From Data to Deployment: The Collaborative Community on Ophthalmic Imaging Roadmap for Artificial Intelligence in Age-Related Macular Degeneration.

OBJECTIVE: Health care systems worldwide are challenged to provide adequate care for the 200 million individuals with age-related macular degeneration (AMD). Artificial intelligence (AI) has the potential to make a significant, positive impact on the diagnosis and management of patients with AMD; however, the development of effective AI devices for clinical care faces numerous considerations and challenges, a fact evidenced by a current absence of Food and Drug Administration (FDA)-approved AI devices for AMD. PURPOSE: To delineate the state of AI for AMD, including current data, standards, achievements, and challenges. METHODS: Members of the Collaborative Community on Ophthalmic Imaging Working Group for AI in AMD attended an inaugural meeting on September 7, 2020, to discuss the topic. Subsequently, they undertook a comprehensive review of the medical literature relevant to the topic. Members engaged in meetings and discussion through December 2021 to synthesize the information and arrive at a consensus. RESULTS: Existing infrastructure for robust AI development for AMD includes several large, labeled data sets of color fundus photography and OCT images; however, image data often do not contain the metadata necessary for the development of reliable, valid, and generalizable models. Data sharing for AMD model development is made difficult by restrictions on data privacy and security, although potential solutions are under investigation. Computing resources may be adequate for current applications, but knowledge of machine learning development may be scarce in many clinical ophthalmology settings. Despite these challenges, researchers have produced promising AI models for AMD for screening, diagnosis, prediction, and monitoring. Future goals include defining benchmarks to facilitate regulatory authorization and subsequent clinical setting generalization. CONCLUSIONS: Delivering an FDA-authorized, AI-based device for clinical care in AMD involves numerous considerations, including the identification of an appropriate clinical application; acquisition and development of a large, high-quality data set; development of the AI architecture; training and validation of the model; and functional interactions between the model output and clinical end user. The research efforts undertaken to date represent starting points for the medical devices that eventually will benefit providers, health care systems, and patients.

Artificial Intelligence Improves Patient Follow-Up in a Diabetic Retinopathy Screening Program.

Purpose: We examine the rate of and reasons for follow-up in an Artificial Intelligence (AI)-based workflow for diabetic retinopathy (DR) screening relative to two human-based workflows. Patients and Methods: A DR screening program initiated September 2019 between one institution and its affiliated primary care and endocrinology clinics screened 2243 adult patients with type 1 or 2 diabetes without a diagnosis of DR in the previous year in the San Francisco Bay Area. For patients who screened positive for more-than-mild-DR (MTMDR), rates of follow-up were calculated under a store-and-forward human-based DR workflow ("Human Workflow"), an AI-based workflow involving IDx-DR ("AI Workflow"), and a two-step hybrid workflow ("AI-Human Hybrid Workflow"). The AI Workflow provided results within 48 hours, whereas the other workflows took up to 7 days. Patients were surveyed by phone about follow-up decisions. Results: Under the AI Workflow, 279 patients screened positive for MTMDR. Of these, 69.2% followed up with an ophthalmologist within 90 days. Altogether 70.5% (N=48) of patients who followed up chose their location based on primary care referral. Among the subset of patients that were seen in person at the university eye institute under the Human Workflow and AI-Human Hybrid Workflow, 12.0% (N=14/117) and 11.7% (N=12/103) of patients with a referrable screening result followed up compared to 35.5% of patients under the AI Workflow (N=99/279; χ2df=2 = 36.70, p < 0.00000001). Conclusion: Ophthalmology follow-up after a positive DR screening result is approximately three-fold higher under the AI Workflow than either the Human Workflow or AI-Human Hybrid Workflow. Improved follow-up behavior may be due to the decreased time to screening result.

AI-Human Hybrid Workflow Enhances Teleophthalmology for the Detection of Diabetic Retinopathy.

