Heyingwuzi formulation alleviates diabetic retinopathy by promoting mitophagy via the HIF-1α/BNIP3/NIX axis.

BACKGROUND: Diabetic retinopathy (DR) is the primary cause of visual problems in patients with diabetes. The Heyingwuzi formulation (HYWZF) is effective against DR. AIM: To determine the HYWZF prevention mechanisms, especially those underlying mitophagy. METHODS: Human retinal capillary endothelial cells (HRCECs) were treated with high glucose (hg), HYWZF serum, PX-478, or Mdivi-1 in vitro. Then, cell counting kit-8, transwell, and tube formation assays were used to evaluate HRCEC proliferation, invasion, and tube formation, respectively. Transmission electron microscopy was used to assess mitochondrial morphology, and Western blotting was used to determine the protein levels. Flow cytometry was used to assess cell apoptosis, reactive oxygen species (ROS) production, and mitochondrial membrane potential. Moreover, C57BL/6 mice were established in vivo using streptozotocin and treated with HYWZF for four weeks. Blood glucose levels and body weight were monitored continuously. Changes in retinal characteristics were evaluated using hematoxylin and eosin, tar violet, and periodic acid-Schiff staining. Protein levels in retinal tissues were determined via Western blotting, immunohistochemistry, and immunostaining. RESULTS: HYWZF inhibited excessive ROS production, apoptosis, tube formation, and invasion in hg-induced HRCECs via mitochondrial autophagy in vitro. It increased the mRNA expression levels of BCL2-interacting protein 3 (BNIP3), FUN14 domain-containing 1, BNIP3-like (BNIP3L, also known as NIX), PARKIN, PTEN-induced kinase 1, and hypoxia-inducible factor (HIF)-1α. Moreover, it downregulated the protein levels of vascular endothelial cell growth factor and increased the light chain 3-II/I ratio. However, PX-478 and Mdivi-1 reversed these effects. Additionally, PX-478 and Mdivi-1 rescued the effects of HYWZF by decreasing oxidative stress and apoptosis and increasing mitophagy. HYWZF intervention improved the symptoms of diabetes, tissue damage, number of acellular capillaries, and oxidative stress in vivo. Furthermore, in vivo experiments confirmed the results of in vitro experiments. CONCLUSION: HYWZF alleviated DR and associated damage by promoting mitophagy via the HIF-1α/BNIP3/NIX axis.

Intraocular lens selection in diabetic patients: How to increase the odds for success.

The incidence of cataracts is significantly higher in diabetic individuals, particularly in younger age groups, with rates quadrupled in those under 65 and doubled in those over 65 compared to non-diabetics. Cataract surgery in diabetic patients poses many challenges: Poor epithelial healing, decreased corneal sensitivity, increased central corneal thickness, decreased endothelial cell count, variable topography, poor pupillary dilatation, anterior capsular phimosis, posterior capsular opacification (PCO), chances of progression of diabetic retinopathy (DR), zonular weakness, and vitreous prolapse and diabetic macular edema. Selection of an appropriate intraocular lens (IOL) is crucial for visual rehabilitation and monitoring DR. The choice of IOL in diabetic cataract patients is a challenging scenario. Square-edge IOLs are favored for their capacity to mitigate PCO, whereas hydrophilic counterparts may incur calcification in the setting of proliferative DR. The advisability of premium IOLs for achieving spectacle independence warrants judicious evaluation, particularly in the presence of advanced retinopathy. Optimal IOL placement within the capsular bag is advocated to minimize postoperative complications. Rigorous preoperative assessment and informed patient counseling regarding IOL options are indispensable for optimizing surgical outcomes. This review article covers various aspects regarding the choice of IOLs in different case scenarios and complications in the diabetic population.

Single-Cell RNA Sequencing Reveals Transcriptional Signatures and Cell-Cell Communication in Diabetic Retinopathy.

