A deep learning system (DLS) is a machine learning technology with potential for screening diabetic retinopathy and related eye diseases.
To evaluate the performance of a DLS in detecting referable ...diabetic retinopathy, vision-threatening diabetic retinopathy, possible glaucoma, and age-related macular degeneration (AMD) in community and clinic-based multiethnic populations with diabetes.
Diagnostic performance of a DLS for diabetic retinopathy and related eye diseases was evaluated using 494 661 retinal images. A DLS was trained for detecting diabetic retinopathy (using 76 370 images), possible glaucoma (125 189 images), and AMD (72 610 images), and performance of DLS was evaluated for detecting diabetic retinopathy (using 112 648 images), possible glaucoma (71 896 images), and AMD (35 948 images). Training of the DLS was completed in May 2016, and validation of the DLS was completed in May 2017 for detection of referable diabetic retinopathy (moderate nonproliferative diabetic retinopathy or worse) and vision-threatening diabetic retinopathy (severe nonproliferative diabetic retinopathy or worse) using a primary validation data set in the Singapore National Diabetic Retinopathy Screening Program and 10 multiethnic cohorts with diabetes.
Use of a deep learning system.
Area under the receiver operating characteristic curve (AUC) and sensitivity and specificity of the DLS with professional graders (retinal specialists, general ophthalmologists, trained graders, or optometrists) as the reference standard.
In the primary validation dataset (n = 14 880 patients; 71 896 images; mean SD age, 60.2 2.2 years; 54.6% men), the prevalence of referable diabetic retinopathy was 3.0%; vision-threatening diabetic retinopathy, 0.6%; possible glaucoma, 0.1%; and AMD, 2.5%. The AUC of the DLS for referable diabetic retinopathy was 0.936 (95% CI, 0.925-0.943), sensitivity was 90.5% (95% CI, 87.3%-93.0%), and specificity was 91.6% (95% CI, 91.0%-92.2%). For vision-threatening diabetic retinopathy, AUC was 0.958 (95% CI, 0.956-0.961), sensitivity was 100% (95% CI, 94.1%-100.0%), and specificity was 91.1% (95% CI, 90.7%-91.4%). For possible glaucoma, AUC was 0.942 (95% CI, 0.929-0.954), sensitivity was 96.4% (95% CI, 81.7%-99.9%), and specificity was 87.2% (95% CI, 86.8%-87.5%). For AMD, AUC was 0.931 (95% CI, 0.928-0.935), sensitivity was 93.2% (95% CI, 91.1%-99.8%), and specificity was 88.7% (95% CI, 88.3%-89.0%). For referable diabetic retinopathy in the 10 additional datasets, AUC range was 0.889 to 0.983 (n = 40 752 images).
In this evaluation of retinal images from multiethnic cohorts of patients with diabetes, the DLS had high sensitivity and specificity for identifying diabetic retinopathy and related eye diseases. Further research is necessary to evaluate the applicability of the DLS in health care settings and the utility of the DLS to improve vision outcomes.
Diabetic retinopathy (DR) is the major ocular complication of diabetes mellitus, and is a problem with significant global health impact. Major advances in diagnostics, technology and treatment have ...already revolutionized how we manage DR in the early part of the 21
century. For example, the accessibility of imaging with optical coherence tomography, and the development of anti-vascular endothelial growth factor (VEGF) treatment are just some of the landmark developments that have shaped the DR landscape over the last few decades. Yet, there are still more exciting advances being made. Looking forward to 2030, many of these ongoing developments are likely to further transform the field. First, epidemiologic projections show that the global burden of DR is not only increasing, but also shifting from high-income countries towards middle- and low-income areas. Second, better understanding of disease pathophysiology is placing greater emphasis on retinal neural dysfunction and non-vascular aspects of diabetic retinal disease. Third, a wealth of information is becoming available from newer imaging modalities such as widefield imaging systems and optical coherence tomography angiography. Fourth, artificial intelligence for screening, diagnosis and prognostication of DR will become increasingly accessible and important. Fifth, new pharmacologic agents targeting other non-VEGF-driven pathways, and novel therapeutic strategies such as gene therapy are being developed for DR. Finally, the classification system for diabetic retinal disease will need to be continually updated to keep pace with new developments. In this article, we discuss these major trends in DR that we expect to see in 2030 and beyond.
