visual field report, and retinal imaging using color fundus 45° photography. Tele-ophthalmological services were provided by consultant ophthalmologists. Within pre-defined periods of pre-COVID-19, COVID-19, and post-COVID-19, we evaluated the rate of referrals to the tele-ophthalmological service, diagnoses made, and referrals to the public healthcare system. Results A total of 1,142,028 unique individuals, which corresponded to 19.1% of the entire population of Denmark, underwent screening by the optometrists; 50,612 (4.4%) of these individuals were referred to the tele-ophthalmological examination by consultant ophthalmologists. A referral for further ophthalmic examination, either at hospital or at an ophthalmic practice, was made for 10,300 individuals (20.4% of those referred for tele-ophthalmology, corresponding to 0.9% of the population screened). The referral rate from the screening to the tele-ophthalmological service increased from before COVID-19 (3.4%) to during COVID-19 (4.3%) and further after COVID-19 (6.4%). This increase coincided with an increasing prevalence of conditions seen in the tele-ophthalmological service. Conclusion During a period of 5 years, 19.1% of the entire population of Denmark underwent retinal screening. This provided an adjunctive health service during a period of severe strain on the public healthcare system, while limiting the number of excessive referrals to the public healthcare system. Temporal trends illustrated an increased pattern of use of a large-scale tele-ophthalmological system.

Purpose: The region of growth (ROG) of geographic atrophy (GA) throughout the macular area has an impact on visual outcomes. Here, we developed multiple deep learning models to predict the 1-year ROG of GA lesions using fundus autofluorescence (FAF) images. Design: In this retrospective analysis, 3 types of models were developed using FAF images collected 6 months after baseline to predict the GA lesion area (segmented lesion mask) at 1.5 years, FAF images collected at baseline and 6 months to predict the GA lesion at 1.5 years, and FAF images collected 6 months after baseline to predict the GA lesion at 1 and 1.5 years. The 1-year ROG from the 6-month visit was derived by taking the difference between the GA lesion area (segmented lesion mask) at the 1.5-year and 6-month visits. Participants: Patients enrolled in the following lampalizumab clinical trials and prospective observational studies: NCT02247479, NCT02247531, NCT02479386, and NCT02399072. Methods: Datasets of study eyes from 597 patients were split into model training (310), validation (78), and test sets (209), stratified by baseline or initial lesion area, lesion growth rate, foveal involvement, and focality. Deep learning experiments were performed using the 2-dimensional U-Net; whole-lesion and multiclass models were developed. Main Outcome Measures: The performance of the models was evaluated by calculating the Dice score, coefficient of determination (R2), and the squared Pearson correlation coefficient (r2) between the true and derived GA lesion 1-year ROG. Results: The model using baseline and 6-month FAF images to predict GA lesion enlargement at 1.5 years had the best performance for the derived 1-year ROG. Mean Dice scores were 0.73, 0.68, and 0.70 in the training, validation, and test sets, respectively. The R2 (0.77, 0.53, and 0.79) and r2 (0.83, 0.61, and 0.79) showed similar trends across the 3 sets. Conclusions: These findings show the potential of using baseline and/or 6-month visit FAF images to predict 1-year GA ROG using a deep learning approach. This work could potentially help support decision-making in clinical trials and more informed treatment decisions in clinical practice. Financial Disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Purpose: Application of artificial intelligence (AI) to macular OCT scans to segment and quantify volumetric change in anatomical and pathological features during intravitreal treatment for neovascular age-related macular degeneration (AMD). Design: Retrospective analysis of OCT images from the Moorfields Eye Hospital AMD Database. Participants: A total of 2115 eyes from 1801 patients starting anti-VEGF treatment between June 1, 2012, and June 30, 2017. Methods: The Moorfields Eye Hospital neovascular AMD database was queried for first and second eyes receiving anti-VEGF treatment and had an OCT scan at baseline and 12 months. Follow-up scans were input into the AI system and volumes of OCT variables were studied at different time points and compared with baseline volume groups. Cross-sectional comparisons between time points were conducted using Mann–Whitney U test. Main Outcome Measures: Volume outputs of the following variables were studied: intraretinal fluid, subretinal fluid, pigment epithelial detachment (PED), subretinal hyperreflective material (SHRM), hyperreflective foci, neurosensory retina, and retinal pigment epithelium. Results: Mean volumes of analyzed features decreased significantly from baseline to both 4 and 12 months, in both first-treated and second-treated eyes. Pathological features that reflect exudation, including pure fluid components (intraretinal fluid and subretinal fluid) and those with fluid and fibrovascular tissue (PED and SHRM), displayed similar responses to treatment over 12 months. Mean PED and SHRM volumes showed less pronounced but also substantial decreases over the first 2 months, reaching a plateau postloading phase, and minimal change to 12 months. Both neurosensory retina and retinal pigment epithelium volumes showed gradual reductions over time, and were not as substantial as exudative features. Conclusions: We report the results of a quantitative analysis of change in retinal segmented features over time, enabled by an AI segmentation system. Cross-sectional analysis at multiple time points demonstrated significant associations between baseline OCT-derived segmented features and the volume of biomarkers at follow-up. Demonstrating how certain OCT biomarkers progress with treatment and the impact of pretreatment retinal morphology on different structural volumes may provide novel insights into disease mechanisms and aid the personalization of care. Data will be made public for future studies. Financial Disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Purpose: To evaluate the performance of a disease activity (DA) model developed to detect DA in participants with neovascular age-related macular degeneration (nAMD). Design: Post hoc analysis. Participants: Patient dataset from the phase III HAWK and HARRIER (H&H) studies. Methods: An artificial intelligence (AI)-based DA model was developed to generate a DA score based on measurements of OCT images and other parameters collected from H&H study participants. Disease activity assessments were classified into 3 categories based on the extent of agreement between the DA model’s scores and the H&H investigators’ decisions: agreement (“easy”), disagreement (“noisy”), and close to the decision boundary (“difficult”). Then, a panel of 10 international retina specialists (“panelists”) reviewed a sample of DA assessments of these 3 categories that contributed to the training of the final DA model. A panelists’ majority vote on the reviewed cases was used to evaluate the accuracy, sensitivity, and specificity of the DA model. Main Outcome Measures: The DA model’s performance in detecting DA compared with the DA assessments made by the investigators and panelists’ majority vote. Results: A total of 4472 OCT DA assessments were used to develop the model; of these, panelists reviewed 425, categorized as “easy” (17.2%), “noisy” (20.5%), and “difficult” (62.4%). False-positive and false negative rates of the DA model’s assessments decreased after changing the assessment in some cases reviewed by the panelists and retraining the DA model. Overall, the DA model achieved 80% accuracy. For “easy” cases, the DA model reached 96% accuracy and performed as well as the investigators (96% accuracy) and panelists (90% accuracy). For “noisy” cases, the DA model performed similarly to panelists and outperformed the investigators (84%, 86%, and 16% accuracies, respectively). The DA model also outperformed the investigators for “difficult” cases (74% and 53% accuracies, respectively) but underperformed the panelists (86% accuracy) owing to lower specificity. Subretinal and intraretinal fluids were the main clinical parameters driving the DA assessments made by the panelists. Conclusions: These results demonstrate the potential of using an AI-based DA model to optimize treatment decisions in the clinical setting and in detecting and monitoring DA in patients with nAMD. Financial Disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

