Implementation of Artificial Intelligence in Retinopathy of Prematurity Care: Challenges and Opportunities.

The diagnosis of retinopathy of prematurity (ROP) is primarily image-based and suitable for implementation of artificial intelligence (AI) systems. Increasing incidence of ROP, especially in low and middle-income countries, has also put tremendous stress on health care systems. Barriers to the implementation of AI include infrastructure, regulatory, legal, cost, sustainability, and scalability. This review describes currently available AI and imaging systems, how a stable telemedicine infrastructure is crucial to AI implementation, and how successful ROP programs have been run in both low and middle-income countries and high-income countries. More work is needed in terms of validating AI systems with different populations with various low-cost imaging devices that have recently been developed. A sustainable and cost-effective ROP screening program is crucial in the prevention of childhood blindness.

Blood biomarkers improve the prediction of prevalent and incident severe chronic kidney disease.

BACKGROUND: The prevalence of chronic kidney disease (CKD) is high. Identification of cases with CKD or at high risk of developing it is important to tailor early interventions. The objective of this study was to identify blood metabolites associated with prevalent and incident severe CKD, and to quantify the corresponding improvement in CKD detection and prediction. METHODS: Data from four cohorts were analyzed: Singapore Epidemiology of Eye Diseases (SEED) (n = 8802), Copenhagen Chronic Kidney Disease (CPH) (n = 916), Singapore Diabetic Nephropathy (n = 714), and UK Biobank (UKBB) (n = 103,051). Prevalent CKD (stages 3-5) was defined as estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m2; incident severe CKD as CKD-related mortality or kidney failure occurring within 10 years. We used multivariable regressions to identify, among 146 blood metabolites, those associated with CKD, and quantify the corresponding increase in performance. RESULTS: Chronic kidney disease prevalence (stages 3-5) and severe incidence were 11.4% and 2.2% in SEED, and 2.3% and 0.2% in UKBB. Firstly, phenylalanine (Odds Ratio [OR] 1-SD increase = 1.83 [1.73, 1.93]), tyrosine (OR = 0.75 [0.71, 0.79]), docosahexaenoic acid (OR = 0.90 [0.85, 0.95]), citrate (OR = 1.41 [1.34, 1.47]) and triglycerides in medium high density lipoprotein (OR = 1.07 [1.02, 1.13]) were associated with prevalent stages 3-5 CKD. Mendelian randomization analyses suggested causal relationships. Adding these metabolites beyond traditional risk factors increased the area under the curve (AUC) by 3% and the sensitivity by 7%. Secondly, lactate (HR = 1.33 [1.08, 1.64]) and tyrosine (HR = 0.74 [0.58, 0.95]) were associated with incident severe CKD among individuals with eGFR < 90 mL/min/1.73 m2 at baseline. These metabolites increased the c-index by 2% and sensitivity by 5% when added to traditional risk factors. CONCLUSION: The performance improvements of CKD detection and prediction achieved by adding metabolites to traditional risk factors are modest and further research is necessary to fully understand the clinical implications of these findings.

Artificial Intelligence, Digital Imaging, and Robotics Technologies for Surgical Vitreoretinal Diseases.

OBJECTIVE: To review recent technological advancement in imaging, surgical visualization, robotics technology, and the use of artificial intelligence in surgical vitreoretinal (VR) diseases. BACKGROUND: Technological advancements in imaging enhance both preoperative and intraoperative management of surgical VR diseases. Widefield imaging in fundal photography and OCT can improve assessment of peripheral retinal disorders such as retinal detachments, degeneration, and tumors. OCT angiography provides a rapid and noninvasive imaging of the retinal and choroidal vasculature. Surgical visualization has also improved with intraoperative OCT providing a detailed real-time assessment of retinal layers to guide surgical decisions. Heads-up display and head-mounted display utilize 3-dimensional technology to provide surgeons with enhanced visual guidance and improved ergonomics during surgery. Intraocular robotics technology allows for greater surgical precision and is shown to be useful in retinal vein cannulation and subretinal drug delivery. In addition, deep learning techniques leverage on diverse data including widefield retinal photography and OCT for better predictive accuracy in classification, segmentation, and prognostication of many surgical VR diseases. CONCLUSION: This review article summarized the latest updates in these areas and highlights the importance of continuous innovation and improvement in technology within the field. These advancements have the potential to reshape management of surgical VR diseases in the very near future and to ultimately improve patient care. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Optical Coherence Tomography and Optical Coherence Tomography Angiography in Pediatric Retinal Diseases.

