Comparison of 1-field, 2-fields, and 3-fields fundus photography for detection and grading of diabetic retinopathy

Diagnostic Accuracy of Artificial Intelligence-Based Automated Diabetic Retinopathy Screening in Real-World Settings: A Systematic Review and Meta-Analysis

PURPOSE

To evaluate the diagnostic accuracy of artificial intelligence (AI)-based automated diabetic retinopathy (DR) screening in real-world settings.

DESIGN

Systematic review and meta-analysis METHODS: We conducted a systematic review of relevant literature from January 2012 to August 2022 using databases including PubMed, Scopus and Web of Science. The quality of studies was evaluated using Quality Assessment for Diagnostic Accuracy Studies 2 (QUADAS-2) checklist. We calculated pooled accuracy, sensitivity, specificity, and diagnostic odds ratio (DOR) as summary measures. The study protocol was registered with the International Prospective Register of Systematic Reviews (PROSPERO - CRD42022367034).

RESULTS

We included 34 studies which utilized AI algorithms for diagnosing DR based on real-world fundus images. Quality assessment of these studies indicated a low risk of bias and low applicability concern. Among gradable images, the overall pooled accuracy, sensitivity, specificity, and DOR were 81%, 94% (95% CI: 92.0-96.0), 89% (95% CI: 85.0-92.0) and 128 (95% CI: 80-204) respectively. Sub-group analysis showed that, when acceptable quality imaging could be obtained, non-mydriatic fundus images had a better DOR of 143 (95% CI: 82-251) and studies using 2 field images had a better DOR of 161 (95% CI 74-347). Our meta-regression analysis revealed a statistically significant association between DOR and variables such as the income status, and the type of fundus camera.

CONCLUSION

Our findings indicate that AI algorithms have acceptable performance in screening for DR using fundus images compared to human graders. Implementing a fundus camera with AI-based software has the potential to assist ophthalmologists in reducing their workload and improving the accuracy of DR diagnosis.

Quantification of Microvascular Lesions in the Central Retinal Field: Could It Predict the Severity of Diabetic Retinopathy?

Diabetic retinopathy (DR) is a neurodegenerative disease characterized by the presence of microcirculatory lesions. Among them, microaneurysms (MAs) are the first observable hallmark of early ophthalmological changes. The present work aims to study whether the quantification of MAs, hemorrhages (Hmas) and hard exudates (HEs) in the central retinal field could have a predictive value on DR severity. These retinal lesions were quantified in a single field NM-1 of 160 retinographies of diabetic patients from the IOBA's reading center. Samples included different disease severity levels and excluded proliferating forms: no DR (n = 30), mild non-proliferative (n = 30), moderate (n = 50) and severe (n = 50). Quantification of MAs, Hmas, and HEs revealed an increasing trend as DR severity progresses. Differences between severity levels were statistically significant, suggesting that the analysis of the central field provides valuable information on severity level and could be used as a clinical tool to assess DR grading in the eyecare routine. Even though further validation is needed, counting microvascular lesions in a single retinal field can be proposed as a rapid screening system to classify DR patients with different stages of severity according to the international classification.

In a Large Healthcare System in the Bronx, Teleretinal Triaging Was Found to Increase Screening and Healthcare Access for an Underserved Population with a High Incidence of T2DM and Retinopathy

The early treatment of diabetic retinopathy (DR) prevents vision-threatening proliferative retinopathy (PDR) and macular edema (DME). Our study evaluates telemedicine (teleretinal) screening for DR in an inner-city healthcare network with a high ethnic diversity and disease burden. Fundus photographs were obtained and graded in a centralized reading center between 2014 and 2016. Patients with positive screenings were referred to a retina specialist. An analysis of sensitivity and specificity and a subgroup analysis of prevalence, disease severity, and follow-up adherence were conducted. In 2251 patients, the '1-year' and 'Overall' follow-ups were 35.1% and 54.8%, respectively. Severe grading, male gender, and age were associated with better follow-up compliance. The DR, PDR, and DME prevalence was 24.9%, 4.1%, and 5.9%, respectively, and was significantly associated with HbA1c. The sensitivity and specificity for DR, PDR, and DME were 70% and 87%, 87% and 75%, and 37% and 95%, respectively. No prevalence differences were noted between ethnicities. Annual diabetic eye exam adherence increased from 55% to 85% during the study period. Teleretinal triaging is sensitive and specific for DR and improved diabetic eye exam compliance for underserved populations when integrated into large healthcare networks. The adherence to follow-up recommendations was better among older patients and among those with more severe retinopathy.