BACKGROUND: Diabetic retinopathy (DR) is a major cause of vision loss in workingage individuals worldwide. Cell-to-cell communication between retinal cells and retinal pigment epithelial cells (RPEs) in DR is still unclear, so this study aimed to generate a single-cell atlas and identify receptor‒ligand communication between retinal cells and RPEs. METHODS: A mouse single-cell RNA sequencing (scRNA-seq) dataset was retrieved from the GEO database (GSE178121) and was further analyzed with the R package Seurat. Cell cluster annotation was performed to further analyze cell‒cell communication. The differentially expressed genes (DEGs) in RPEs were explored through pathway enrichment analysis and the protein‒ protein interaction (PPI) network. Core genes in the PPI were verified by quantitative PCR in ARPE-19 cells. RESULTS: We observed an increased proportion of RPEs in STZ mice. Although some overall intercellular communication pathways did not differ significantly in the STZ and control groups, RPEs relayed significantly more signals in the STZ group. In addition, THBS1, ITGB1, COL9A3, ITGB8, VTN, TIMP2, and FBN1 were found to be the core DEGs of the PPI network in RPEs. qPCR results showed that the expression of ITGB1, COL9A3, ITGB8, VTN, TIMP2, and FBN1 was higher and consistent with scRNA-seq results in ARPE-19 cells under hyperglycemic conditions. CONCLUSION: Our study, for the first time, investigated how signals that RPEs relay to and from other cells underly the progression of DR based on scRNA-seq. These signaling pathways and hub genes may provide new insights into DR mechanisms and therapeutic targets.

The circular RNA circSLC16A10 alleviates diabetic retinopathy by improving mitochondrial function via the miR-761-5p/MFN2 axis.

It has been demonstrated that circular RNAs (circRNAs) are associated with the development of diabetic retinopathy (DR). Nevertheless, the function of circSLC16A10 in the development of DR remains unclear. In order to investigate the role of circSLC16A10, we employed cell and animal models of DR. An analysis of a public database revealed that hsa_circSLC16A10 was expressed at lower levels in DR patients than in diabetic patients without DR or healthy controls. Additionally, the level of hsa_circSLC16A10 was lower in high glucose (HG)-exposed ARPE-19 cells and diabetic mice. hsa_circSLC16A10 was observed to be mainly distributed in the cytoplasm. Moreover, overexpression of hsa_circSLC16A10 alleviated HG-induced endoplasmic reticulum stress and cell apoptosis in vitro. Furthermore, overexpression of hsa_circSLC16A10 ameliorated HG-induced mitochondrial dysfunction, as evidenced by improvements in mitochondrial structure and function. Hsa_circSLC16A10 acted as a hsa-miR-761-5p sponge to increase MFN2 expression. MFN2 knockdown or hsa-miR-761-5p overexpression partially reversed the protective effect of hsa_circSLC16A10 in vitro. The protective effect of mmu_circSLC16A10 against DR was confirmed in an animal model of DR. These findings indicate that circSLC16A10 may regulate DR progression by improving mitochondrial function via the miR-761-5p/MFN2 axis.

Quantitative Biomarkers of Diabetic Retinopathy Using Ultra-Widefield Fluorescein Angiography.

Introduction: Diabetic retinopathy (DR) is a leading cause of blindness. Retinal imaging is an important tool to monitor the progression of DR. While seven-standard retinal fields are the traditional method for evaluating DR, ultra-widefield (UWF) imaging allows for improved visualization of peripheral areas of nonperfusion (NP) and neovascularization (NV), which could be used as biomarkers to monitor and predict progression of DR. Methods: A retrospective, cross-sectional study was conducted on 651 eyes from 363 patients diagnosed with type 1 or type 2 diabetes who received UWF-FA over 10 years. Fluorescein Angiography (FA) images were segmented, and surface areas of NP and NV were analyzed using multivariate regression to determine if biomarkers of DR and DR severity are associated with increasing areas of NP and NV. Results: Each additional year with a diagnosis of DR was associated with a 10.75 mm2 increase in the total NP (95% CI, 1.94-19.56; P = 0.02) and 7.87 mm2 increase in NP far-periphery (95% CI, 1.62-14.13; P = 0.01). A one-unit change in severity as defined by the Early Treatment of Diabetic Retinopathy Study (ETDRS) classification was associated with a 25.75 mm2 increase in total NP (95% CI, 11.16-40.33; P = 0.001), a 13.15 mm2 increase in mid-periphery NP (95% CI, 6.93-19.38; P < 0.0001), and a 12.29 mm2 increase in far-periphery NP (95% CI, 3.62-20.97; P = 0.01). Discussion: Biomarkers identified through UWF imaging such as total and regional areas of NP can be used to monitor and predict the progression of DR. This may provide a quantitative method for prognostication in patients with DR.