Indirect ophthalmoscopy and handheld retinal imaging are the most common and traditional modalities for the evaluation and documentation of the pediatric fundus, especially for pre-verbal children. Optical coherence tomography (OCT) allows for in vivo visualization that resembles histology, and optical coherence tomography angiography (OCTA) allows for non-invasive depth-resolved imaging of the retinal vasculature. Both OCT and OCTA were extensively used and studied in adults, but not in children. The advent of prototype handheld OCT and OCTA have allowed for detailed imaging in younger infants and even neonates in the neonatal care intensive unit with retinopathy of prematurity (ROP). In this review, we discuss the use of OCTA and OCTA in various pediatric retinal diseases, including ROP, familial exudative vitreoretinopathy (FEVR), Coats disease and other less common diseases. For example, handheld portable OCT was shown to detect subclinical macular edema and incomplete foveal development in ROP, as well as subretinal exudation and fibrosis in Coats disease. Some challenges in the pediatric age group include the lack of a normative database and the difficulty in image registration for longitudinal comparison. We believe that technological improvements in the use of OCT and OCTA will improve our understanding and care of pediatric retina patients in the future.

Acceptance and Perception of Artificial Intelligence Usability in Eye Care (APPRAISE) for Ophthalmologists: A Multinational Perspective.

Background: Many 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. Methods: This 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. Results: One 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. Conclusion: Ophthalmologists 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.

Evaluating a Deep Learning Diabetic Retinopathy Grading System Developed on Mydriatic Retinal Images When Applied to Non-Mydriatic Community Screening.

Artificial Intelligence has showcased clear capabilities to automatically grade diabetic retinopathy (DR) on mydriatic retinal images captured by clinical experts on fixed table-top retinal cameras within hospital settings. However, in many low- and middle-income countries, screening for DR revolves around minimally trained field workers using handheld non-mydriatic cameras in community settings. This prospective study evaluated the diagnostic accuracy of a deep learning algorithm developed using mydriatic retinal images by the Singapore Eye Research Institute, commercially available as Zeiss VISUHEALTH-AI DR, on images captured by field workers on a Zeiss Visuscout® 100 non-mydriatic handheld camera from people with diabetes in a house-to-house cross-sectional study across 20 regions in India. A total of 20,489 patient eyes from 11,199 patients were used to evaluate algorithm performance in identifying referable DR, non-referable DR, and gradability. For each category, the algorithm achieved precision values of 29.60 (95% CI 27.40, 31.88), 92.56 (92.13, 92.97), and 58.58 (56.97, 60.19), recall values of 62.69 (59.17, 66.12), 85.65 (85.11, 86.18), and 65.06 (63.40, 66.69), and F-score values of 40.22 (38.25, 42.21), 88.97 (88.62, 89.31), and 61.65 (60.50, 62.80), respectively. Model performance reached 91.22 (90.79, 91.64) sensitivity and 65.06 (63.40, 66.69) specificity at detecting gradability and 72.08 (70.68, 73.46) sensitivity and 85.65 (85.11, 86.18) specificity for the detection of all referable eyes. Algorithm accuracy is dependent on the quality of acquired retinal images, and this is a major limiting step for its global implementation in community non-mydriatic DR screening using handheld cameras. This study highlights the need to develop and train deep learning-based screening tools in such conditions before implementation.