BackgroundDiabetic retinopathy (DR) affects about 25% of people with diabetes in Canada. Early detection of DR is essential for preventing vision loss. ObjectiveWe evaluated the real-world performance of an artificial intelligence (AI) system that analyzes fundus images for DR screening in a Quebec tertiary care center. MethodsWe prospectively recruited adult patients with diabetes at the Centre hospitalier de l’Université de Montréal (CHUM) in Montreal, Quebec, Canada. Patients underwent dual-pathway screening: first by the Computer Assisted Retinal Analysis (CARA) AI system (index test), then by standard ophthalmological examination (reference standard). We measured the AI system's sensitivity and specificity for detecting referable disease at the patient level, along with its performance for detecting any retinopathy and diabetic macular edema (DME) at the eye level, and potential cost savings. ResultsThis study included 115 patients. CARA demonstrated a sensitivity of 87.5% (95% CI 71.9-95.0) and specificity of 66.2% (95% CI 54.3-76.3) for detecting referable disease at the patient level. For any retinopathy detection at the eye level, CARA showed 88.2% sensitivity (95% CI 76.6-94.5) and 71.4% specificity (95% CI 63.7-78.1). For DME detection, CARA had 100% sensitivity (95% CI 64.6-100) and 81.9% specificity (95% CI 75.6-86.8). Potential yearly savings from implementing CARA at the CHUM were estimated at CAD $245,635 (US $177,643.23, as of July 26, 2024) considering 5000 patients with diabetes. ConclusionsOur study indicates that integrating a semiautomated AI system for DR screening demonstrates high sensitivity for detecting referable disease in a real-world setting. This system has the potential to improve screening efficiency and reduce costs at the CHUM, but more work is needed to validate it.

Introduction: Automated machine learning (AutoML) is a novel artificial intelligence (AI) strategy that enables clinicians without coding experience to develop their own AI models. This study assessed the discriminative performance of AutoML in differentiating diabetic retinopathy (DR), central retinal vein occlusion (CRVO) and branch retinal vein occlusion (BRVO) from normal fundi using color fundus photographs (CFPs). Methods: We carried out AutoML model design using CFPs retrieved from a publicly available CFP data set (3200 labelled images). The retrieved CFPs were reviewed for quality and then uploaded to the Google Cloud Vertex AI platform for AutoML training and testing. We trained a multi-class classification model to differentiate DR, CRVO, BRVO from normal fundi using 875 CFPs and externally validated the model using 210 CFPs obtained from another dataset. Performance metrics, including area under receiver operator curve (AUROC) and sensitivity were reported. We then compared the AutoML model to state-of-the-art deep learning (DL)-based DR and RVO multi-class models identified through a literature search. Results: Our AutoML model showed high discriminative performance in the multi-class classification of DR, CRVO and BRVO based on CFPs, with an AUROC, precision and recall reaching 0.995, 95.4% and 95.4% respectively at the 0.5 confidence threshold. The per-label sensitivity and specificity, respectively, were normal fundi (97.5%, 100%), DR (100%, 93.88%), CRVO (66.67%, 100%) and BRVO (71.43%, 98.73%). Our AutoML model generally showed similar performance to the state-of-the-art DL classifiers. Conclusion: Our AutoML model can detect DR, CRVO, and BRVO in CFPs with good diagnostic accuracy and is a potentially useful screening tool.

Abstract 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 =

Abstract 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.

Abstract Retinal thickness may serve as a biomarker in Parkinson’s disease (PD). In this prospective longitudinal study, we aimed to determine if PD patients present accelerated thinning rate in the parafoveal ganglion cell-inner plexiform layer (pfGCIPL) and peripapillary retinal nerve fiber layer (pRNFL) compared to controls. Additionally, we evaluated the relationship between retinal neurodegeneration and clinical progression in PD. A cohort of 156 PD patients and 72 controls underwent retinal optical coherence tomography, visual, and cognitive assessments between February 2015 and December 2021 in two Spanish tertiary hospitals. The pfGCIPL thinning rate was twice as high in PD (β [SE] = −0.58 [0.06]) than in controls (β [SE] = −0.29 [0.06], p

Abstract Introduction To evaluate the effect pegcetacoplan, a C3 and C3b inhibitor, on the rate of progression of geographic atrophy (GA) as assessed by spectral domain optical coherence tomography (SD-OCT) using a split-person study design and deep-learning quantification. Methods A post hoc analysis of phase 2 FILLY trial data comparing study (treated monthly, treated every other month and sham-treated) and fellow (untreated) eyes in a split-person study design was performed. This analysis included 288 eyes from 144 patients with bilateral GA from the FILLY phase 2 trial (Clinical Trials identifier: NCT02503332). Only patients with bilateral GA and without evidence of choroidal neovascularisation in either eye were included. Patient study eyes were treated with sham injections or with pegcetacoplan monthly (PM) or every other month (PEOM) for 12 months. SD-OCT scans of study and fellow eyes taken at baseline and 12 months were used for the analysis. The main outcomes were the annual change in the area of retinal pigment epithelial and outer retinal atrophy (RORA), its constituent features (photoreceptor degeneration [PRD], retinal pigment epithelium [RPE] loss, hypertransmission) and intact macula as compared to the untreated fellow eye. Results Annual GA growth was reduced in eyes treated with PM versus untreated fellow eyes for OCT features, including RORA (study eye 0.792 vs. fellow eye 1.13 mm2; P = 0.003), PRD (0.739 vs. 1.23 mm2; P = 0.015), RPE-loss (0.789 vs. 1.17 mm2; P = 0.007) and intact macula (− 0.735 vs. − 1.29 mm2; P = 0.011). Similar (but not statistically significant) trends were observed with the PEOM treatment or when GA was quantified with fundus autofluorescence (FAF). The sham treatment demonstrated no effect. Pearson correlation coefficients showed concordance in the enlargement rate of GA between the study and fellow eyes in the sham (R = 0.64) and PEOM (R = 0.68) groups, but not in the PM group (R = 0.21). Conclusions Pegcetacoplan-treated eyes demonstrated a reduction in spatial GA progression compared to their untreated counterparts. This effect was more evident on OCT than with FAF. Trial Registration Clinical Trials identifier: NCT02503332.