The effect of the intercellular adhesion molecule-1 and glycated haemoglobin on the management of diabetic neovascular glaucoma.

Introduction: The study hypothesizes that some patients with diabetic neovascular glaucoma (NVG) do not fully respond to transscleral (TSC) cyclophotocoagulation (CPC) due to significant inflammation and insufficient glucose control. Objective: The study aimed to determine the effect of baseline blood levels of intercellular adhesion molecule-1 (ICAM-1) and glycated haemoglobin (HbA1c) on the management of patients with diabetic NVG by TSC CPC. Methods: This open prospective study included 70 diabetic patients (75 eyes; aged Ме 63.0 years) with painful NVG and 20 healthy individuals (aged Ме 61.5 years) as an immunological control. All patients underwent TSC СPC with a diode laser. Baseline HbA1c levels and ICAM-1 expression in blood samples were determined. Follow-up was 12 months. Results: One month after TSC CPC, IOP decreased by 28% compared to baseline. The effectiveness of laser treatment after 12 months of follow-up was 63% with IOP decrease by 46%. In patients with NVG, the initial level of ICAM-1 was 2.5 times higher than in the control group. Patients who did not fully respond to the first TSC CPC (30 eyes) and required additional laser procedure, had high initial HbA1c (9.5%) and high expression values of the ICAM-1 (609.0 cells/µL). Conclusions: Repeated procedures of TSC CPC at high IOP in diabetic patients with NVG are associated with high initial values of expression of ICAM-1 in peripheral blood and high HbA1c. The strategy of management of patients with diabetic NVG should be aimed at intensive glucose control and local anti-inflammatory treatment. Abbreviations: PDR = proliferative diabetic retinopathy, DR = diabetic retinopathy, NVG = neovascular glaucoma, TSC CPC = transscleral cyclophotocoagulation, ICAM-1 = intercellular adhesion molecule-1, HbA1c = glycated haemoglobin, IOP = intraocular pressure.

Rationale of Basic and Cellular Mechanisms Considered in Updating the Staging System for Diabetic Retinal Disease.

Purpose: Hyperglycemia is a major risk factor for early lesions of diabetic retinal disease (DRD). Updating the DRD staging system to incorporate relevant basic and cellular mechanisms pertinent to DRD is necessary to better address early disease, disease progression, the use of therapeutic interventions, and treatment effectiveness. Design: We sought to review preclinical and clinical evidence on basic and cellular mechanisms potentially pertinent to DRD that might eventually be relevant to update the DRD staging system. Participants: Not applicable. Methods: The Basic and Cellular Mechanisms Working Group (BCM-WG) of the Mary Tyler Moore Vision Initiative carefully and extensively reviewed available preclinical and clinical evidence through multiple iterations and classified these. Main Outcome Measures: Classification was made into evidence grids, level of supporting evidence, and anticipated future relevance to DRD. Results: A total of 40 identified targets based on pathophysiology and other parameters for DRD were grouped into concepts or evaluated as specific candidates. VEGFA, peroxisome proliferator-activated receptor-alpha related pathways, plasma kallikrein, and angiopoietin 2 had strong agreement as promising for use as biomarkers in diagnostic, monitoring, predictive, prognostic, and pharmacodynamic responses as well as for susceptibility/risk biomarkers that could underlie new assessments and eventually be considered within an updated DRD staging system or treatment, based on the evidence and need for research that would fit within a 2-year timeline. The BCM-WG found there was strong reason also to pursue the following important concepts regarding scientific research of DRD acknowledging their regulation by hyperglycemia: inflammatory/cytokines, oxidative signaling, vasoprotection, neuroprotection, mitophagy, and nutrients/microbiome. Conclusion: Promising targets that might eventually be considered within an updated DRD staging system or treatment were identified. Although the BCM-WG recognizes that at this stage little can be incorporated into a new DRD staging system, numerous potential targets and important concepts deserve continued support and research, as they may eventually serve as biomarkers and/or therapeutic targets with measurable benefits to patients with diabetes. Financial Disclosures: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Relationship Between the Hemoglobin, Albumin, Lymphocyte Count, Platelet Count (HALP) Score and Type 2 Diabetes Retinopathy.

