Search

Recent advancements in artificial intelligence (AI) have prompted researchers to expand into the field of oculomics; the association between the retina and systemic health. Unlike conventional AI models trained on well-recognized retinal features, the retinal phenotypes that most oculomics models use are more subtle. Consequently, applying conventional tools, such as saliency maps, to understand how oculomics models arrive at their inference is problematic and open to bias. We hypothesized that neuron activation patterns (NAPs) could be an alternative way to interpret oculomics models, but currently, most existing implementations focus on failure diagnosis. In this study, we designed a novel NAP framework to interpret an oculomics model. We then applied our framework to an AI model predicting systolic blood pressure from fundus images in the United Kingdom Biobank dataset. We found that the NAP generated from our framework was correlated to the clinically relevant endpoint of cardiovascular risk. Our NAP was also able to discern two biologically distinct groups among participants who were assigned the same predicted systolic blood pressure. These results demonstrate the feasibility of our proposed NAP framework for gaining deeper insights into the functioning of oculomics models. Further work is required to validate these results on external datasets. [ABSTRACT FROM AUTHOR]

Deep learning (DL) models have shown promise in detecting chronic kidney disease (CKD) from fundus photographs. However, previous studies have utilized a serum creatinine-only estimated glomerular rate (eGFR) equation to measure kidney function despite the development of more up-to-date methods. In this study, we developed two sets of DL models using fundus images from the UK Biobank to ascertain the effects of using a creatinine and cystatin-C eGFR equation over the baseline creatinine-only eGFR equation on fundus image-based DL CKD predictors. Our results show that a creatinine and cystatin-C eGFR significantly improved classification performance over the baseline creatinine-only eGFR when the models were evaluated conventionally. However, these differences were no longer significant when the models were assessed on clinical labels based on ICD10. Furthermore, we also observed variations in model performance and systemic condition incidence between our study and the ones conducted previously. We hypothesize that limitations in existing eGFR equations and the paucity of retinal features uniquely indicative of CKD may contribute to these inconsistencies. These findings emphasize the need for developing more transparent models to facilitate a better understanding of the mechanisms underpinning the ability of DL models to detect CKD from fundus images. [ABSTRACT FROM AUTHOR]

Objective: To assess diabetes eye service use in New Zealand among people aged ≥15 years by estimating service attendance, biennial screening rate, and disparities in the use of screening and treatment services. Methods: We obtained Ministry of Health data from the National Non-Admitted Patient Collection on diabetes eye service events between 1 July 2006 and 31 December 2019 and sociodemographic and mortality data from the Virtual Diabetes Register and linked these using a unique patient identifier (encrypted National Health Index). We 1) summarized attendance at retinal screening and ophthalmology services, 2) calculated biennial and triennial screening rate, 3) summarized treatment with laser and anti-VEGF and used log-binomial regression to examine associations of all of these with age group, ethnicity, and area-level deprivation. Results: In total, 245,844 people aged ≥15 years had at least one diabetes eye service appointment attended or scheduled; half of these (n = 125,821, 51.2%) attended only retinal screening, one-sixth attended only ophthalmology (n = 35,883, 14.6%) and one-third attended both (n = 78,300, 31.8%). The biennial retinal screening rate was 62.1%, with large regional variation (73.9% in Southern District to 29.2% in West Coast). Compared with NZ Europeans, Māori were approximately twice as likely to never receive diabetes eye care or to access ophthalmology when referred from retinal screening, 9% relatively less likely to receive biennial screening and received the fewest anti-VEGF injections when treatment was commenced. Disparities in service access were also present for Pacific Peoples compared to NZ Europeans, younger and older age groups compared to those aged 50–59 years and those living in areas with higher deprivation. Conclusions: Access to diabetes eye care is suboptimal, with substantial disparity between age groups, ethnicity groups, area level deprivation quintile and across districts. Efforts to improve access to and quality of diabetes eye care services must include strengthening data collection and monitoring. [ABSTRACT FROM AUTHOR]