Objective: Detection of diabetic retinopathy (DR) outside of specialized eye care settings is an important means of access to vision-preserving health maintenance. Remote interpretation of fundus photographs acquired in a primary care or other nonophthalmic setting in a store-and-forward manner is a predominant paradigm of teleophthalmology screening programs. Artificial intelligence (AI)-based image interpretation offers an alternative means of DR detection. IDx-DR (Digital Diagnostics Inc) is a Food and Drug Administration-authorized autonomous testing device for DR. We evaluated the diagnostic performance of IDx-DR compared with human-based teleophthalmology over 2 and a half years. Additionally, we evaluated an AI-human hybrid workflow that combines AI-system evaluation with human expert-based assessment for referable cases. Design: Prospective cohort study and retrospective analysis. Participants: Diabetic patients ≥ 18 years old without a prior DR diagnosis or DR examination in the past year presenting for routine DR screening in a primary care clinic. Methods: Macula-centered and optic nerve-centered fundus photographs were evaluated by an AI algorithm followed by consensus-based overreading by retina specialists at the Stanford Ophthalmic Reading Center. Detection of more-than-mild diabetic retinopathy (MTMDR) was compared with in-person examination by a retina specialist. Main Outcome Measures: Sensitivity, specificity, accuracy, positive predictive value, and gradability achieved by the AI algorithm and retina specialists. Results: The AI algorithm had higher sensitivity (95.5% sensitivity; 95% confidence interval [CI], 86.7%-100%) but lower specificity (60.3% specificity; 95% CI, 47.7%-72.9%) for detection of MTMDR compared with remote image interpretation by retina specialists (69.5% sensitivity; 95% CI, 50.7%-88.3%; 96.9% specificity; 95% CI, 93.5%-100%). Gradability of encounters was also lower for the AI algorithm (62.5%) compared with retina specialists (93.1%). A 2-step AI-human hybrid workflow in which the AI algorithm initially rendered an assessment followed by overread by a retina specialist of MTMDR-positive encounters resulted in a sensitivity of 95.5% (95% CI, 86.7%-100%) and a specificity of 98.2% (95% CI, 94.6%-100%). Similarly, a 2-step overread by retina specialists of AI-ungradable encounters improved gradability from 63.5% to 95.6% of encounters. Conclusions: Implementation of an AI-human hybrid teleophthalmology workflow may both decrease reliance on human specialist effort and improve diagnostic accuracy. Financial Disclosures: Proprietary or commercial disclosure may be found after the references.

A Deep-Learning Algorithm to Predict Short-Term Progression to Geographic Atrophy on Spectral-Domain Optical Coherence Tomography.

Importance: The identification of patients at risk of progressing from intermediate age-related macular degeneration (iAMD) to geographic atrophy (GA) is essential for clinical trials aimed at preventing disease progression. DeepGAze is a fully automated and accurate convolutional neural network-based deep learning algorithm for predicting progression from iAMD to GA within 1 year from spectral-domain optical coherence tomography (SD-OCT) scans. Objective: To develop a deep-learning algorithm based on volumetric SD-OCT scans to predict the progression from iAMD to GA during the year following the scan. Design, Setting, and Participants: This retrospective cohort study included participants with iAMD at baseline and who either progressed or did not progress to GA within the subsequent 13 months. Participants were included from centers in 4 US states. Data set 1 included patients from the Age-Related Eye Disease Study 2 AREDS2 (Ancillary Spectral-Domain Optical Coherence Tomography) A2A study (July 2008 to August 2015). Data sets 2 and 3 included patients with imaging taken in routine clinical care at a tertiary referral center and associated satellites between January 2013 and January 2023. The stored imaging data were retrieved for the purpose of this study from July 1, 2022, to February 1, 2023. Data were analyzed from May 2021 to July 2023. Exposure: A position-aware convolutional neural network with proactive pseudointervention was trained and cross-validated on Bioptigen SD-OCT volumes (data set 1) and validated on 2 external data sets comprising Heidelberg Spectralis SD-OCT scans (data sets 2 and 3). Main Outcomes and Measures: Prediction of progression to GA within 13 months was evaluated with area under the receiver-operator characteristic curves (AUROC) as well as area under the precision-recall curve (AUPRC), sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. Results: The study included a total of 417 patients: 316 in data set 1 (mean [SD] age, 74 [8]; 185 [59%] female), 53 in data set 2, (mean [SD] age, 83 [8]; 32 [60%] female), and 48 in data set 3 (mean [SD] age, 81 [8]; 32 [67%] female). The AUROC for prediction of progression from iAMD to GA within 1 year was 0.94 (95% CI, 0.92-0.95; AUPRC, 0.90 [95% CI, 0.85-0.95]; sensitivity, 0.88 [95% CI, 0.84-0.92]; specificity, 0.90 [95% CI, 0.87-0.92]) for data set 1. The addition of expert-annotated SD-OCT features to the model resulted in no improvement compared to the fully autonomous model (AUROC, 0.95; 95% CI, 0.92-0.95; P = .19). On an independent validation data set (data set 2), the model predicted progression to GA with an AUROC of 0.94 (95% CI, 0.91-0.96; AUPRC, 0.92 [0.89-0.94]; sensitivity, 0.91 [95% CI, 0.74-0.98]; specificity, 0.80 [95% CI, 0.63-0.91]). At a high-specificity operating point, simulated clinical trial recruitment was enriched for patients progressing to GA within 1 year by 8.3- to 20.7-fold (data sets 2 and 3). Conclusions and Relevance: The fully automated, position-aware deep-learning algorithm assessed in this study successfully predicted progression from iAMD to GA over a clinically meaningful time frame. The ability to predict imminent GA progression could facilitate clinical trials aimed at preventing the condition and could guide clinical decision-making regarding screening frequency or treatment initiation.