Purpose: To explore the correlation between hemoglobin, albumin, lymphocyte count, platelet count (HALP) score and type 2 diabetic retinopathy (DR). Methods: The study was conducted on 674 patients with type 2 diabetes (T2DM). According to the results of the fundus examination, they were divided into non-diabetic retinopathy group (NDR, n=388) and diabetic retinopathy group (DR, n=286). Collected patients baseline data, calculated HALP score, analyzed the correlation between HALP score and DR. Results: In all patients, male patients and female patients, the HALP score of the DR group was lower than that of the NDR group (P<0.001), and the HALP score was negatively correlated with the incidence of DR (P<0.05). HALP score was independent risk factors for DR, regardless of gender. In male patients, HALP score and DR had a linear relationship, but in female patients, HALP score and DR showed a nonlinear relationship, and HALP score was more sensitive to the onset of DR in male patients. The combined diagnostic model of HALP score, course of disease, SBP and BUN was used to diagnose DR, and it was found that the diagnostic value was the highest among male patients, with AUC of 0.761, sensitivity of 58.3% and specificity of 80.3%. Conclusion: HALP score was an independent risk factor for DR, attention should be paid to monitoring HALP score, especially in male T2DM patients. The accuracy of HALP score, disease course, SBP and BUN combined model diagnosis of DR was high, which can become a biological indicator for early screening of DR.

Prediction of proliferative diabetic retinopathy using machine learning in Latino and non-Hispanic black cohorts with routine blood and urine testing.

PURPOSE: The objective was to predict proliferative diabetic retinopathy (PDR) in non-Hispanic Black (NHB) and Latino (LA) patients by applying machine learning algorithms to routinely collected blood and urine laboratory results. METHODS: Electronic medical records of 1124 type 2 diabetes patients treated at the Bronxcare Hospital eye clinic between January and December 2019 were analysed. Data collected included demographic information (ethnicity, age and sex), blood (fasting glucose, haemoglobin A1C [HbA1c] high-density lipoprotein [HDL], low-density lipoprotein [LDL], serum creatinine and estimated glomerular filtration rate [eGFR]) and urine (albumin-to-creatinine ratio [ACR]) test results and the outcome measure of retinopathy status. The efficacy of different machine learning models was assessed and compared. SHapley Additive exPlanations (SHAP) analysis was employed to evaluate the contribution of each feature to the model's predictions. RESULTS: The balanced random forest model surpassed other models in predicting PDR for both NHB and LA cohorts, achieving an AUC (area under the curve) of 83%. Regarding sex, the model exhibited remarkable performance for the female LA demographic, with an AUC of 87%. The SHAP analysis revealed that PDR-related factors influenced NHB and LA patients differently, with more pronounced disparity between sexes. Furthermore, the optimal cut-off values for these factors showed variations based on sex and ethnicity. CONCLUSIONS: This study demonstrates the potential of machine learning in identifying individuals at higher risk for PDR by leveraging routine blood and urine test results. It allows clinicians to prioritise at-risk individuals for timely evaluations. Furthermore, the findings emphasise the importance of accounting for both ethnicity and sex when analysing risk factors for PDR in type 2 diabetes individuals.

Mendelian randomization study and mediation analysis about the relation of inflammatory bowel disease and diabetic retinopathy: the further exploration of gut-retina axis.