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 Disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

BackgroundArtificial 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. ObjectiveThis 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. MethodsThe 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. ResultsPreliminary 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. ConclusionsThis 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

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.

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.

Summary: Background: A substantial proportion of attendances to ophthalmic emergency departments are for non-urgent presentations. We developed and evaluated a machine learning system (DemDx Ophthalmology Triage System: DOTS) to optimise triage, with the aim of reducing inappropriate emergency attendances and streamlining case referral when necessary. Methods: DOTS was built using retrospective tabular data from 11,315 attendances between July 1st, 2021, to June 15th, 2022 at Moorfields Eye Hospital Emergency Department (MEH) in London, UK. Demographic and clinical features were used as inputs and a triage recommendation was given (“see immediately”, “see within a week”, or “see electively”). DOTS was validated temporally and compared with triage nurses’ performance (1269 attendances at MEH) and validated externally (761 attendances at the Federal University of Minas Gerais - UFMG, Brazil). It was also tested for biases and robustness to variations in disease incidences. All attendances from patients aged at least 18 years with at least one confirmed diagnosis were included in the study. Findings: For identifying ophthalmic emergency attendances, on temporal validation, DOTS had a sensitivity of 94.5% [95% CI 92.3–96.1] and a specificity of 42.4% [38.8–46.1]. For comparison within the same dataset, triage nurses had a sensitivity of 96.4% [94.5–97.7] and a specificity of 25.1% [22.0–28.5]. On external validation at UFMG, DOTS had a sensitivity of 95.2% [92.5–97.0] and a specificity of 32.2% [27.4–37.0]. In simulated scenarios with varying disease incidences, the sensitivity was ≥92.2% and the specificity was ≥36.8%. No differences in sensitivity were found in subgroups of index of multiple deprivation, but the specificity was higher for Q2 when compared to Q4 (Q4 is less deprived than Q2). Interpretation: At MEH, DOTS had similar sensitivity to triage nurses in determining attendance priority; however, with a specificity of 17.3% higher, DOTS resulted in lower rates of patients triaged to be seen immediately at emergency. DOTS showed consistent performance in temporal and external validation, in social-demographic subgroups and was robust to varying relative disease incidences. Further trials are necessary to validate these findings. This system will be prospectively evaluated, considering human-computer interaction, in a clinical trial. Funding: The Artificial Intelligence in Health and Care Award (AI_AWARD01671) of the NHS AI Lab under National Institute for Health and Care Research (NIHR) and the Accelerated Access Collaborative (AAC).

The rapid progress of large language models (LLMs) driving generative artificial intelligence applications heralds the potential of opportunities in health care. We conducted a review up to April 2023 on Google Scholar, Embase, MEDLINE, and Scopus using the following terms: “large language models,” “generative artificial intelligence,” “ophthalmology,” “ChatGPT,” and “eye,” based on relevance to this review. From a clinical viewpoint specific to ophthalmologists, we explore from the different stakeholders’ perspectives—including patients, physicians, and policymakers—the potential LLM applications in education, research, and clinical domains specific to ophthalmology. We also highlight the foreseeable challenges of LLM implementation into clinical practice, including the concerns of accuracy, interpretability, perpetuating bias, and data security. As LLMs continue to mature, it is essential for stakeholders to jointly establish standards for best practices to safeguard patient safety. Financial Disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Abstract Background The aim of this study was to describe features of disease activity in patients with treated stable macular neovascularisation (MNV) in neovascular age related macular degeneration (nAMD) using optical coherence tomography angiography (OCTA). Methods Thirty-two eyes of 32 patients with nAMD were included in this prospective, observational study. These patients were undergoing treatment with aflibercept on a treat-and-extend regimen attending an extension to a 12-week treatment interval. Results All subjects had no macular haemorrhage and no structural OCT markers of active MNV activity at the index 12-week treatment extension visit. 31/32 OCTA images were gradeable without significant imaging artefact. The mean MNV size was 3.6mm2 ± 4.6mm2 and 27 (87.1%) had detectable MNV blood flow. 29/31 (93.5%) subjects had MNV with mature phenotypes including 10 non-specific, 10 tangle and 3 deadtree phenotypes. MNV halo and MNV central feeder vessel were noted in 18 (58.1%) and 19 (61.3%) of subjects respectively; only 1 (3.2%) subject was noted to have a MNV capillary fringe. Conclusions MNV blood flow is still detectable using OCTA in the majority of subjects in this study with treated stable MNV. OCTA features associated included MNV mature phenotype, MNV feeder vessel, MNV halo and absence of capillary fringe.

Deep learning–based clinical imaging analysis underlies diagnostic artificial intelligence (AI) models, which can match or even exceed the performance of clinical experts, having the potential to revolutionize clinical practice. A wide variety of automated machine learning (autoML) platforms lower the technical barrier to entry to deep learning, extending AI capabilities to clinicians with limited technical expertise, and even autonomous foundation models such as multimodal large language models. Here, we provide a technical overview of autoML with descriptions of how autoML may be applied in education, research, and clinical practice. Each stage of the process of conducting an autoML project is outlined, with an emphasis on ethical and technical best practices. Specifically, data acquisition, data partitioning, model training, model validation, analysis, and model deployment are considered. The strengths and limitations of available code-free, code-minimal, and code-intensive autoML platforms are considered. AutoML has great potential to democratize AI in medicine, improving AI literacy by enabling “hands-on” education. AutoML may serve as a useful adjunct in research by facilitating rapid testing and benchmarking before significant computational resources are committed. AutoML may also be applied in clinical contexts, provided regulatory requirements are met. The abstraction by autoML of arduous aspects of AI engineering promotes prioritization of data set curation, supporting the transition from conventional model-driven approaches to data-centric development. To fulfill its potential, clinicians must be educated on how to apply these technologies ethically, rigorously, and effectively; this tutorial represents a comprehensive summary of relevant considerations.