MACULAR INFARCTION IN A PATIENT WITH SICKLE CELL TRAIT.

PURPOSE: Sickle cell trait affects 8% of African Americans. Once believed to represent a benign carrier state, it has been linked to an increased risk of several of the pathological conditions that arise in sickle cell disease in at-risk individuals with hematologic and vascular comorbidities. Macular infarction is a known complication of sickle cell disease; this article illustrates this unique presentation in a patient with sickle cell trait. METHODS: Case report. PATIENT: A 74-year-old African American man presented with the complaint of a central scotoma of the right eye. RESULTS: Multimodal retinal imaging identified central macular infarction with severe inner retinal atrophy. Laboratory testing confirmed the presence of sickle cell trait. Other pertinent positives included low levels of protein C and protein S, untreated obstructive sleep apnea, and elevated levels of homocysteine in the setting of alcoholic liver cirrhosis and chronic kidney disease. CONCLUSION: Ocular manifestations of sickle cell trait have most frequently been reported in individuals with systemic medical comorbidities that predispose to erythrocyte sickling and vaso-occlusive disease. This case identifies a novel complication of sickle cell trait disorder, macular infarction, in a patient with comorbid associations.

Posterior Uveitis Associated with Cemiplimab.

PURPOSE: The immune checkpoint inhibitors (ICPIs) comprise a class of oncologic immunotherapies. The most recent US Food and Drug Administration-approved ICPI is cemiplimab (Libtayo®). Cemiplimab, like the other ICPIs, blocks checkpoint receptors in order to disinhibit T-cells so that they may detect and eliminate tumor cells. Consequently, treatment with ICPIs is associated with immune-related adverse events including uveitis. METHODS: Case report. RESULTS: A 63-year-old man with a history of metastatic squamous cell carcinoma developed blurry vision 3 months after starting treatment with cemiplimab. The patient was found to have posterior uveitis with retinal vasculitis that was successfully controlled with discontinuation of the medication as well as treatment with local and systemic steroids. CONCLUSION: Similar to other ICPIs, uveitis may be associated with cemiplimab. In the setting of posterior uveitis, treatment may require cessation of cemiplimab and intensive steroid treatment.

Multimodal Imaging, Tele-Education, and Telemedicine in Retinopathy of Prematurity.

Retinopathy of prematurity (ROP) is a disease that affects retinal vasculature in premature infants and remains one of the leading causes of blindness in childhood worldwide. ROP screening can encounter some difficulties such as the lack of specialists and services in rural areas. The evolution of technology has helped address these issues and led to the emergence of state-of-the-art multimodal digital imaging devices such fundus cameras with its variable properties, optical coherence tomography (OCT), OCT angiography, and fluorescein angiography which has helped immensely in the process of improving ROP care and understanding the disease pathophysiology. Computer-based imaging analysis and deep learning have recently been demonstrating promising outcomes in regard to ROP diagnosis. Telemedicine is considered an acceptable alternative to clinical examination when optimal circumstances for ROP screening in certain areas are lacking, and the expansion of these programs has been reported. Tele-education programs in ROP have the potential to improve the quality of training to physicians to optimize ROP care.