Background: The concept of the gut-retinal axis proposed by previous scholars primarily focused on the relationship between intestinal microbiota and retinal diseases, and few further expanded the relationship between intestinal diseases and retinal diseases. To further substantiate the concept of the gut-retinal axis, we analyzed inflammatory bowel disease (IBD) and diabetic retinopathy (DR) using Mendelian randomization (MR), and use mediation analysis to further explore the potential substances that influence this causal relationship. Methods: The genome-wide association study's (GWAS) summary statistics for genetic variations were utilized in a Mendelian randomization (MR) investigation. GWAS data on IBD (including ulcerative colitis (UC), Crohn's disease (CD), and IBD) for non-Finnish Europeans (NFE) were sourced from published articles. In contrast, data on DR (including DR and diabetic maculopathy (DMP)) were obtained from FinnGen R9. The causal relationship has been investigated using inverse variance weighted (IVW), MR-Egger, and weighted median and sensitivity analysis was applied to verify the stability of the results. In addition, we applied mediation analysis to investigate whether circulating inflammatory proteins and plasma lipids played a mediating role, and calculated its effect ratio. Results: The causal relationship between IBD and DR was discovered by employing the inverse variance weighted (IVW) method and weighted median method. In forward MR, UC was significantly associated with lower risk of DR (IVW: OR=0.874; 95%CI= 0.835-0.916; P value= 1.28E-08) (Weighted median: OR=0.893; 95%CI= 0.837-0.954; P value= 7.40E-04). In reverse MR, it was shown that DR (IVW: OR=0.870; 95%CI= 0.828-0.914; P value= 2.79E-08)(Weighted median: OR=0.857; 95%CI= 0.801-0.916; P value= 6.40E-06) and DMP (IVW: OR=0.900; 95%CI= 0.865-0.937; P value= 3.34E-07)(Weighted median: OR=0.882; 95%CI= 0.841-0.924; P value= 1.82E-07) could reduce the risk of CD. What's more, DR is associated with a lower risk of IBD according to genetic prediction (IVW: OR=0.922; 95%CI= 0.873-0.972; P value= 0.002) (Weighted median: OR=0.924; 95%CI= 0.861-0.992; P value= 0.029). Fibroblast growth factor 21 (FGF21), phosphatidylcholine (PC), and triacylglycerol (TG) serve as mediators in these relationships. Conclusions: Our research offers novel insights and sources for investigating the gut-retina axis in the genetic relationship between IBD and DR. We discover four mediators and more about the association between the intestine and retinal disorders and provide more evidence for the gut-retinal axis theory.

Clinical serum lipidomic profiling revealed potential lipid biomarkers for early diabetic retinopathy.

Diabetic retinopathy (DR) is a serious complication of diabetes featuring abnormal lipid metabolism. However, the specific lipid molecules associated with onset and progression remain unclear. We used a broad-targeted lipidomics approach to assess the lipid changes that occur before the proliferative retinopathy stage and to identify novel lipid biomarkers to distinguish between patients without DR (NDR) and with non-proliferative DR (NPDR). Targeted lipomics analysis was carried out on serum samples from patients with type I diabetes, including 20 NDRs and 20 NPDRs. The results showed that compared with the NDR group, 102 lipids in the NPDR group showed specific expressions. Four lipid metabolites including TAG58:2-FA18:1 were obtained using the Least Absolute Shrink And Selection Operator (LASSO) and Support Vector Machine Recursive Feature Elimination (SVM-RFE) methods. The four-lipid combination diagnostic models showed good predictive ability in both the discovery and validation sets, and were able to distinguish between NDR patients and NPDR patients. The identified lipid markers significantly improved diagnostic accuracy within the NPDR group. Our findings help to better understand the complexity and individual differences of DR lipid metabolism.

Identification of key biomarkers for early warning of diabetic retinopathy using BP neural network algorithm and hierarchical clustering analysis.