The Cost-Effectiveness of a Telemedicine Screening Program for Diabetic Retinopathy in New York City

Background

A telemedicine screening initiative was implemented by the Montefiore Health System to improve access to eyecare for a multi-ethnic, at-risk population of diabetic patients in a largely underserved urban community in the Bronx, New York. This retrospective, cross-sectional analysis evaluates the societal benefit and financial sustainability of this program by analyzing both cost and revenue generation based on current standard Medicare reimbursement rates.

Methods

Non-mydriatic fundus cameras were placed in collaboration with a vendor in eight outpatient primary care sites throughout the Montefiore Health Care System, and data was collected between June 2014 and July 2016. Fundus photos were electronically transmitted to a central reading center to be systematically reviewed and coded by faculty ophthalmologists, and patients were subsequently scheduled for ophthalmic evaluation based upon a predetermined treatment algorithm. A retrospective chart review of 2251 patients was performed utilizing our electronic medical record system (Epic Systems, Verona WI). Revenue was projected utilizing standard Medicare rates for our region while societal benefit was calculated using quality adjusted life years (QALY).

Results

Of the 2251 patient charts reviewed, 791 patients (35.1%) were seen by Montefiore ophthalmologists within a year of the original screening date. Estimated revenue generated by these visits was $276,800, with the majority from the treatment of retinal disease ($208,535), and the remainder from other ophthalmic conditions detected in the fundus photos ($68,265). There was a societal benefit of 14.66 quality adjusted life years (QALYs) with an estimated value of $35,471/QALY.

Conclusion

This telemedicine initiative was successful in identifying many patients with diabetic retinopathy and other ophthalmic conditions who may otherwise not have been formally evaluated. Our analysis demonstrates the program to generate a downstream revenue of nearly $280K with a cost benefit below <50% of the threshold of $100,000/QALY, and therefore cost-effective in marginalized communities.

Real-world outcomes of a clinical decision support system for diabetic retinopathy in Spain

Objective

The aim of present study was to evaluate our clinical decision support system (CDSS) for predicting risk of diabetic retinopathy (DR). We selected randomly a real population of patients with type 2 diabetes (T2DM) who were attending our screening programme.

Methods and analysis

The sample size was 602 patients with T2DM randomly selected from those who attended the DR screening programme. The algorithm developed uses nine risk factors: current age, sex, body mass index (BMI), duration and treatment of diabetes mellitus (DM), arterial hypertension, Glicated hemoglobine (HbA1c), urine–albumin ratio and glomerular filtration.

Results

The mean current age of 67.03±10.91, and 272 were male (53.2%), and DM duration was 10.12±6.4 years, 222 had DR (35.8%). The CDSS was employed for 1 year. The prediction algorithm that the CDSS uses included nine risk factors: current age, sex, BMI, DM duration and treatment, arterial hypertension, HbA1c, urine–albumin ratio and glomerular filtration. The area under the curve (AUC) for predicting the presence of any DR achieved a value of 0.9884, the sensitivity of 98.21%, specificity of 99.21%, positive predictive value of 98.65%, negative predictive value of 98.95%, α error of 0.0079 and β error of 0.0179.

Conclusion

Our CDSS for predicting DR was successful when applied to a real population.

AUTOMATED SCREENING OF DIABETIC RETINOPATHY WITH OPTIMIZED DEEP CONVOLUTIONAL NEURAL NETWORK: ENHANCED MOTH FLAME MODEL

Nowadays, analysis on retinal image exists as one of the challenging area for study. Numerous retinal diseases could be recognized by analyzing the variations taking place in retina. However, the main disadvantage among those studies is that, they do not have higher recognition accuracy. The proposed framework includes four phases namely, (i) Blood Vessel Segmentation (ii) Feature Extraction (iii) Optimal Feature Selection and (iv) Classification. Initially, the input fundus image is subjected to blood vessel segmentation from which two binary thresholded images (one from High Pass Filter (HPF) and other from top-hat reconstruction) are acquired. These two images are differentiated and the areas that are common to both are said to be the major vessels and the left over regions are fused to form vessel sub-image. These vessel sub-images are classified with Gaussian Mixture Model (GMM) classifier and the resultant is summed up with the major vessels to form the segmented blood vessels. The segmented images are subjected to feature extraction process, where the features like proposed Local Binary Pattern (LBP), Gray-Level Co-Occurrence Matrix (GLCM) and Gray Level Run Length Matrix (GLRM) are extracted. As the curse of dimensionality seems to be the greatest issue, it is important to select the appropriate features from the extracted one for classification. In this paper, a new improved optimization algorithm Moth Flame with New Distance Formulation (MF-NDF) is introduced for selecting the optimal features. Finally, the selected optimal features are subjected to Deep Convolutional Neural Network (DCNN) model for classification. Further, in order to make the precise diagnosis, the weights of DCNN are optimally tuned by the same optimization algorithm. The performance of the proposed algorithm will be compared against the conventional algorithms in terms of positive and negative measures.