Predicting Systemic Health Features from Retinal Fundus Images Using Transfer-Learning-Based Artificial Intelligence Models.

While color fundus photos are used in routine clinical practice to diagnose ophthalmic conditions, evidence suggests that ocular imaging contains valuable information regarding the systemic health features of patients. These features can be identified through computer vision techniques including deep learning (DL) artificial intelligence (AI) models. We aim to construct a DL model that can predict systemic features from fundus images and to determine the optimal method of model construction for this task. Data were collected from a cohort of patients undergoing diabetic retinopathy screening between March 2020 and March 2021. Two models were created for each of 12 systemic health features based on the DenseNet201 architecture: one utilizing transfer learning with images from ImageNet and another from 35,126 fundus images. Here, 1277 fundus images were used to train the AI models. Area under the receiver operating characteristics curve (AUROC) scores were used to compare the model performance. Models utilizing the ImageNet transfer learning data were superior to those using retinal images for transfer learning (mean AUROC 0.78 vs. 0.65, p-value < 0.001). Models using ImageNet pretraining were able to predict systemic features including ethnicity (AUROC 0.93), age > 70 (AUROC 0.90), gender (AUROC 0.85), ACE inhibitor (AUROC 0.82), and ARB medication use (AUROC 0.78). We conclude that fundus images contain valuable information about the systemic characteristics of a patient. To optimize DL model performance, we recommend that even domain specific models consider using transfer learning from more generalized image sets to improve accuracy.

Immune Checkpoint Inhibitor-associated Uveitis: Review of Treatments and Outcomes.

Purpose: Immune checkpoint inhibitors (ICPIs), novel immunotherapy agents employed in the treatment of metastatic melanoma and other solid tumors, are associated with immune-related adverse events, including ocular inflammation. We review the current literature on immune checkpoint inhibitor-associated uveitis (ICIPU).Methods: A comprehensive literature review utilizing MEDLINE/PubMed, Cochrane, and Web of Science databases was conducted. One hundred and twenty-six cases of ICPIU reported in the literature prior to January 31, 2020 were identified and reviewed.Results: ICPIs were associated with 126 cases of anterior uveitis, intermediate uveitis, posterior uveitis, and panuveitis from 67 reports in the literature. Patients typically developed intraocular inflammation a median of 9 weeks after initiation of ICPI and 83.6% of the patients developed uveitis within 6 months. The vast majority of patients recovered to within one line of baseline vision in response to topical, local, and/or systemic steroid treatment as well as the cessation of medication.Conclusions: Prompt recognition and steroid treatment of ICPIU are critical to the care of patients receiving ICPIs.

Involvement of AMPA receptors in the antidepressant-like effects of lithium in the mouse tail suspension test and forced swim test.

In addition to its clinical antimanic effects, lithium also has efficacy in the treatment of depression. However, the mechanism by which lithium exerts its antidepressant effects is unclear. Our objective was to further characterize the effects of peripheral and central administration of lithium in mouse models of antidepressant efficacy as well as to investigate the role of alpha-amino-3-hydroxy-5-methyl-4-isoxazole-propionic acid (AMPA) receptors in these behaviors. We utilized the mouse forced swim test (FST) and tail suspension test (TST), intracerebroventricular (ICV) lithium administration, AMPA receptor inhibitors, and BS3 crosslinking followed by Western blot. Both short- and long-term administration of lithium resulted in robust antidepressant-like effects in the mouse FST and TST. Using ICV administration of lithium, we show that these effects are due to actions of lithium on the brain, rather than to peripheral effects of the drug. Both ICV and rodent chow (0.4% LiCl) administration paradigms resulted in brain lithium concentrations within the human therapeutic range. The antidepressant-like effects of lithium in the FST and TST were blocked by administration of AMPA receptor inhibitors. Additionally, administration of lithium increased the cell surface expression of GluR1 and GluR2 in the mouse hippocampus. Collectively, these data show that lithium exerts centrally mediated antidepressant-like effects in the mouse FST and TST that require AMPA receptor activation. Lithium may exert its antidepressant effects in humans through AMPA receptors, thus further supporting a role of targeting AMPA receptors as a therapeutic approach for the treatment of depression.