Purpose: To create an ensemble of Convolutional Neural Networks (CNNs), capable of detecting and stratifying the risk of progressive age-related macular degeneration (AMD) from retinal photographs. Design: Retrospective cohort study. Methods: Three individual CNNs are trained to accurately detect 1) advanced AMD, 2) drusen size and 3) the presence or otherwise of pigmentary abnormalities, from macular centered retinal images were developed. The CNNs were then arranged in a "cascading" architecture to calculate the Age-related Eye Disease Study (AREDS) Simplified 5-level risk Severity score (Risk Score 0 – Risk Score 4), for test images. The process was repeated creating a simplified binary "low risk" (Scores 0– 2) and "high risk" (Risk Score 3– 4) classification. Participants: There were a total of 188,006 images, of which 118,254 images were deemed gradable, representing 4591 patients, from the AREDS1 dataset. The gradable images were split into 50%/25%/25% ratios for training, validation and test purposes. Main Outcome Measures: The ability of the ensemble of CNNs using retinal images to predict an individual's risk of experiencing progression of their AMD based on the AREDS 5-step Simplified Severity Scale. Results: When assessed against the 5-step Simplified Severity Scale, the results generated by the ensemble of CNN's achieved an accuracy of 80.43% (quadratic kappa 0.870). When assessed against a simplified binary (Low Risk/High Risk) classification, an accuracy of 98.08%, sensitivity of ≥ 85% and specificity of ≥ 99% was achieved. Conclusion: We have created an ensemble of neural networks, trained on the AREDS 1 dataset, that is able to accurately calculate an individual's score on the AREDS 5-step Simplified Severity Scale for AMD. If the results presented were replicated, then this ensemble of CNNs could be used as a screening tool that has the potential to significantly improve health outcomes by identifying asymptomatic individuals who would benefit from AREDS2 macular supplements. [ABSTRACT FROM AUTHOR]

Purpose: The main purpose of the study was to report the estimated incidence, cumulative rate, risk factors and outcomes of submacular haemorrhage (SMH) with loss of vision in neovascular age‐related macular degeneration (nAMD) receiving intravitreal injections (IVT) of vascular endothelial growth factor (VEGF) inhibitor in routine clinical practice. Methods: Retrospective analysis of treatment‐naïve eyes receiving IVTs of VEGF inhibitors (ranibizumab, aflibercept or bevacizumab) for nAMD from 1 January 2010 to 31 December 2020 that were tracked the Fight Retinal Blindness! registry. Estimated incidence, cumulative rate and hazard ratios (HR) of SMH with loss of vision during treatment were measured using the Poisson regression, Kaplan–Meier survival curves and Cox proportional hazard models. Results: We identified 7642 eyes (6425 patients) with a total of 135 095 IVT over a 10‐year period. One hundred five eyes developed SMH with loss of vision with a rate of 1 per 1283 injections (0.08% 95% confidence interval [95% CI] [0.06; 0.09]). The estimated incidence [95% CI] was 4.6 [3.8; 5.7] SMH with loss of vision per year per 1000 treated patients during the study. The cumulative [95% CI] rate of SMH per patient did not increase significantly with each successive injection (p = 0.947). SMH cases had a mean VA drop of around 6 lines at diagnosis, which then improved moderately to a 4‐line loss at 1 year. Conclusions: Submacular haemorrhage (SMH) with loss of vision is an uncommon complication that can occur at any time in eyes treated for nAMD in routine clinical practice, with only limited recovery of vision 1 year later. [ABSTRACT FROM AUTHOR]

This systematic review was performed to identify the specifics of an optimal diabetic retinopathy deep learning algorithm, by identifying the best exemplar research studies of the field, whilst highlighting potential barriers to clinical implementation of such an algorithm. Searching five electronic databases (Embase, MEDLINE, Scopus, PubMed, and the Cochrane Library) returned 747 unique records on 20 December 2019. Predetermined inclusion and exclusion criteria were applied to the search results, resulting in 15 highest-quality publications. A manual search through the reference lists of relevant review articles found from the database search was conducted, yielding no additional records. A validation dataset of the trained deep learning algorithms was used for creating a set of optimal properties for an ideal diabetic retinopathy classification algorithm. Potential limitations to the clinical implementation of such systems were identified as lack of generalizability, limited screening scope, and data sovereignty issues. It is concluded that deep learning algorithms in the context of diabetic retinopathy screening have reported impressive results. Despite this, the potential sources of limitations in such systems must be evaluated carefully. An ideal deep learning algorithm should be clinic-, clinician-, and camera-agnostic; complying with the local regulation for data sovereignty, storage, privacy, and reporting; whilst requiring minimum human input. [ABSTRACT FROM AUTHOR]