Detecting visually significant cataract using retinal photograph-based deep learning.

Age-related cataracts are the leading cause of visual impairment among older adults. Many significant cases remain undiagnosed or neglected in communities, due to limited availability or accessibility to cataract screening. In the present study, we report the development and validation of a retinal photograph-based, deep-learning algorithm for automated detection of visually significant cataracts, using more than 25,000 images from population-based studies. In the internal test set, the area under the receiver operating characteristic curve (AUROC) was 96.6%. External testing performed across three studies showed AUROCs of 91.6-96.5%. In a separate test set of 186 eyes, we further compared the algorithm's performance with 4 ophthalmologists' evaluations. The algorithm performed comparably, if not being slightly more superior (sensitivity of 93.3% versus 51.7-96.6% by ophthalmologists and specificity of 99.0% versus 90.7-97.9% by ophthalmologists). Our findings show the potential of a retinal photograph-based screening tool for visually significant cataracts among older adults, providing more appropriate referrals to tertiary eye centers.

Identifying the content for an item bank and computerized adaptive testing system to measure the impact of age-related macular degeneration on health-related quality of life.

PURPOSE: We are developing an age-related macular degeneration (AMD) health-related quality of life (HRQoL) item bank, applicable to Western and Asian populations. We report primarily on content generation and refinement, but also compare the HRQoL issues reported in our study with Western studies and current AMD-HRQoL questionnaires. METHODS: In this cross-sectional, qualitative study of AMD patients attending the Singapore National Eye Centre (May-December 2019), items/domains were generated from: (1) AMD-specific questionnaires; (2) published articles; (3) focus groups/semi-structured interviews with AMD patients (n = 27); and (4) written feedback from retinal experts. Following thematic analysis, items were systematically refined to a minimally representative set and pre-tested using cognitive interviews with 16 AMD patients. RESULTS: Of the 27 patients (mean ± standard deviation age 67.9 ± 7.0; 59.2% male), 18 (66.7%), two (7.4%), and seven (25.9%) had no, early-intermediate, and late/advanced AMD (better eye), respectively. Whilst some HRQoL issues, e.g. activity limitation, mobility, lighting, and concerns were similarly reported by Western patients and covered by other questionnaires, others like anxiety about intravitreal injections, work tasks, and financial dependency were novel. Overall, 462 items within seven independent HRQoL domains were identified: Activity limitation, Lighting, Mobility, Emotional, Concerns, AMD management, and Work. Following item refinement, items were reduced to 219, with 31 items undergoing amendment. CONCLUSION: Our 7-domain, 219-item AMD-specific HRQoL instrument will undergo psychometric testing and calibration for computerized adaptive testing. The future instrument will enable users to precisely, rapidly, and comprehensively quantify the HRQoL impact of AMD and associated treatments, with item coverage relevant across several populations.

Treat-and-Extend Regimens for the Management of Neovascular Age-related Macular Degeneration and Polypoidal Choroidal Vasculopathy: Consensus and Recommendations From the Asia-Pacific Vitreo-retina Society.