Background and Purpose
Diabetic retinopathy, a secondary complication of diabetes mellitus, can lead to irreversible vision loss. Currently, no treatment is approved for early phases of diabetic ...retinopathy. Modifications of the expression pattern of miRNAs could be involved in the early retinal damage of diabetic subjects. Therefore, we aimed at identification of dysregulated miRNAs–mRNA interactions that might be biomarkers and pharmacological targets for diagnosis and treatment of early diabetic retinopathy.
Methods
A focused set of miRNAs was predicted through a bioinformatic analysis accessing to Gene Expression Omnibus dataset and enrichment of information approach (GENEMANIA‐Cytoscape). Identification of miRNAs–mRNA interactions was carried out with miRNET analysis. Diabetes was induced in C57BL6J mice by streptozotocin and samples analysed at 5 and 10 weeks after diabetes induction. Retinal ultrastructure of diabetic mice was analysed through electron microscopy. We used Real‐time PCR, western blot analysis, elisa, and immunohistochemistry to study expression of miRNAs and possible targets of dysregulated miRNAs.
Key Results
We found that miR‐20a‐5p, miR‐20a‐3p, miR‐20b, miR‐106a‐5p, miR‐27a‐5p, miR‐27b‐3p, miR‐206‐3p, and miR‐381‐3p were dysregulated in the retina and serum of diabetic mice. VEGF, brain‐derived neurotrophic factor (BDNF), PPAR‐α, and cAMP response element‐binding protein 1 (CREB1) are targets of dysregulated miRNAs, which then modulated protein expression in diabetic retina. We found structural modifications in retinas from diabetic mice.
Conclusions and Implications
Serum and retina of diabetic mice express eight dysregulated miRNAs, which modified the expression of VEGF, BDNF, PPAR‐α, and CREB1, before vasculopathy in diabetic retinas.
Diabetic retinopathy (DR) is the most frequent complication of diabetes. The main risk factors are disease duration, a poor glycemic control, and the presence of hypertension. However, there is an ...important variation in risk which indicates that other factors, such as genetic heritability or glycemic variability, play an important role in accounting for the susceptibility to DR development. Another important concept is that DR is an independent predictor of both microvascular and macrovascular complications. Thus, the presence of DR should be taken into account when evaluating the cardiovascular risk of a diabetic subject. Moreover, the evaluation of retinal neurodegeneration could help to identify those diabetic subjects at risk of cognitive impairment, an emerging complication of the type 2 diabetic population. When evaluating a diabetic subject, the awareness of the presence of DR has also therapeutic implications. In this regard, a worsening of DR could occur after a rapid improvement of blood glucose. In summary, a critical review on the importance of the presence of DR in the general management of subjects with diabetes is provided.
Display omitted
We evaluated whether specifically and directly targeting human antigen R (HuR), a member of embryonic lethal abnormal vision (ELAV) proteins family, may represent a new potential ...therapeutic strategy to manage diabetic retinopathy. Nanosystems loaded with siRNA silencing HuR expression (lipoplexes), consisting of solid lipid nanoparticles (SLN) and liposomes (SUV) were prepared. Photon correlation spectroscopy analysis, Zeta potential measurement and atomic force microscopy (AFM) studies were carried out to characterize the complexation of siRNA with the lipid nanocarriers. Nanosystems were evaluated by using AFM and scanning electron microscopy. The lipoplexes were injected into the eye of streptozotocin (STZ)-induced diabetic rats. Retinal HuR and VEGF levels were detected by Western blot and ELISA, respectively. Retinal histology was also carried out. The results demonstrated that retinal HuR and VEGF are significantly increased in STZ-rats and are blunted by HuR siRNA treatment. Lipoplexes with a weak positive surface charge and with a 4:1 N/P (cationic lipid nitrogen to siRNA phosphate) ratio exert a better transfection efficiency, significantly dumping retinal HuR and VEGF levels. In conclusion, we demonstrated that siRNA can be efficiently delivered into the rat retina using lipid-based nanocarriers, and some of the lipoplexes loaded with siRNA silencing HuR expression are potential candidates to manage retinal diseases.