Convolutional Neural Networks (CNNs) have become a prominent method of AI implementation in medical classification tasks. Grading Diabetic Retinopathy (DR) has been at the forefront of the development of AI for ophthalmology. However, major obstacles remain in the generalization of these CNNs onto real-world DR screening programs. We believe these difficulties are due to use of 1) small training datasets (<5,000 images), 2) private and 'curated' repositories, 3) locally implemented CNN implementation methods, while 4) relying on measured Area Under the Curve (AUC) as the sole measure of CNN performance. To address these issues, the public EyePACS Kaggle Diabetic Retinopathy dataset was uploaded onto Microsoft Azure™ cloud platform. Two CNNs were trained; 1 a "Quality Assurance", and 2. a "Classifier". The Diabetic Retinopathy classifier CNN (DRCNN) performance was then tested both on 'un-curated' as well as the 'curated' test set created by the "Quality Assessment" CNN model. Finally, the sensitivity of the DRCNNs was boosted using two post-training techniques. Our DRCNN proved to be robust, as its performance was similar on 'curated' and 'un-curated' test sets. The implementation of 'cascading thresholds' and 'max margin' techniques led to significant improvements in the DRCNN's sensitivity, while also enhancing the specificity of other grades. [ABSTRACT FROM AUTHOR]

Background and Objective. To determine if using a multi-input deep learning approach in the image analysis of optical coherence tomography (OCT), OCT angiography (OCT-A), and colour fundus photographs increases the accuracy of a CNN to diagnose intermediate dry age-related macular degeneration (AMD). Patients and Methods. Seventy-five participants were recruited and divided into three cohorts: young healthy (YH), old healthy (OH), and patients with intermediate dry AMD. Colour fundus photography, OCT, and OCT-A scans were performed. The convolutional neural network (CNN) was trained on multiple image modalities at the same time. Results. The CNN trained using OCT alone showed a diagnostic accuracy of 94%, whilst the OCT-A trained CNN resulted in an accuracy of 91%. When multiple modalities were combined, the CNN accuracy increased to 96% in the AMD cohort. Conclusions. Here we demonstrate that superior diagnostic accuracy can be achieved when deep learning is combined with multimodal image analysis. [ABSTRACT FROM AUTHOR]

Diabetic retinopathy is a vascular disease of the retina characterised by hyperglycaemic and inflammatory processes. Most animal models of diabetic retinopathy are hyperglycaemia-only models that do not account for the significant role that inflammation plays in the development of the disease. In the present study, we present data on the establishment of a new animal model of diabetic retinopathy that incorporates both hyperglycaemia and inflammation. We hypothesized that inflammation may trigger and worsen the development of diabetic retinopathy in a hyperglycaemic environment. Pro-inflammatory cytokines, IL-1β and TNF-α, were therefore injected into the vitreous of non-obese diabetic (NOD) mice. CD1 mice were used as same genetic background controls. Fundus and optical coherence tomography images were obtained before (day 0) as well as on days 2 and 7 after intravitreal cytokine injection to assess vessel dilation and beading, retinal and vitreous hyper-reflective foci and retinal thickness. Astrogliosis and microgliosis were assessed using immunohistochemistry. Results showed that intravitreal cytokines induced vessel dilation, beading, severe vitreous hyper-reflective foci, retinal oedema, increased astrogliosis and microglia upregulation in diabetic NOD mice. Intravitreal injection of inflammatory cytokines into the eyes of diabetic mice therefore appears to provide a new model of diabetic retinopathy that could be used for the study of disease progression and treatment strategies. [ABSTRACT FROM AUTHOR]