PURPOSE: Review and provide consensus recommendations on use of treat-and-extend (T&E) regimens for neovascular age-related macular degeneration (nAMD) and polypoidal choroidal vasculopathy (PCV) management with relevance for clinicians in the Asia-Pacific region. METHODS: A systematic search of MEDLINE, EMBASE, and Cochrane databases, and abstract databases of the Asia-Pacific Vitreo-retina Society, European Society of Retina Specialists, American Academy of Ophthalmology, and Controversies in Ophthalmology: Asia-Australia congresses, was conducted to assess evidence for T&E regimens in nAMD. Only studies with ≥100 study eyes were included. An expert panel reviewed the results and key factors potentially influencing the use of T&E regimens in nAMD and PCV, and subsequently formed consensus recommendations for their application in the Asia-Pacific region. RESULTS: Twenty-seven studies were included. Studies demonstrated that T&E regimens with aflibercept, ranibizumab, or bevacizumab in nAMD, and with aflibercept in PCV, were efficacious and safe. The recommendation for T&E is, after ≥3 consecutive monthly loading doses, treatment intervals can be extended by 2 to 4 weeks up to 12 to 16 weeks. When disease activity recurs, the recommendation is to reinject and shorten intervals by 2 to 4 weeks until fluid resolution, after which treatment intervals can again be extended. Intraretinal fluid should be treated until resolved; however, persistent minimal subretinal fluid after consecutive treatments may be tolerated with treatment intervals maintained or extended if the clinical condition is stable. CONCLUSIONS: T&E regimens are efficacious and safe for nAMD and PCV, can reduce the number of visits, and minimize the overall burden for clinicians and patients.

Considerations for Artificial Intelligence Real-World Implementation in Ophthalmology: Providers' and Patients' Perspectives.

ABSTRACT: Artificial Intelligence (AI), in particular deep learning, has made waves in the health care industry, with several prominent examples shown in ophthalmology. Despite the burgeoning reports on the development of new AI algorithms for detection and management of various eye diseases, few have reached the stage of regulatory approval for real-world implementation. To better enable real-world translation of AI systems, it is important to understand the demands, needs, and concerns of both health care professionals and patients, as providers and recipients of clinical care are impacted by these solutions. This review outlines the advantages and concerns of incorporating AI in ophthalmology care delivery, from both the providers' and patients' perspectives, and the key enablers for seamless transition to real-world implementation.

cGMP-grade human iPSC-derived retinal photoreceptor precursor cells rescue cone photoreceptor damage in non-human primates.

BACKGROUND: Retinal regenerative therapies hold great promise for the treatment of inherited retinal degenerations (IRDs). Studies in preclinical lower mammal models of IRDs have suggested visual improvement following retinal photoreceptor precursors transplantation, but there is limited evidence on the ability of these transplants to rescue retinal damage in higher mammals. The purpose of this study was to evaluate the therapeutic potential of photoreceptor precursors derived from clinically compliant induced pluripotent stem cells (iPSCs). METHODS: Photoreceptor precursors were sub-retinally transplanted into non-human primates (Macaca fascicularis). The cells were transplanted both in naïve and cobalt chloride-induced retinal degeneration models who had been receiving systemic immunosuppression for one week prior to the procedure. Optical coherence tomography, fundus autofluorescence imaging, electroretinography, ex vivo histology and immunofluorescence staining were used to evaluate retinal structure, function and survival of transplanted cells. RESULTS: There were no adverse effects of iPSC-derived photoreceptor precursors on retinal structure or function in naïve NHP models, indicating good biocompatibility. In addition, photoreceptor precursors injected into cobalt chloride-induced retinal degeneration NHP models demonstrated an ability both to survive and to mature into cone photoreceptors at 3 months post-transplant. Optical coherence tomography showed restoration of retinal ellipsoid zone post-transplantation. CONCLUSIONS: These findings demonstrate the safety and therapeutic potential of clinically compliant iPSC-derived photoreceptor precursors as a cell replacement source for future clinical trials.

Submacular integration of hESC-RPE monolayer xenografts in a surgical non-human primate model.