•Fibrovascular membranes from PDR showed significantly increased M2 macrophage infiltration compared to control retinas.•Single-cell analysis confirmed that HMOX1 was located in M2 macrophages.•The ...levels of HMOX1 were upregulated in the vitreous humor of PDR patients and OIR retinas.•Immunofluorescence staining showed that HMOX1 co-localized with M2 macrophages in the retinas of OIR mice.
The roles of immune cell infiltration and ferroptosis in the progression of proliferative diabetic retinopathy (PDR) remain unclear. To identify upregulated molecules associated with immune infiltration and ferroptosis in PDR, GSE60436 and GSE102485 datasets were downloaded from the Gene Expression Omnibus (GEO). Genes associated with immune cell infiltration were examined through Weighted Gene Co-expression Network Analysis (WGCNA) and CIBERSORT algorithm. Common differentially expressed genes (DEGs) were intersected with ferroptosis-associated and immune cell infiltration-related genes. Localization of cellular expression was confirmed by single-cell analysis of GSE165784 dataset. Findings were validated by qRT-PCR, ELISA, Western blotting, and immunofluorescence staining. As a result, the infiltration of M2 macrophages was significantly elevated in fibrovascular membrane samples from PDR patients than the retinas of control subjects. Analysis of DEGs, M2 macrophage-related genes and ferroptosis-related genes identified three hub intersecting genes, TP53, HMOX1 and PPARA. qRT-PCR showed that HMOX1 was significantly higher in the oxygen-induced retinopathy (OIR) mouse model retinas than in controls. Single-cell analysis confirmed that HMOX1 was located in M2 macrophages. ELISA and western blotting revealed elevated levels of HMOX1 in the vitreous humor of PDR patients and OIR retinas, and immunofluorescence staining showed that HMOX1 co-localized with M2 macrophages in the retinas of OIR mice. This study offers novel insights into the mechanisms associated with immune cell infiltration and ferroptosis in PDR. HMOX1 expression correlated with M2 macrophage infiltration and ferroptosis, which may play a crucial role in PDR pathogenesis.
Worsening of diabetic retinopathy (DR) is associated with the initiation of effective treatment of glycaemia in some patients with diabetes. It has been associated with risk factors such as poor ...blood‐glucose control and hypertension, and it manifests prior to the long‐term benefits of optimizing glycaemic control. The majority of evidence supports an association of large and rapid reductions in blood‐glucose levels with early worsening of DR. Despite a general awareness of early worsening within the diabetes community, mechanisms to explain the phenomenon remain speculative. We provide an overview of early worsening of DR and its pathophysiology based on current data. We describe the phenomenon in various settings, including in patients receiving insulin‐ or non‐insulin‐based treatments, in those undergoing bariatric surgery, and in pregnant women. We discuss various mechanisms and theories that have been suggested to explain this paradoxical phenomenon, and we summarize the implications of these in clinical practice.
Aims
To identify and synthesize studies reporting modifiable barriers/enablers associated with retinopathy screening attendance in people with Type 1 or Type 2 diabetes, and to identify those most ...likely to influence attendance.
Methods
We searched MEDLINE, EMBASE, PsycINFO, Cochrane Library and the ‘grey literature’ for quantitative and qualitative studies to February 2017. Data (i.e. participant quotations, interpretive summaries, survey results) reporting barriers/enablers were extracted and deductively coded into domains from the Theoretical Domains Framework; with domains representing categories of theoretical barriers/enablers proposed to mediate behaviour change. Inductive thematic analysis was conducted within domains to describe the role each domain plays in facilitating or hindering screening attendance. Domains that were more frequently coded and for which more themes were generated were judged more likely to influence attendance.
Results
Sixty‐nine primary studies were included. We identified six theoretical domains ‘environmental context and resources’ (75% of included studies), ‘social influences’ (51%), ‘knowledge’ (51%), ‘memory, attention, decision processes’ (50%), ‘beliefs about consequences’ (38%) and ‘emotions’ (33%) as the key mediators of diabetic retinopathy screening attendance. Examples of barriers populating these domains included inaccurate diabetic registers and confusion between routine eye care and retinopathy screening. Recommendations by healthcare professionals and community‐level media coverage acted as enablers.