Purpose: Diabetic retinopathy (DR) is the most common microvascular complication associated with diabetes mellitus (DM), affecting approximately 40% of this patient population. Early detection of DR is vital to ensure monitoring of disease progression and prompt sight saving treatments as required. This article describes the data contained within the INSIGHT Birmingham, Solihull, and Black Country Diabetic Retinopathy Dataset. Design: Dataset descriptor for routinely collected eye screening data. Participants: All diabetic patients aged 12 years and older, attending annual digital retinal photography-based screening within the Birmingham, Solihull, and Black Country Eye Screening Programme. Methods: The INSIGHT Health Data Research Hub for Eye Health is a National Health Service (NHS)–led ophthalmic bioresource that provides researchers with safe access to anonymized, routinely collected data from contributing NHS hospitals to advance research for patient benefit. This report describes the INSIGHT Birmingham, Solihull, and Black Country DR Screening Dataset, a dataset of anonymized images and linked screening data derived from the United Kingdom’s largest regional DR screening program. Main Outcome Measures: This dataset consists of routinely collected data from the eye screening program. The data primarily include retinal photographs with the associated DR grading data. Additional data such as corresponding demographic details, information regarding patients’ diabetic status, and visual acuity data are also available. Further details regarding available data points are available in the supplementary information, in addition to the INSIGHT webpage included below. Results: At the time point of this analysis (December 31, 2019), the dataset comprised 6 202 161 images from 246 180 patients, with a dataset inception date of January 1, 2007. The dataset includes 1 360 547 grading episodes between R0M0 and R3M1. Conclusions: This dataset descriptor article summarizes the content of the dataset, how it has been curated, and what its potential uses are. Data are available through a structured application process for research studies that support discovery, clinical evidence analyses, and innovation in artificial intelligence technologies for patient benefit. Further information regarding the data repository and contact details can be found at https://www.insight.hdrhub.org/. Financial Disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Purpose: Rare disease diagnosis is challenging in medical image-based artificial intelligence due to a natural class imbalance in datasets, leading to biased prediction models. Inherited retinal diseases (IRDs) are a research domain that particularly faces this issue. This study investigates the applicability of synthetic data in improving artificial intelligence-enabled diagnosis of IRDs using generative adversarial networks (GANs). Design: Diagnostic study of gene-labeled fundus autofluorescence (FAF) IRD images using deep learning. Participants: Moorfields Eye Hospital (MEH) dataset of 15 692 FAF images obtained from 1800 patients with confirmed genetic diagnosis of 1 of 36 IRD genes. Methods: A StyleGAN2 model is trained on the IRD dataset to generate 512 × 512 resolution images. Convolutional neural networks are trained for classification using different synthetically augmented datasets, including real IRD images plus 1800 and 3600 synthetic images, and a fully rebalanced dataset. We also perform an experiment with only synthetic data. All models are compared against a baseline convolutional neural network trained only on real data. Main Outcome Measures: We evaluated synthetic data quality using a Visual Turing Test conducted with 4 ophthalmologists from MEH. Synthetic and real images were compared using feature space visualization, similarity analysis to detect memorized images, and Blind/Referenceless Image Spatial Quality Evaluator (BRISQUE) score for no-reference-based quality evaluation. Convolutional neural network diagnostic performance was determined on a held-out test set using the area under the receiver operating characteristic curve (AUROC) and Cohen’s Kappa (κ). Results: An average true recognition rate of 63% and fake recognition rate of 47% was obtained from the Visual Turing Test. Thus, a considerable proportion of the synthetic images were classified as real by clinical experts. Similarity analysis showed that the synthetic images were not copies of the real images, indicating that copied real images, meaning the GAN was able to generalize. However, BRISQUE score analysis indicated that synthetic images were of significantly lower quality overall than real images (P < 0.05). Comparing the rebalanced model (RB) with the baseline (R), no significant change in the average AUROC and κ was found (R-AUROC = 0.86[0.85-88], RB-AUROC = 0.88[0.86-0.89], R-κ = 0.51[0.49-0.53], and RB-κ = 0.52[0.50-0.54]). The synthetic data trained model (S) achieved similar performance as the baseline (S-AUROC = 0.86[0.85-87], S-κ = 0.48[0.46-0.50]). Conclusions: Synthetic generation of realistic IRD FAF images is feasible. Synthetic data augmentation does not deliver improvements in classification performance. However, synthetic data alone deliver a similar performance as real data, and hence may be useful as a proxy to real data.Financial Disclosure(s): Proprietary or commercial disclosure may be found after the references.

Background: The glymphatic system removes neurodegenerative debris. The ocular glymphatic outflow is from the eye to the proximal optic nerve. In multiple sclerosis (MS), atrophy of the optic nerve increases the glymphatic outflow space. Here, we tested whether vitreous haze (VH) can provide novel insights into the relationship between neurodegeneration and the ocular glymphatic system in MS. Methods: This cross-sectional study comprised 315 persons with MS and 87 healthy controls (HCs). VH was quantified from optical coherence tomography (OCT) volume scans. Neurodegeneration was determined on three-dimensional T1 (3DT1) MRI, lesion detection on fluid-attenuated inversion (FLAIR), and layer thickness on OCT. Generalized estimating equations, corrected for age, were used to analyze associations between VH and metrics for neurodegeneration, demographics, and clinical scales. Group differences were determined between mild, moderate, and severe disability. Results: On the group level, VH scores were comparable between MS and control (p = 0.629). In MS, VH scores declined with disease duration (β = −0.009, p = 0.004) and age (β = −0.007, p = 0.001). There was no relation between VH scores and higher age in HCs. In MS patients, VH was related to normalized gray (NGMV, β = 0.001, p = 0.011) and white matter volume (NWMV, β = 0.001, p = 0.003), macular ganglion cell–inner plexiform layer thickness (mGCIPL, β = 0.006, p < 0.001), and peripapillary retinal nerve fiber layer thickness (pRNFL, β = 0.004, p = 0.008). VH was significantly lower in severe compared to mild disability (mean difference −28.86%, p = 0.058). Conclusions: There is a correlation between VH on OCT and disease duration, more severe disability and lower brain volumes in MS. Biologically, these relationships suggest accelerated glymphatic clearance with disease-related atrophy.