Generation and behavioral characterization of beta-catenin forebrain-specific conditional knock-out mice.

The canonical Wnt pathway and beta-catenin have been implicated in the pathophysiology of mood disorders. We generated forebrain-specific CRE-mediated conditional beta-catenin knock-out mice to begin exploring the behavioral implications of decreased Wnt pathway signaling in the central nervous system. In situ hybridization revealed a progressive knock-out of beta-catenin that began between 2 and 4 weeks of age, and by 12 weeks resulted in considerably decreased beta-catenin expression in regions of the forebrain, including the frontal cortex, hippocampus, and striatum. A significant decrease in protein levels of beta-catenin in these brain regions was observed by Western blot. Behavioral characterization of these mice in several tests (including the forced swim test, tail suspension test (TST), learned helplessness, response and sensitization to stimulants, and light/dark box among other tests) revealed relatively circumscribed alterations. In the TST, knock-out mice spent significantly less time struggling (a depression-like phenotype). However, knock-out mice did not differ from their wild-type littermates in the other behavioral tests of mood-related or anxiety-related behaviors. These results suggest that a 60-70% beta-catenin reduction in circumscribed brain regions is only capable of inducing subtle behavioral changes. Alternatively, regulating beta-catenin may modulate drug effects rather than being a model of mood disorder pathophysiology per se.

Maternal aggression persists following lipopolysaccharide-induced activation of the immune system.

Lactating females direct aggressive behaviors towards intruders presumably to reduce the likelihood of infanticide of their pups. Infected animals display a constellation of responses that include lethargy, anorexia, and decreased social interactions. This suite of responses is referred to as sickness behavior, and is putatively part of an adaptive strategy to aid the organism in recovery from infection. Previous work has suggested that animals can suppress the behavioral symptoms of sickness in order to engage in adaptive behaviors. To test whether adaptive nest defense is affected by illness, dams received a peripheral injection of either saline or lipopolysaccharide (LPS [50, 400, or 1000 microg/kg]), a non-replicating component of bacterial cell walls that activates the immune system. Simulated infection with LPS reduced body mass and food intake in dams and interfered with litter growth in a dose-dependent manner. Generally, nest defense was unaffected by LPS; the proportion of dams displaying maternal aggression against a male intruder, as well as the latency and duration of aggressive encounters were only suppressed at the highest LPS dose tested. Further, LPS treatment also altered non-agonistic behavior during the aggression test as indicated by reduced social investigation of the intruder and an increased time spent immobile during the session. LPS administration also significantly increased serum corticosterone concentrations in lactating females. These findings suggest that maternal aggression is not suppressed by LPS-evoked immune activation at doses that attenuate other aspects of maternal and social behavior.

Targeting signal transduction pathways in the treatment of mood disorders: recent insights into the relevance of the Wnt pathway.

Regulation of complex signaling pathways plays a critical role in higher-order brain functions including the regulation of mood, cognition, appetite, sexual arousal, sleep patterns, and weight, all of which are altered in mood disorders, suggesting the involvement of signaling pathways in mood disorder pathogenesis and pathophysiology. Most existing medications used to treat mood disorders take many weeks to exert their full clinical effects, a fact which implicates changes in gene and protein expression, as well as neuroplasticity, in their mechanism of action. Modulation of signaling pathways has many downstream effects on gene expression and protein function, causing changes in synaptic function, plasticity, and response to various inputs such as neurohormones. The Wnt signaling pathway has recently been linked to the therapeutically relevant actions of available treatments of mood disorders. We provide a brief introduction to signaling cascades and their potential roles in mood disorder pathophysiology and treatment. Subsequently, we describe the Wnt signaling pathway, and glycogen synthase kinase-3 (GSK-3) and beta-catenin specifically, discussing studies that have implicated these proteins as relevant to the pathophysiology and treatment of mood disorders. Future directions, aimed at understanding mood disorders and developing more efficacious treatments, are also discussed.