Diabetic retinopathy is one of the most common microangiopathy in diabetes, essentially caused by abnormal blood glucose metabolism resulting from insufficient insulin secretion or reduced insulin activity. Epidemiological survey results show that about one third of diabetes patients have signs of diabetic retinopathy, and another third may suffer from serious retinopathy that threatens vision. However, the pathogenesis of diabetic retinopathy is still unclear, and there is no systematic method to detect the onset of the disease and effectively predict its occurrence. In this study, we used medical detection data from diabetic retinopathy patients to determine key biomarkers that induce disease onset through back propagation neural network algorithm and hierarchical clustering analysis, ultimately obtaining early warning signals of the disease. The key markers that induce diabetic retinopathy have been detected, which can also be used to explore the induction mechanism of disease occurrence and deliver strong warning signal before disease occurrence. We found that multiple clinical indicators that form key markers, such as glycated hemoglobin, serum uric acid, alanine aminotransferase are closely related to the occurrence of the disease. They respectively induced disease from the aspects of the individual lipid metabolism, cell oxidation reduction, bone metabolism and bone resorption and cell function of blood coagulation. The key markers that induce diabetic retinopathy complications do not act independently, but form a complete module to coordinate and work together before the onset of the disease, and transmit a strong warning signal. The key markers detected by this algorithm are more sensitive and effective in the early warning of disease. Hence, a new method related to key markers is proposed for the study of diabetic microvascular lesions. In clinical prediction and diagnosis, doctors can use key markers to give early warning of individual diseases and make early intervention.

Peripheral retinal lesions in diabetic retinopathy on ultra-widefield imaging.

Peripheral retinal imaging plays a crucial role in the diagnosis, management, and prognosis of diabetic retinopathy (DR). Traditional fundus imaging techniques have limited coverage of the retina, resulting in missed peripheral lesions. The advent of ultra-widefield (UWF) imaging has revolutionized the assessment of the peripheral retina. UWF imaging modalities provide comprehensive visualization of the retina, enabling the detection of peripheral lesions without the need for mydriasis. Integration of UWF imaging with other modalities, including fluorescein angiography (FA), indocyanine green angiography, pseudocolor imaging, and fundus autofluorescence, further enhances our understanding of peripheral retinal lesions. UWF imaging has demonstrated improved detection of DR lesions and presumably more accurate management of DR compared to traditional fundus photography and dilated fundus examination. UWF-FA and UWF-optical coherence tomography angiography have emerged as valuable tools for assessing retinal and choroidal vascular abnormalities, nonperfusion areas, neovascularization, and microvascular abnormalities. The presence and increasing extent of predominantly peripheral lesions detected using UWF FA are associated with a higher risk of DR progression and proliferative DR. UWF imaging provides a comprehensive evaluation of DR severity, aiding in more accurate risk stratification and treatment decision-making. Overall, UWF imaging modalities have significantly advanced our understanding of peripheral retinal lesions in DR, facilitating early detection and targeted management for better visual outcomes.

The role of ultra-widefield imaging with navigated central and peripheral cross-sectional and three-dimensional swept source optical coherence tomography in ophthalmology: Clinical applications.

PURPOSE: To assess central and peripheral retinal and choroidal diseases using ultra-widefield (UWF) fundus imaging in combination with navigated central and peripheral cross-sectional and three-dimensional (3D) swept source optical coherence tomography (SS-OCT) scans. METHODS: Retrospective study involving 332 consecutive patients, with a nearly equal distribution of males and females. The mean age of patients was 52 years (range 18-92 years). Average refractive error was -3.80 D (range +7.75 to -20.75 D). RESULTS: The observations in this study demonstrate the efficacy of peripheral navigated SS-OCT in assessing various ocular conditions. The technology provides high-quality images of the peripheral vitreous, vitreoretinal interface, retina, and choroid, enabling visualization of vitreous floaters and opacities, retinal holes and tears, pigmented lesions, and peripheral retinal degenerations. 3D OCT scans enhance the visualization of these abnormalities and improve diagnostic and therapeutic decisions. CONCLUSION: Navigated central and peripheral cross-sectional and 3D SS-OCT scans offer significant complementary benefits in the assessment and management of retinal diseases. Their addition to UWF imaging provides a comprehensive view of central and peripheral ocular structures, aiding in early detection, precise anatomical measurements, and objective monitoring of disease progression. In addition, this technology serves as a valuable tool for patient education, a teaching tool for trainees, and documentation for medico-legal purposes.