BackgroundThe rhetoric surrounding clinical artificial intelligence (AI) often exaggerates its effect on real-world care. Limited understanding of the factors that influence its implementation can perpetuate this. ObjectiveIn this qualitative systematic review, we aimed to identify key stakeholders, consolidate their perspectives on clinical AI implementation, and characterize the evidence gaps that future qualitative research should target. MethodsOvid-MEDLINE, EBSCO-CINAHL, ACM Digital Library, Science Citation Index-Web of Science, and Scopus were searched for primary qualitative studies on individuals’ perspectives on any application of clinical AI worldwide (January 2014-April 2021). The definition of clinical AI includes both rule-based and machine learning–enabled or non–rule-based decision support tools. The language of the reports was not an exclusion criterion. Two independent reviewers performed title, abstract, and full-text screening with a third arbiter of disagreement. Two reviewers assigned the Joanna Briggs Institute 10-point checklist for qualitative research scores for each study. A single reviewer extracted free-text data relevant to clinical AI implementation, noting the stakeholders contributing to each excerpt. The best-fit framework synthesis used the Nonadoption, Abandonment, Scale-up, Spread, and Sustainability (NASSS) framework. To validate the data and improve accessibility, coauthors representing each emergent stakeholder group codeveloped summaries of the factors most relevant to their respective groups. ResultsThe initial search yielded 4437 deduplicated articles, with 111 (2.5%) eligible for inclusion (median Joanna Briggs Institute 10-point checklist for qualitative research score, 8/10). Five distinct stakeholder groups emerged from the data: health care professionals (HCPs), patients, carers and other members of the public, developers, health care managers and leaders, and regulators or policy makers, contributing 1204 (70%), 196 (11.4%), 133 (7.7%), 129 (7.5%), and 59 (3.4%) of 1721 eligible excerpts, respectively. All stakeholder groups independently identified a breadth of implementation factors, with each producing data that were mapped between 17 and 24 of the 27 adapted Nonadoption, Abandonment, Scale-up, Spread, and Sustainability subdomains. Most of the factors that stakeholders found influential in the implementation of rule-based clinical AI also applied to non–rule-based clinical AI, with the exception of intellectual property, regulation, and sociocultural attitudes. ConclusionsClinical AI implementation is influenced by many interdependent factors, which are in turn influenced by at least 5 distinct stakeholder groups. This implies that effective research and practice of clinical AI implementation should consider multiple stakeholder perspectives. The current underrepresentation of perspectives from stakeholders other than HCPs in the literature may limit the anticipation and management of the factors that influence successful clinical AI implementation. Future research should not only widen the representation of tools and contexts in qualitative research but also specifically investigate the perspectives of all stakeholder HCPs and emerging aspects of non–rule-based clinical AI implementation. Trial RegistrationPROSPERO (International Prospective Register of Systematic Reviews) CRD42021256005; https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=256005 International Registered Report Identifier (IRRID)RR2-10.2196/33145

Purpose: To describe a novel optical coherence tomography (OCT) finding at the vitreomacular interface (VMI), and report its association with advanced choroidal neovascularisation (CNV). Observations: Optical coherence tomography (OCT) scans performed at three retinal imaging centres at Amanat Eye Hospital, Pakistan from May 2016 till May 2021 were reviewed. A specific change at the vitreomacular interface was noted consisting of abnormal hyper reflectivity at the point of attachment of the posterior hyaloid membrane to the foveal center which appears to ‘fill in’ the foveolar depression.Eight eyes of eight patients were identified. All affected eyes had advanced CNV and persistent vitreofoveolar adhesion. In all eyes, the foveal contour (concavity) was maintained and there was no inner retinal surface wrinkling which differentiates this OCT feature from vitreomacular traction or epiretinal membranes. The authors propose the term Central Posterior Hyaloidal Fibrosis (CPHF) for this specific OCT finding. Conclusions and Importance: Central Posterior Hyaloidal Fibrosis (CPHF) is a newly reported OCT finding associated with advanced CNV, which may represent a possible profibrotic influence of a choroidal neovascular membrane to the overlying posterior hyaloid adhesion.

BackgroundMany artificial intelligence (AI) studies have focused on development of AI models, novel techniques, and reporting guidelines. However, little is understood about clinicians' perspectives of AI applications in medical fields including ophthalmology, particularly in light of recent regulatory guidelines. The aim for this study was to evaluate the perspectives of ophthalmologists regarding AI in 4 major eye conditions: diabetic retinopathy (DR), glaucoma, age-related macular degeneration (AMD) and cataract.MethodsThis was a multi-national survey of ophthalmologists between March 1st, 2020 to February 29th, 2021 disseminated via the major global ophthalmology societies. The survey was designed based on microsystem, mesosystem and macrosystem questions, and the software as a medical device (SaMD) regulatory framework chaired by the Food and Drug Administration (FDA). Factors associated with AI adoption for ophthalmology analyzed with multivariable logistic regression random forest machine learning.ResultsOne thousand one hundred seventy-six ophthalmologists from 70 countries participated with a response rate ranging from 78.8 to 85.8% per question. Ophthalmologists were more willing to use AI as clinical assistive tools (88.1%, n = 890/1,010) especially those with over 20 years' experience (OR 3.70, 95% CI: 1.10–12.5, p = 0.035), as compared to clinical decision support tools (78.8%, n = 796/1,010) or diagnostic tools (64.5%, n = 651). A majority of Ophthalmologists felt that AI is most relevant to DR (78.2%), followed by glaucoma (70.7%), AMD (66.8%), and cataract (51.4%) detection. Many participants were confident their roles will not be replaced (68.2%, n = 632/927), and felt COVID-19 catalyzed willingness to adopt AI (80.9%, n = 750/927). Common barriers to implementation include medical liability from errors (72.5%, n = 672/927) whereas enablers include improving access (94.5%, n = 876/927). Machine learning modeling predicted acceptance from participant demographics with moderate to high accuracy, and area under the receiver operating curves of 0.63–0.83.ConclusionOphthalmologists are receptive to adopting AI as assistive tools for DR, glaucoma, and AMD. Furthermore, ML is a useful method that can be applied to evaluate predictive factors on clinical qualitative questionnaires. This study outlines actionable insights for future research and facilitation interventions to drive adoption and operationalization of AI tools for Ophthalmology.

Objective To train and validate a code-free deep learning system (CFDLS) on classifying high-resolution digital retroillumination images of posterior capsule opacification (PCO) and to discriminate between clinically significant and non-significant PCOs.Methods and analysis For this retrospective registry study, three expert observers graded two independent datasets of 279 images three separate times with no PCO to severe PCO, providing binary labels for clinical significance. The CFDLS was trained and internally validated using 179 images of a training dataset and externally validated with 100 images. Model development was through Google Cloud AutoML Vision. Intraobserver and interobserver variabilities were assessed using Fleiss kappa (κ) coefficients and model performance through sensitivity, specificity and area under the curve (AUC).Results Intraobserver variability κ values for observers 1, 2 and 3 were 0.90 (95% CI 0.86 to 0.95), 0.94 (95% CI 0.90 to 0.97) and 0.88 (95% CI 0.82 to 0.93). Interobserver agreement was high, ranging from 0.85 (95% CI 0.79 to 0.90) between observers 1 and 2 to 0.90 (95% CI 0.85 to 0.94) for observers 1 and 3. On internal validation, the AUC of the CFDLS was 0.99 (95% CI 0.92 to 1.0); sensitivity was 0.89 at a specificity of 1. On external validation, the AUC was 0.97 (95% CI 0.93 to 0.99); sensitivity was 0.84 and specificity was 0.92.Conclusion This CFDLS provides highly accurate discrimination between clinically significant and non-significant PCO equivalent to human expert graders. The clinical value as a potential decision support tool in different models of care warrants further research.