Conclusions
Across a variety of contexts, we found common barriers to and enablers of retinopathy screening that could be targeted in interventions aiming to increase screening attendance.
What's new?
Diabetic retinopathy screening is effective but uptake is sub‐optimal.
Theoretical determinants (barriers and enablers) of screening attendance were identified that operate at the level of the person with diabetes (e.g. confusion between retinopathy screening and routine eye care), the healthcare professionals (e.g. lack of recommendation to screen), the healthcare system (e.g. inaccurate registers), and the wider community (e.g. lack of media coverage).
Findings from this study will help to inform which theoretical determinants to target in interventions that seek to improve attendance at diabetic retinopathy screening.
Display omitted
Diabetic retinopathy (DR) is a complication caused by abnormal glucose metabolism, which affects the vision and quality of life of patients and severely impacts the society at ...large.DR has a complex pathogenic process. Evidence from multiple studies have shown that oxidative stress and inflammation play pivotal roles in DR.Additionally, with the rapid development of various genetic detection methods, the abnormal expression of long non-coding RNAs (lncRNAs) have been confirmed to promote the development of DR.Research has demonstrated the potential of lncRNAs as ideal biomarkers and theranostic targets in DR. In this narrative review, we will focus on the research results on mechanisms underlying DR, list lncRNAs confirmed to be closely related to these mechanisms, and discuss their potential clinical application value and limitations.
The prevalence of diabetes in the United States has increased. People with diabetes are at risk for diabetic retinopathy. No recent national population-based estimate of the prevalence and severity ...of diabetic retinopathy exists.
To describe the prevalence and risk factors of diabetic retinopathy among US adults with diabetes aged 40 years and older.
Analysis of a cross-sectional, nationally representative sample of the National Health and Nutrition Examination Survey 2005-2008 (N = 1006). Diabetes was defined as a self-report of a previous diagnosis of the disease (excluding gestational diabetes mellitus) or glycated hemoglobin A(1c) of 6.5% or greater. Two fundus photographs were taken of each eye with a digital nonmydriatic camera and were graded using the Airlie House classification scheme and the Early Treatment Diabetic Retinopathy Study severity scale. Prevalence estimates were weighted to represent the civilian, noninstitutionalized US population aged 40 years and older.
Diabetic retinopathy and vision-threatening diabetic retinopathy.
The estimated prevalence of diabetic retinopathy and vision-threatening diabetic retinopathy was 28.5% (95% confidence interval CI, 24.9%-32.5%) and 4.4% (95% CI, 3.5%-5.7%) among US adults with diabetes, respectively. Diabetic retinopathy was slightly more prevalent among men than women with diabetes (31.6%; 95% CI, 26.8%-36.8%; vs 25.7%; 95% CI, 21.7%-30.1%; P = .04). Non-Hispanic black individuals had a higher crude prevalence than non-Hispanic white individuals of diabetic retinopathy (38.8%; 95% CI, 31.9%-46.1%; vs 26.4%; 95% CI, 21.4%-32.2%; P = .01) and vision-threatening diabetic retinopathy (9.3%; 95% CI, 5.9%-14.4%; vs 3.2%; 95% CI, 2.0%-5.1%; P = .01). Male sex was independently associated with the presence of diabetic retinopathy (odds ratio OR, 2.07; 95% CI, 1.39-3.10), as well as higher hemoglobin A(1c) level (OR, 1.45; 95% CI, 1.20-1.75), longer duration of diabetes (OR, 1.06 per year duration; 95% CI, 1.03-1.10), insulin use (OR, 3.23; 95% CI, 1.99-5.26), and higher systolic blood pressure (OR, 1.03 per mm Hg; 95% CI, 1.02-1.03).
In a nationally representative sample of US adults with diabetes aged 40 years and older, the prevalence of diabetic retinopathy and vision-threatening diabetic retinopathy was high, especially among Non-Hispanic black individuals.