BACKGROUND: Human pluripotent stem cells (hPSCs) provide a promising cell source for retinal cell replacement therapy but often lack standardized cell production and live-cell shipment logistics as well as rigorous analyses of surgical procedures for cell transplantation in the delicate macula area. We have previously established a xeno- and feeder cell-free production system for hPSC differentiated retinal pigment epithelial (RPE) cells, and herein, a novel immunosuppressed non-human primate (NHP) model with a disrupted ocular immune privilege is presented for transplanting human embryonic stem cell (hESC)-derived RPE on a scaffold, and the safety and submacular graft integration are assessed. Furthermore, the feasibility of intercontinental shipment of live hESC-RPE is examined. METHODS: Cynomolgus monkeys were systemically immunosuppressed and implanted with a hESC-RPE monolayer on a permeable polyester-terephthalate (PET) scaffold. Microscope-integrated intraoperative optical coherence tomography (miOCT)-guided surgery, postoperative follow-up incorporated scanning laser ophthalmoscopy, spectral domain (SD-) OCT, and full-field electroretinography (ERG) were used as outcome measures. In addition, histology was performed after a 28-day follow-up. RESULTS: Intercontinental cell shipment, which took >30 h from the manufacturing to the transplantation site, did not alter the hESC-RPE quality. The submacular hESC-RPE xenotransplantation was performed in 11 macaques. The miOCT typically revealed foveal disruption. ERG showed amplitude and peak time preservation in cases with favorable surgical outcomes. Histology confirmed photoreceptor preservation above the grafts and in vivo phagocytosis by hESC-RPE, albeit evidence of cytoplasmic redistribution of opsin in photoreceptors and glia hypertrophy. The immunosuppression protocol efficiently suppressed retinal T cell infiltration and microglia activation. CONCLUSION: These results suggest both structural and functional submacular integrations of hESC-RPE xenografts. It is anticipated that surgical technique refinement will further improve the engraftment of macular cell therapeutics with significant translational relevance to improve future clinical trials.

Retinal microvascular signs and risk of diabetic kidney disease in asian and white populations.

The objective was to examine prospectively the association between retinal microvascular signs and development of diabetic kidney disease (DKD) in Asian and White populations. We analysed two population-based cohorts, composing of 1,221 Asians (SEED) and 703 White (WESDR) adults with diabetes. Retinal microvascular signs at baseline included vascular caliber (arteriolar-CRAE, and venular-CRVE) and diabetic retinopathy (DR). Incident cases of DKD were identified after ~ 6-year. Incident cases were defined based on eGFR in SEED and proteinuria or history of renal dialysis in WESDR. The incidence of DKD were 11.8% in SEED and 14.0% in WESDR. Wider CRAE in SEED (OR = 1.58 [1.02, 2.45]) and wider CRVE (OR = 1.69 [1.02, 2.80)) in WESDR were associated with increased risk of DKD. Presence of DR was associated with an increased risk of DKD in both cohorts (SEED: OR = 1.91 [1.21, 3.01] in SEED, WESDR: OR = 1.99 [1.18, 3.35]). Adding DR and retinal vascular calibers in the model beyond traditional risk factors led to an improvement of predictive performance of DKD risk between 1.1 and 2.4%; and improved classification (NRI 3 between 9%). Microvascular changes in the retina are longitudinally associated with risk of DKD.

Diabetic Macular Ischemia: Influence of Optical Coherence Tomography Angiography Parameters on Changes in Functional Outcomes Over One Year.

Purpose: To prospectively evaluate whether diabetic macular ischemia detected with coherence tomography angiography (OCTA) is associated with change in functional outcomes over a period of one year. Methods: This is a one-year prospective, observational study that included 56 eyes with varying levels of diabetic retinopathy. All participants underwent best corrected visual acuity evaluation, swept-source OCTA and microperimetry at baseline and repeated at one year. Parafoveal vessel densities (VD) and foveal avascular zone (FAZ) areas were generated from OCTA in the superficial and deep vascular plexuses. The influence of baseline and change in OCTA parameters on change in visual acuity and retinal sensitivity over one year was evaluated. Results: Over the one-year follow-up period, 16% (9) of eyes had at least one line worsening in BCVA and 7% (4) of eyes had at least 5% decrease in retinal sensitivity compared to baseline. Diabetic retinopathy progressed in 12.5%. Mean superficial vascular plexus (SVP) FAZ area increased (0.32 ± 0.15 to 0.39 ± 0.18 mm2, P = 0.003) and parafoveal VD in deep vascular plexus (DVP) decreased (49.8 ± 3.7% to 48.8 ± 2.9%, P = 0.040) at one year compared to baseline. In the multivariate regression analysis, larger baseline DVP FAZ area was associated with worsening of BCVA over one year (ß = 0.16 logMAR per mm2, 95% CI 0.02 to 0.31, P = 0.032). In addition, larger decreases in SVP VD (ß = -4.18 db per 10% decrease, 95% CI -6.55 to -1.80, P = 0.002) was associated with worsening of retinal sensitivity over one year. Conclusions: Progression of parafoveal microvasculature changes over one year can be detected using OCTA. Larger baseline DVP FAZ area on OCTA is predictive of worsening in visual outcomes, and larger decreases in SVP VD were associated with worsening of retinal sensitivity over a course of one year in diabetic individuals.