Background Recent clinical trials on proliferative diabetic retinopathy (PDR) show that presenting visual acuity can be stabilised with panretinal photocoagulation (PRP) within 2 years despite the need for supplementary PRP or anti-vascular endothelial growth factor therapy for concomitant diabetic macular oedema (DMO). It is unclear whether similar results can be obtained in daily clinical practice. Here, we query the probability of vision loss in patients with treatment-naïve PDR who have attained stability after PRP and its predictors.Methods Retrospective cohort study at a tertiary eye centre between 01 January 2015 and 31 December 2019, wherein 2336 eyes met study criteria with first record of stable PRP-treated PDR in at least one eye. Kaplan-Meier and Cox proportional hazards modelling were used to report the probability of vision loss of at least five Early Treatment Diabetic Retinopathy Study (ETDRS) letters.Results The probability of losing at least five ETDRS letters was 50% at 3.32 (95% CI, 2.94 to 3.78) years after achieving first stability post PRP in treatment-naïve PDR. The mean decrease at this event was 14.2 (SD 13.0) ETDRS letters irrespective of the presence of DMO. The strongest risk factor for vision loss was a history of DMO at baseline (HR 1.62 (95% CI, 1.34 to 1.95), p

Abstract Artificial intelligence (AI)‐based diagnostic algorithms have achieved ambitious aims through automated image pattern recognition. For neurological disorders, this includes neurodegeneration and inflammation. Scalable imaging technology for big data in neurology is optical coherence tomography (OCT). We highlight that OCT changes observed in the retina, as a window to the brain, are small, requiring rigorous quality control pipelines. There are existing tools for this purpose. Firstly, there are human‐led validated consensus quality control criteria (OSCAR‐IB) for OCT. Secondly, these criteria are embedded into OCT reporting guidelines (APOSTEL). The use of the described annotation of failed OCT scans advances machine learning. This is illustrated through the present review of the advantages and disadvantages of AI‐based applications to OCT data. The neurological conditions reviewed here for the use of big data include Alzheimer disease, stroke, multiple sclerosis (MS), Parkinson disease, and epilepsy. It is noted that while big data is relevant for AI, ownership is complex. For this reason, we also reached out to involve representatives from patient organizations and the public domain in addition to clinical and research centers. The evidence reviewed can be grouped in a five‐point expansion of the OSCAR‐IB criteria to embrace AI (OSCAR‐AI). The review concludes by specific recommendations on how this can be achieved practically and in compliance with existing guidelines.

Abstract Deep learning may transform health care, but model development has largely been dependent on availability of advanced technical expertise. Herein we present the development of a deep learning model by clinicians without coding, which predicts reported sex from retinal fundus photographs. A model was trained on 84,743 retinal fundus photos from the UK Biobank dataset. External validation was performed on 252 fundus photos from a tertiary ophthalmic referral center. For internal validation, the area under the receiver operating characteristic curve (AUROC) of the code free deep learning (CFDL) model was 0.93. Sensitivity, specificity, positive predictive value (PPV) and accuracy (ACC) were 88.8%, 83.6%, 87.3% and 86.5%, and for external validation were 83.9%, 72.2%, 78.2% and 78.6% respectively. Clinicians are currently unaware of distinct retinal feature variations between males and females, highlighting the importance of model explainability for this task. The model performed significantly worse when foveal pathology was present in the external validation dataset, ACC: 69.4%, compared to 85.4% in healthy eyes, suggesting the fovea is a salient region for model performance OR (95% CI): 0.36 (0.19, 0.70) p = 0.0022. Automated machine learning (AutoML) may enable clinician-driven automated discovery of novel insights and disease biomarkers.

Plants display a range of temporal patterns of inter-annual reproduction, from relatively constant seed production to mast seeding, the synchronized and highly variable interannual seed production of plants within a population. Previous efforts have compiled global records of seed production in long-lived plants to gain insight into seed production, forest and animal population dynamics, and the effects of global change on masting. Existing datasets focus on seed production dynamics at the population scale but are limited in their ability to examine community-level mast seeding dynamics across different plant species at the continental scale. We harmonized decades of plant reproduction data for 141 woody plant species across nine Long-Term Ecological Research (LTER) or long-term ecological monitoring sites from a wide range of habitats across the United States. Plant reproduction data are reported annually between 1957 and 2021 and based on either seed traps or seed and/or cone counts on individual trees. A wide range of woody plant species including trees, shrubs, and lianas are represented within sites allowing for direct community-level comparisons among species. We share code for filtering of data that enables the comparison of plot and individual tree data across sites. For each species, we compiled relevant life history attributes (e.g., seed mass, dispersal syndrome, seed longevity, sexual system) that may serve as important predictors of mast seeding in future analyses. To aid in phylogenetically informed analyses, we also share a phylogeny and phylogenetic distance matrix for all species in the dataset. These data can be used to investigate continent-scale ecological properties of seed production, including individual and population variability, synchrony within and across species, and how these properties of seed production vary in relation to plant species traits and environmental conditions. In addition, these data can be used to assess how annual variability in seed production is associated with climate conditions and how that varies across populations, species, and regions. The dataset is released under a CC0 1.0 Universal public domain license.

BACKGROUND: The rapidly changing policy climate related to cannabis legalization has led to drastic changes in cannabis use in the United States (US). Medical cannabis use is increasing overall, but at a faster rate among older adults compared to other age groups. OBJECTIVE: The aim was to investigate older adults cannabis use behaviors and attitudes around disclosing medical cannabis use to their primary healthcare providers (HCPs). METHODS: Nineteen older adults (ages 65+ years) with self-reported medical cannabis use were recruited from flyers posted in ambulatory clinics in San Diego, CA. Surveys and semi-structured interviews on cannabis use were completed. A multi-methods approach was used to analyze data. RESULTS: Participants mean age was 75.3 years; 52.6% identified as women, and 89.5% as White. Cannabis was used by all participants to treat pain and by 75% for insomnia, with 25-33% reductions in use of prescription medications to treat these symptoms. Approximately 89% reported their primary HCPs were aware of their cannabis use, and 84.2% felt very comfortable/comfortable talking to HCPs about cannabis. Common themes from interviews included participants (1) being motivated to disclose cannabis use to their HCPs to seek medical advice on dosing, side effects, and benefits of cannabis, (2) feeling comfortable disclosing cannabis use as legalization has eased the stigma around cannabis use, and (3) perceiving mostly neutral attitudes from HCPs on their cannabis use. CONCLUSION: The study emphasizes the pivotal role of HCPs as educators in addressing patient inquiries about cannabis, underlining the need for equipping healthcare professionals with evidence-based knowledge through education and training initiatives.