Cross-Sectional Analysis of the Involvement of Interleukin-17A in Diabetic Retinopathy in Elderly Individuals with Type 2 Diabetes Mellitus

BACKGROUND

To investigate the correlation between serum interleukin-17A (IL-17A) levels and diabetic retinopathy (DR) in elderly individuals with type 2 diabetes mellitus (T2DM).

METHODS

The study included 194 elderly patients (94 males and 100 females) with T2DM. Digital retinal photography as well as fundus fluorescein angiography was employed to distinguish between nonproliferative diabetic retinopathy (NPDR) and proliferative diabetic retinopathy (PDR). In addition, multiple logistic regression analysis was conducted to determine the correlation between serum IL-17A levels and DR status.

RESULTS

The average age of the study cohort was 69.14 ± 6.33 years, of which 52.08% were male. The study participants with the highest IL-17A (Q4) levels had higher TC, DBP, and low-density lipoprotein cholesterol (LDL-C) values than those the other groups. Analysis using unadjusted and adjusted linear regression revealed that the effect size of 1.09 for DR in the unadjusted model indicates that IL-17A is associated with an increase of 1.09 in DR (mmol/L) (β 1.09, 95% confidence interval (CI) 1.03, 1.16). Using the minimum-adjusted model (the model 2), as IL-17A increased, DR was higher by 1.11 (β 1.11, 95% CI 1.04, 1.18). With the fully adjusted model (the model 3), for each additional IL-17A increase, DR was higher by 1.15 (β 1.15, 95% CI 1.06, 1.24).

CONCLUSION

Serum IL-17A levels are apparently positively correlated to DR in Chinese elderly individuals with T2DM.

A Review of Advancements and Evidence Gaps in Diabetic Retinopathy Screening Models

Diabetic retinopathy (DR) is a microvascular complication of diabetes with a prevalence of ~35%, and is one of the leading causes of visual impairment in people of working age in most developed countries. The earliest stage of DR, non-proliferative DR (NPDR), may progress to sight-threatening DR (STDR). Thus, early detection of DR and active regular screening of patients with diabetes are necessary for earlier intervention to prevent sight loss. While some countries offer systematic DR screening, most nations are reliant on opportunistic screening or do not offer any screening owing to limited healthcare resources and infrastructure. Currently, retinal imaging approaches for DR screening include those with and without mydriasis, imaging in single or multiple fields, and the use of conventional or ultra-wide-field imaging. Advances in telescreening and automated detection facilitate screening in previously hard-to-reach communities. Despite the heterogeneity in approaches to fit local needs, an evidence base must be created for each model to inform practice. In this review, we appraise different aspects of DR screening, including technological advances, identify evidence gaps, and propose several studies to improve DR screening globally, with a view to identifying patients with moderate-to-severe NPDR who would benefit if a convenient treatment option to delay progression to STDR became available.

Screening for diabetic retinopathy: new perspectives and challenges

Although the prevalence of all stages of diabetic retinopathy has been declining since 1980 in populations with improved diabetes control, the crude prevalence of visual impairment and blindness caused by diabetic retinopathy worldwide increased between 1990 and 2015, largely because of the increasing prevalence of type 2 diabetes, particularly in low-income and middle-income countries. Screening for diabetic retinopathy is essential to detect referable cases that need timely full ophthalmic examination and treatment to avoid permanent visual loss. In the past few years, personalised screening intervals that take into account several risk factors have been proposed, with good cost-effectiveness ratios. However, resources for nationwide screening programmes are scarce in many countries. New technologies, such as scanning confocal ophthalmology with ultrawide field imaging and handheld mobile devices, teleophthalmology for remote grading, and artificial intelligence for automated detection and classification of diabetic retinopathy, are changing screening strategies and improving cost-effectiveness. Additionally, emerging evidence suggests that retinal imaging could be useful for identifying individuals at risk of cardiovascular disease or cognitive impairment, which could expand the role of diabetic retinopathy screening beyond the prevention of sight-threatening disease.