Hints for Gentle Submacular Injection in Non-Human Primates Based on Intraoperative OCT Guidance.

Purpose: Delivery of Advanced Therapy Medicinal Products to the submacular space is increasingly evolving into a therapeutic modality. Cell replacement for age-related macular degeneration (AMD) and gene therapy for RPE65 are recent successful examples. Herein, a nonhuman primate (NHP) model was used to investigate surgical means to detach the macula. Methods: Sixteen eyes of 13 healthy macaques underwent a 25-gauge vitrectomy and subretinal injection of balanced salt solution monitored by microscope-integrated intraoperative optical coherence tomography (miOCT). The animals were followed with OCT and histology. Results: The miOCT monitoring allowed a more precise definition of surgical trauma ranging from an initial full-thickness foveal tear, or induction of a cystoid macular edema (CME), until no foveal defect was discernible, as the technique improved. However, as the subretinal fluid wave detached the fovea, the aforementioned lesions formed, whereas persistent retinal adhesion reproducibly proved to remain in the distal parafoveal semi-annulus. Measures to reduce foveal trauma during submacular fluid injection included reducing intraocular pressure, injection volume, and velocity, as well as the retinal location for bleb initiation, use of a vitreous tamponade, and a dual-bore subretinal cannula. Conclusions: A stable very low intraocular pressure and careful subretinal injection may avoid tangential macular stretching or mechanical CME formation, while vitreous tamponade may facilitate a more lamellar subretinal flow, all thereby reducing foveal trauma during submacular injection in NHP. Translational Relevance: These results can be relevant to any submacular surgery procedure used today, as they synergistically reduce the risk of compromising foveal integrity.

Surgical Transplantation of Human RPE Stem Cell-Derived RPE Monolayers into Non-Human Primates with Immunosuppression.

Recent trials of retinal pigment epithelium (RPE) transplantation for the treatment of disorders such as age-related macular degeneration have been promising. However, limitations of existing strategies include the uncertain survival of RPE cells delivered by cell suspension and the inherent risk of uncontrolled cell proliferation in the vitreous cavity. Human RPE stem cell-derived RPE (hRPESC-RPE) transplantation can rescue vision in a rat model of retinal dystrophy and survive in the rabbit retina for at least 1 month. The present study placed hRPESC-RPE monolayers under the macula of a non-human primate model for 3 months. The transplant was able to recover in vivo and maintained healthy photoreceptors. Importantly, there was no evidence that subretinally transplanted monolayers underwent an epithelial-mesenchymal transition. Neither gliosis in adjacent retina nor epiretinal membranes were observed. These findings suggest that hRPESC-RPE monolayers are safe and may be a useful source for RPE cell replacement therapy.

Digital health during COVID-19: lessons from operationalising new models of care in ophthalmology.