Abstract Quantifying intraocular inflammation is crucial in managing uveitis patients. We assessed the minimum B-scan density for reliable automated vitreous intensity (VI) assessment, using a novel approach based on optical coherence tomography (OCT). OCT volume scans centered on the macula were retrospectively collected in patients with uveitis. Nine B-scans per volume scan at fixed locations were automatically analyzed. The following B-scan selections were compared against the average score of 9 B-scans per volume scan as a reference standard: 1/3/5/7 central scans (1c/3c/5c/7c), 3 widely distributed scans (3w). Image data of 49 patients (31 females) were included. The median VI was 0.029 (IQR: 0.032). The intra-class-correlation coefficient of the VI across the 9 B-scans was 0.923. The median difference from the reference standard ranged between 0.001 (7c) and 0.006 (1c). It was significantly lower for scan selection 3w than 5c, p(adjusted) = 0.022, and lower for selection 7c than 3w, p(adjusted) = 0.003. The scan selections 7c and 3w showed the two highest areas under the receiver operating curve (0.985 and 0.965, respectively). Three widely distributed B-scans are sufficient to quantify VI reliably. Highest reliability was achieved using 7 central B-scans. Automated quantification of VI in uveitis is reliable and requires only few OCT B-scans.

Abstract Benchmark datasets have a powerful normative influence: by determining how the real world is represented in data, they define which problems will first be solved by algorithms built using the datasets and, by extension, who these algorithms will work for. It is desirable for these datasets to serve four functions: (1) enabling the creation of clinically relevant algorithms; (2) facilitating like-for-like comparison of algorithmic performance; (3) ensuring reproducibility of algorithms; (4) asserting a normative influence on the clinical domains and diversity of patients that will potentially benefit from technological advances. Without benchmark datasets that satisfy these functions, it is impossible to address two perennial concerns of clinicians experienced in computational research: “the data scientists just go where the data is rather than where the needs are,” and, “yes, but will this work for my patients?” If algorithms are to be developed and applied for the care of patients, then it is prudent for the research community to create benchmark datasets proactively, across specialties. As yet, best practice in this area has not been defined. Broadly speaking, efforts will include design of the dataset; compliance and contracting issues relating to the sharing of sensitive data; enabling access and reuse; and planning for translation of algorithms to the clinical environment. If a deliberate and systematic approach is not followed, not only will the considerable benefits of clinical algorithms fail to be realized, but the potential harms may be regressively incurred across existing gradients of social inequity.

BackgroundQuantitative systematic reviews have identified clinical artificial intelligence (AI)-enabled tools with adequate performance for real-world implementation. To our knowledge, no published report or protocol synthesizes the full breadth of stakeholder perspectives. The absence of such a rigorous foundation perpetuates the “AI chasm,” which continues to delay patient benefit. ObjectiveThe aim of this research is to synthesize stakeholder perspectives of computerized clinical decision support tools in any health care setting. Synthesized findings will inform future research and the implementation of AI into health care services. MethodsThe search strategy will use MEDLINE (Ovid), Scopus, CINAHL (EBSCO), ACM Digital Library, and Science Citation Index (Web of Science). Following deduplication, title, abstract, and full text screening will be performed by 2 independent reviewers with a third topic expert arbitrating. The quality of included studies will be appraised to support interpretation. Best-fit framework synthesis will be performed, with line-by-line coding completed by 2 independent reviewers. Where appropriate, these findings will be assigned to 1 of 22 a priori themes defined by the Nonadoption, Abandonment, Scale-up, Spread, and Sustainability framework. New domains will be inductively generated for outlying findings. The placement of findings within themes will be reviewed iteratively by a study advisory group including patient and lay representatives. ResultsStudy registration was obtained from PROSPERO (CRD42021256005) in May 2021. Final searches were executed in April, and screening is ongoing at the time of writing. Full text data analysis is due to be completed in October 2021. We anticipate that the study will be submitted for open-access publication in late 2021. ConclusionsThis paper describes the protocol for a qualitative evidence synthesis aiming to define barriers and facilitators to the implementation of computerized clinical decision support tools from all relevant stakeholders. The results of this study are intended to expedite the delivery of patient benefit from AI-enabled clinical tools. Trial RegistrationPROSPERO CRD42021256005; https://tinyurl.com/r4x3thvp International Registered Report Identifier (IRRID)DERR1-10.2196/33145

Purpose: Diabetic retinopathy is an increasingly common complication of diabetes mellitus that benefits from early diagnosis and frequent monitoring, especially in pregnancy where there is a greater risk of progression. Fundus fluorescein angiography is currently the gold standard method of investigation for neovascularization. However, this has risks of unpleasant and potentially dangerous side effects, with an added theoretical risk to the fetus in pregnancy. Wide-field optical coherence tomography angiography (PLEX Elite 9000, Carl Zeiss Meditec, Inc. USA and DRI Swept Source OCT Triton, Topcon Medical Systems, Inc. USA) potentially offers a safer, faster and equally effective alternative method for diagnosis and monitoring of diabetic retinopathy in pregnant patients. Observations: In the 3 cases reviewed from our clinic, we found that wide-field optical coherence tomography angiography was successfully used as an alternative to fundus fluorescein angiography in early diagnosis and ongoing monitoring of diabetic retinopathy in pregnant patients. Conclusions and importance: Wide-field optical coherence tomography angiography can be considered as a useful alternative to fundus fluorescein angiography when reviewing pregnant diabetic retinopathy patients. This allows avoidance of any unpleasant and potentially dangerous side effects associated with fluorescein injection, and appears to be just as effective in diagnosis and detecting disease progression.

Purpose Retinal signatures of systemic disease (‘oculomics’) are increasingly being revealed through a combination of high-resolution ophthalmic imaging and sophisticated modelling strategies. Progress is currently limited not mainly by technical issues, but by the lack of large labelled datasets, a sine qua non for deep learning. Such data are derived from prospective epidemiological studies, in which retinal imaging is typically unimodal, cross-sectional, of modest number and relates to cohorts, which are not enriched with subpopulations of interest, such as those with systemic disease. We thus linked longitudinal multimodal retinal imaging from routinely collected National Health Service (NHS) data with systemic disease data from hospital admissions using a privacy-by-design third-party linkage approach.Participants Between 1 January 2008 and 1 April 2018, 353 157 participants aged 40 years or older, who attended Moorfields Eye Hospital NHS Foundation Trust, a tertiary ophthalmic institution incorporating a principal central site, four district hubs and five satellite clinics in and around London, UK serving a catchment population of approximately six million people.Findings to date Among the 353 157 individuals, 186 651 had a total of 1 337 711 Hospital Episode Statistics admitted patient care episodes. Systemic diagnoses recorded at these episodes include 12 022 patients with myocardial infarction, 11 735 with all-cause stroke and 13 363 with all-cause dementia. A total of 6 261 931 retinal images of seven different modalities and across three manufacturers were acquired from 1 54 830 patients. The majority of retinal images were retinal photographs (n=1 874 175) followed by optical coherence tomography (n=1 567 358).Future plans AlzEye combines the world’s largest single institution retinal imaging database with nationally collected systemic data to create an exceptional large-scale, enriched cohort that reflects the diversity of the population served. First analyses will address cardiovascular diseases and dementia, with a view to identifying hidden retinal signatures that may lead to earlier detection and risk management of these life-threatening conditions.