Thyroid Hormone Signaling in Retinal Development and Function: Implications for Diabetic Retinopathy and Age-Related Macular Degeneration.

Thyroid Hormones (THs) play a central role in the development, cell growth, differentiation, and metabolic homeostasis of neurosensory systems, including the retina. The coordinated activity of various components of TH signaling, such as TH receptors (THRs) and the TH processing enzymes deiodinases 2 and 3 (DIO2, DIO3), is required for proper retinal maturation and function of the adult photoreceptors, Müller glial cells, and pigmented epithelial cells. Alterations of TH homeostasis, as observed both in frank or subclinical thyroid disorders, have been associated with sight-threatening diseases leading to irreversible vision loss i.e., diabetic retinopathy (DR), and age-related macular degeneration (AMD). Although observational studies do not allow causal inference, emerging data from preclinical models suggest a possible correlation between TH signaling imbalance and the development of retina disease. In this review, we analyze the most important features of TH signaling relevant to retinal development and function and its possible implication in DR and AMD etiology. A better understanding of TH pathways in these pathological settings might help identify novel targets and therapeutic strategies for the prevention and management of retinal disease.

Hybrid Explainable Artificial Intelligence Models for Targeted Metabolomics Analysis of Diabetic Retinopathy.

BACKGROUND: Diabetic retinopathy (DR) is a prevalent microvascular complication of diabetes mellitus, and early detection is crucial for effective management. Metabolomics profiling has emerged as a promising approach for identifying potential biomarkers associated with DR progression. This study aimed to develop a hybrid explainable artificial intelligence (XAI) model for targeted metabolomics analysis of patients with DR, utilizing a focused approach to identify specific metabolites exhibiting varying concentrations among individuals without DR (NDR), those with non-proliferative DR (NPDR), and individuals with proliferative DR (PDR) who have type 2 diabetes mellitus (T2DM). METHODS: A total of 317 T2DM patients, including 143 NDR, 123 NPDR, and 51 PDR cases, were included in the study. Serum samples underwent targeted metabolomics analysis using liquid chromatography and mass spectrometry. Several machine learning models, including Support Vector Machines (SVC), Random Forest (RF), Decision Tree (DT), Logistic Regression (LR), and Multilayer Perceptrons (MLP), were implemented as solo models and in a two-stage ensemble hybrid approach. The models were trained and validated using 10-fold cross-validation. SHapley Additive exPlanations (SHAP) were employed to interpret the contributions of each feature to the model predictions. Statistical analyses were conducted using the Shapiro-Wilk test for normality, the Kruskal-Wallis H test for group differences, and the Mann-Whitney U test with Bonferroni correction for post-hoc comparisons. RESULTS: The hybrid SVC + MLP model achieved the highest performance, with an accuracy of 89.58%, a precision of 87.18%, an F1-score of 88.20%, and an F-beta score of 87.55%. SHAP analysis revealed that glucose, glycine, and age were consistently important features across all DR classes, while creatinine and various phosphatidylcholines exhibited higher importance in the PDR class, suggesting their potential as biomarkers for severe DR. CONCLUSION: The hybrid XAI models, particularly the SVC + MLP ensemble, demonstrated superior performance in predicting DR progression compared to solo models. The application of SHAP facilitates the interpretation of feature importance, providing valuable insights into the metabolic and physiological markers associated with different stages of DR. These findings highlight the potential of hybrid XAI models combined with explainable techniques for early detection, targeted interventions, and personalized treatment strategies in DR management.

Determinants of progression of diabetic retinopathy in pregnancy.