The COVID-19 pandemic has resulted in massive disruptions within health care, both directly as a result of the infectious disease outbreak, and indirectly because of public health measures to mitigate against transmission. This disruption has caused rapid dynamic fluctuations in demand, capacity, and even contextual aspects of health care. Therefore, the traditional face-to-face patient-physician care model has had to be re-examined in many countries, with digital technology and new models of care being rapidly deployed to meet the various challenges of the pandemic. This Viewpoint highlights new models in ophthalmology that have adapted to incorporate digital health solutions such as telehealth, artificial intelligence decision support for triaging and clinical care, and home monitoring. These models can be operationalised for different clinical applications based on the technology, clinical need, demand from patients, and manpower availability, ranging from out-of-hospital models including the hub-and-spoke pre-hospital model, to front-line models such as the inflow funnel model and monitoring models such as the so-called lighthouse model for provider-led monitoring. Lessons learnt from operationalising these models for ophthalmology in the context of COVID-19 are discussed, along with their relevance for other specialty domains.

Artificial intelligence for teleophthalmology-based diabetic retinopathy screening in a national programme: an economic analysis modelling study.

BACKGROUND: Deep learning is a novel machine learning technique that has been shown to be as effective as human graders in detecting diabetic retinopathy from fundus photographs. We used a cost-minimisation analysis to evaluate the potential savings of two deep learning approaches as compared with the current human assessment: a semi-automated deep learning model as a triage filter before secondary human assessment; and a fully automated deep learning model without human assessment. METHODS: In this economic analysis modelling study, using 39 006 consecutive patients with diabetes in a national diabetic retinopathy screening programme in Singapore in 2015, we used a decision tree model and TreeAge Pro to compare the actual cost of screening this cohort with human graders against the simulated cost for semi-automated and fully automated screening models. Model parameters included diabetic retinopathy prevalence rates, diabetic retinopathy screening costs under each screening model, cost of medical consultation, and diagnostic performance (ie, sensitivity and specificity). The primary outcome was total cost for each screening model. Deterministic sensitivity analyses were done to gauge the sensitivity of the results to key model assumptions. FINDINGS: From the health system perspective, the semi-automated screening model was the least expensive of the three models, at US$62 per patient per year. The fully automated model was $66 per patient per year, and the human assessment model was $77 per patient per year. The savings to the Singapore health system associated with switching to the semi-automated model are estimated to be $489 000, which is roughly 20% of the current annual screening cost. By 2050, Singapore is projected to have 1 million people with diabetes; at this time, the estimated annual savings would be $15 million. INTERPRETATION: This study provides a strong economic rationale for using deep learning systems as an assistive tool to screen for diabetic retinopathy. FUNDING: Ministry of Health, Singapore.

Artificial intelligence for diabetic retinopathy screening, prediction and management.

PURPOSE OF REVIEW: Diabetic retinopathy is the most common specific complication of diabetes mellitus. Traditional care for patients with diabetes and diabetic retinopathy is fragmented, uncoordinated and delivered in a piecemeal nature, often in the most expensive and high-resource tertiary settings. Transformative new models incorporating digital technology are needed to address these gaps in clinical care. RECENT FINDINGS: Artificial intelligence and telehealth may improve access, financial sustainability and coverage of diabetic retinopathy screening programs. They enable risk stratifying patients based on individual risk of vision-threatening diabetic retinopathy including diabetic macular edema (DME), and predicting which patients with DME best respond to antivascular endothelial growth factor therapy. SUMMARY: Progress in artificial intelligence and tele-ophthalmology for diabetic retinopathy screening, including artificial intelligence applications in 'real-world settings' and cost-effectiveness studies are summarized. Furthermore, the initial research on the use of artificial intelligence models for diabetic retinopathy risk stratification and management of DME are outlined along with potential future directions. Finally, the need for artificial intelligence adoption within ophthalmology in response to coronavirus disease 2019 is discussed. Digital health solutions such as artificial intelligence and telehealth can facilitate the integration of community, primary and specialist eye care services, optimize the flow of patients within healthcare networks, and improve the efficiency of diabetic retinopathy management.