Summary: Background: We undertook a Grand Challenges in Global Eye Health prioritisation exercise to identify the key issues that must be addressed to improve eye health in the context of an ageing population, to eliminate persistent inequities in health-care access, and to mitigate widespread resource limitations. Methods: Drawing on methods used in previous Grand Challenges studies, we used a multi-step recruitment strategy to assemble a diverse panel of individuals from a range of disciplines relevant to global eye health from all regions globally to participate in a three-round, online, Delphi-like, prioritisation process to nominate and rank challenges in global eye health. Through this process, we developed both global and regional priority lists. Findings: Between Sept 1 and Dec 12, 2019, 470 individuals complete round 1 of the process, of whom 336 completed all three rounds (round 2 between Feb 26 and March 18, 2020, and round 3 between April 2 and April 25, 2020) 156 (46%) of 336 were women, 180 (54%) were men. The proportion of participants who worked in each region ranged from 104 (31%) in sub-Saharan Africa to 21 (6%) in central Europe, eastern Europe, and in central Asia. Of 85 unique challenges identified after round 1, 16 challenges were prioritised at the global level; six focused on detection and treatment of conditions (cataract, refractive error, glaucoma, diabetic retinopathy, services for children and screening for early detection), two focused on addressing shortages in human resource capacity, five on other health service and policy factors (including strengthening policies, integration, health information systems, and budget allocation), and three on improving access to care and promoting equity. Interpretation: This list of Grand Challenges serves as a starting point for immediate action by funders to guide investment in research and innovation in eye health. It challenges researchers, clinicians, and policy makers to build collaborations to address specific challenges. Funding: The Queen Elizabeth Diamond Jubilee Trust, Moorfields Eye Charity, National Institute for Health Research Moorfields Biomedical Research Centre, Wellcome Trust, Sightsavers, The Fred Hollows Foundation, The Seva Foundation, British Council for the Prevention of Blindness, and Christian Blind Mission. Translations: For the French, Spanish, Chinese, Portuguese, Arabic and Persian translations of the abstract see Supplementary Materials section.

Diabetic eye disease is a cause of preventable blindness and accurate and timely referral of patients with diabetic macular edema is important to start treatment. Here the authors present a deep learning model that can predict the presence of diabetic macular edema from color fundus photographs with superior specificity and positive predictive value compared to retinal specialists.

Abstract Background Inflammation in anterior uveitis is characterised by breakdown of the blood-ocular barrier, which allows leakage of blood constituents of higher molecular weight into the aqueous humour. In routine clinical care, increase in aqueous protein levels can be observed at the slit lamp as ‘flare’ and the severity can be graded using various clinical grading systems, of which the Standardization of Uveitis Nomenclature (SUN) grading system is most commonly used. Alternative instrument-based technologies are available, which can detect aqueous protein levels in an objective and quantifiable way. This review will identify instruments capable of measuring anterior chamber inflammation in this way, their level of reliability, and how well the measurements correlate with clinical grading and/or actual aqueous protein concentration. Methods Standard systematic review methodology will be used to identify, select and extract data from studies that report the use of any instrument-based technology in the assessment of aqueous protein levels. Searches will be conducted through bibliographic databases (MEDLINE, EMBASE and Cochrane Library), clinical trial registries and the grey literature. No restrictions will be placed on language or year of publication. The outcomes of interest are the level of correlation between identified instrument-based test measurements, clinical grading and/or actual aqueous protein concentration, as well as the reliability of each index test identified. Study quality assessment will be based on QUADAS2. Correlation and reliability outcomes will be pooled and meta-analysed if appropriate. Discussion The assessment of inflammation in anterior chamber protein levels currently relies on crude and subjective clinical examination. The findings of this review will identify non-invasive technologies which show good correlation with actual protein concentration, which could be used in routine clinical practice for objective monitoring of AC inflammation. Systematic review registration PROSPERO CRD42017084167. Study screening stage has just been completed.

Abstract Background Childhood uveitis is a collection of chronic rare inflammatory eye disorders which result in visual loss in at least one eye of one fifth of affected children. Despite the introduction of novel systemic immunochemotherapies, it remains a blinding disease. We have undertaken a systematic review of outcome measures used in interventional trials of children with, or at risk of uveitis, in order to investigate metric quality and heterogeneity, as possible barriers to the translation of clinical research into improved outcomes. Methods Systematic review of trials registered within databases approved by the International Committee of Medical Journal Editors (ICMJE). Eligible trials for were those which involved participants aged under 18 years with or at risk of non-infectious uveitis. Data on date of study commencement, uveitis site, inclusion age criteria, and outcome measure characteristics including type, dimension and quality were extracted independently by two authors. Quality was determined using the reproducibility, validity and age-appropriateness of the metric. Results Of 917 identified trials, 57 were eligible for inclusion. Twenty different domains across five dimensions were used as primary outcome measures. The structure most commonly used was multiple separate outcome measures. In a quarter of studies, outcomes were assessed less than 3 months following the intervention. Disease activity was the most commonly assessed dimension, with only 30 studies (60%) using reproducible methodologies to assess activity. Only 2/12 (18%) studies on intermediate or posterior uveitis used reproducible activity grading schemes. Of 18 studies involving children aged under 6 years old which used outcome measures related to visual function, only 8/18 (44%) described the use of age-appropriate acuity assessment measures. None of the studies used a vision related quality of life metrics which had been validated for use in childhood. Conclusion This review of outcome measures in childhood uveitis interventional trials has identified under-utilisation of reproducible or child appropriate measures, and considerable heterogeneity in metric type, and structure. Clinicians and researchers interested in improving outcomes for affected children must identify a patient and family centred core outcome set, and work to validate both objective and patient (or proxy) reported disease age appropriate outcome measures.

The fundamental trade-off between current and future reproduction has long been considered to result in a tendency for species that can grow large to begin reproduction at a larger size. Due to the prolonged time required to reach maturity, estimates of tree maturation size remain very rare and we lack a global view on the generality and the shape of this trade-off. Using seed production from five continents, we estimate tree maturation sizes for 486 tree species spanning tropical to boreal climates. Results show that a species' maturation size increases with maximum size, but in a non-proportional way: the largest species begin reproduction at smaller sizes than would be expected if maturation were simply proportional to maximum size. Furthermore, the decrease in relative maturation size is steepest in cold climates. These findings on maturation size drivers are key to accurately represent forests' responses to disturbance and climate change.