AIMS: To assess the rate of diabetic retinopathy (DR) progression in an Australian cohort and to identify the determinants of DR progression in pregnancy. METHODS: A total of 367 pregnancies of women with Type 1 or 2 diabetes mellitus attending King Edward Memorial Hospital, Western Australia, between June 2020 and July 2023 were included. These women were screened for the presence and severity of DR in the first trimester and/or at 28-32weeks gestation via retinal imaging with a DRS camera. RESULTS: DR was seen in 121 (33 %) pregnancies at baseline and DR progression was seen in 62 (17 %) pregnancies. Only 11 (4 %) women with no baseline DR developed DR and none of these progressed to more than moderate non-proliferative DR. A total of 51 (42 %) women with baseline DR had DR progression. The presence of baseline DR was the only significant predictor for DR progression on multivariate analysis (OR 9.88 (4.43-22.07), p < 0.001). CONCLUSIONS: Women without DR at baseline are unlikely to progress to more severe forms of DR and usually do not require treatment. The presence of DR at baseline screening during pregnancy is a strong predictor of DR progression during pregnancy.

Screening for diabetic retinopathy with artificial intelligence: a real world evaluation.

AIM: Periodic screening for diabetic retinopathy (DR) is effective for preventing blindness. Artificial intelligence (AI) systems could be useful for increasing the screening of DR in diabetic patients. The aim of this study was to compare the performance of the DAIRET system in detecting DR to that of ophthalmologists in a real-world setting. METHODS: Fundus photography was performed with a nonmydriatic camera in 958 consecutive patients older than 18 years who were affected by diabetes and who were enrolled in the DR screening in the Diabetes and Endocrinology Unit and in the Eye Unit of ULSS8 Berica (Italy) between June 2022 and June 2023. All retinal images were evaluated by DAIRET, which is a machine learning algorithm based on AI. In addition, all the images obtained were analysed by an ophthalmologist who graded the images. The results obtained by DAIRET were compared with those obtained by the ophthalmologist. RESULTS: We included 958 patients, but only 867 (90.5%) patients had retinal images sufficient for evaluation by a human grader. The sensitivity for detecting cases of moderate DR and above was 1 (100%), and the sensitivity for detecting cases of mild DR was 0.84 ± 0.03. The specificity of detecting the absence of DR was lower (0.59 ± 0.04) because of the high number of false-positives. CONCLUSION: DAIRET showed an optimal sensitivity in detecting all cases of referable DR (moderate DR or above) compared with that of a human grader. On the other hand, the specificity of DAIRET was low because of the high number of false-positives, which limits its cost-effectiveness.

The Role of Widefield Optical Coherence Tomography Angiography in Assessing the Severity of Diabetic Retinopathy.

INTRODUCTION: Physicians need an accurate understanding of diabetic retinopathy (DR) severity to optimally manage patients. The aim of this prospective study is to correlate the severity of macular and peripheral retinal vascular abnormalities seen on widefield (WF) optical coherence tomography angiography (OCTA) with DR grading based on WF fundus photography. METHODS: The study included 150 eyes from 82 patients with treatment-naïve DR. All patients were imaged with WF fundus photography and swept-source WF OCTA. Quantitative and qualitative analyses of the foveal avascular zone (FAZ) size and shape, and measurement of capillary nonperfusion (CNP) areas, were performed from the OCTA images. The mixed-effects model was used to compare the DR grading from WF photography with the vascular changes seen on WF-OCTA, and Bonferroni correction was applied to the gradings. RESULTS: The mean [± standard deviation (SD)] age of patients was 55.5 (± 9.4) years. The WF-OCTA showed that an increasing size of the FAZ (from 0.442 (± 0.059) µm to 0.933 (± 0.086) µm) correlated with increasing severity of the DR (as determined with WF photography). The deep capillary plexus, FAZ size, and CNP areas in eyes with proliferative diabetic retinopathy (PDR) differed from those with mild nonproliferative diabetic retinopathy (NPDR) (p < 0.001). Most eyes with severe nonproliferative DR were found to have CNP in four quadrants [superficial capillary plexus (SCP) 60%, deep capillary plexus (DCP) 50%]. The WF-OCTA detected subtle neovascularization of the disc (NVD) in 7 eyes (10%) and neovascularization elsewhere (NVE) in 13 eyes (18%) that had been diagnosed with only moderate NPDR on WF photography. CONCLUSIONS: FAZ and CNP areas as measured by WF-OCTA correlate with DR severity. WF-OCTA can also detect subtle NVE and NVD that cannot be seen with fundus photography.