OBJECTIVE: To establish and validate a novel diabetic retinopathy (DR) risk-prediction model using a whole-exome sequencing (WES)-based machine learning (ML) method. METHODS: WES was performed to identify potential single nucleotide polymorphism (SNP) or mutation sites in a DR pedigree comprising 10 members. A prediction model was established and validated in a cohort of 420 type 2 diabetic patients based on both genetic and demographic features. The contribution of each feature was assessed using Shapley Additive explanation analysis. The efficacies of the models with and without SNP were compared. RESULTS: WES revealed that seven SNPs/mutations ( rs116911833 in TRIM7, 1997T>C in LRBA, 1643T>C in PRMT10, rs117858678 in C9orf152, rs201922794 in CLDN25, rs146694895 in SH3GLB2, and rs201407189 in FANCC) were associated with DR. Notably, the model including rs146694895 and rs201407189 achieved better performance in predicting DR (accuracy: 80.2%; sensitivity: 83.3%; specificity: 76.7%; area under the receiver operating characteristic curve [AUC]: 80.0%) than the model without these SNPs (accuracy: 79.4%; sensitivity: 80.3%; specificity: 78.3%; AUC: 79.3%). CONCLUSION Novel SNP sites associated with DR were identified in the DR pedigree. Inclusion of rs146694895 and rs201407189 significantly enhanced the performance of the ML-based DR prediction model.
Diabetic Retinopathy/diagnosis* ; Humans ; Machine Learning ; Male ; Female ; Polymorphism, Single Nucleotide ; Middle Aged ; Exome Sequencing ; Aged ; Adult ; Pedigree ; Diabetes Mellitus, Type 2/complications* ; Genetic Predisposition to Disease ; Mutation

Diabetic retinopathy (DR), one of the chronic complications of diabetes, is a serious and irreversible blinding disease. It is difficult to detect in the early stage, to control in the progressive stage, to operate in the advanced stage of DR. Recently, the "14th Five-year plan" for National Eye Health proposed to "improve the management mode of chronic eye disease, and build a chronic disease management system". The project team used artificial intelligence technology based on cloud platform, joint outpatient service, virtual ward to explore the comprehensive management of DR from the aspects of early screening, multidisciplinary collaborative diagnosis and treatment, and refined blood glucose management during perioperative period. In the future, it is urgent to integrate DR chronic disease management with other systemic chronic diseases to reduce the blindness caused by DR.
Humans ; Diabetic Retinopathy/diagnosis* ; Artificial Intelligence ; Mass Screening ; Blindness/prevention & control* ; Diabetes Mellitus

Diabetic retinopathy (DR) is an important cause of blindness globally, and its prevalence is increasing. Early detection and intervention can help change the outcomes of the disease. The rapid development of artificial intelligence (AI) in recent years has led to new possibilities for the screening and diagnosis of DR. An AI-based diagnostic system for the detection of DR has significant advantages, such as high efficiency, high accuracy, and lower demand for human resources. At the same time, there are shortcomings, such as the lack of standards for development and evaluation and the limited scope of application. This article demonstrates the current applications of AI in the field of DR, existing problems, and possible future development directions.
Artificial Intelligence ; Blindness ; Diabetes Mellitus ; Diabetic Retinopathy/diagnosis* ; Humans ; Mass Screening

PURPOSE: To investigate the risk factors of diabetic nephropathy in patients with diabetic retinopathy requiring panretinal photocoagulation (PRP) and the visual prognosis. METHODS: A retrospective review of electronic medical records was conducted at Seoul St. Mary's Hospital, comprising 103 patients with type 2 diabetes mellitus and diabetic retinopathy who underwent PRP from 1996 to 2005. Patients with type 1 diabetes mellitus, non-diabetic renal disease, non-diabetic retinal disease, visually significant ocular disease, high-risk proliferative diabetic retinopathy, and advanced diabetic retinopathy were excluded. The patients were divided into three groups: no nephropathy (group 1, n = 45), microalbuminuria (group 2, n = 16), and advanced nephropathy (group 3, n = 42). Duration of diagnosis of retinopathy and nephropathy, glycosylated hemoglobin, visual acuity, complications, and treatment history were investigated. RESULTS: The mean glycosylated hemoglobin of group 3 (8.4 ± 1.2) was higher than that of group 1 (7.7 ± 1.0) or group 2 (7.7 ± 1.0) (p = 0.04). Mean interval from PRP to diagnosis of nephropathy was 8.8 ± 6.0 years in group 2 and 8.7 ± 4.9 years in group 3. The significant decrease in visual acuity in group 3 (28 eyes, 35.9%) was significantly higher than that in group 1 (15 eyes, 18.1%, p = 0.01) or group 2 (6 eyes, 20.7%, p = 0.03). Only vitreous hemorrhage showed a significantly higher incidence in groups 2 and 3 than in group 1 (p = 0.02). Multivariate regression analysis revealed that female sex and lower glycosylated hemoglobin were significantly associated with a protective effect on development of nephropathy. CONCLUSIONS: In the clinical setting, many patients with PRP-requiring diabetic retinopathy develop nephropathy an average of 8 to 9 years after PRP. Male sex and higher glycosylated hemoglobin could be risk factors of nephropathy.
Diabetes Mellitus, Type 1 ; Diabetes Mellitus, Type 2 ; Diabetic Nephropathies ; Diabetic Retinopathy ; Diagnosis ; Electronic Health Records ; Female ; Hemoglobin A, Glycosylated ; Humans ; Incidence ; Light Coagulation ; Male ; Prognosis ; Retinal Diseases ; Retrospective Studies ; Risk Factors ; Seoul ; Visual Acuity ; Vitreous Hemorrhage

PURPOSE: To clarify the relationship between various factors (sociodemographic factors, health behavioral risk factors and health status, and diabetic factors) related to diabetic retinopathy and to suggest improvements regarding the associated medical examination. METHODS: The subjects were 1,444 diabetic patients diagnosed in the 5th Korean National Health and Nutrition Examination Survey (KNHANES), aged 19 years or older, who underwent non-mydriatic fundus photography. The criteria for diagnosing diabetes were a fasting glucose level ≥ 126 mg/dL and a previous diagnosis of diabetes or currently undergoing treatment. The diagnosis of diabetic retinopathy followed the modified Airlie House classification. Univariate and multivariate analyses of diabetic retinopathy were performed. RESULTS: Among the 1,444 patients who were diagnosed with diabetes, 277 had diabetic retinopathy; the prevalence rate was 19.18%. The higher the body mass index, the lower the risk of diabetic retinopathy by 0.924 times (p = 0.001; 95% confidence interval [CI], 0.883–0.966). The longer the duration of diabetes, the greater the risk of diabetic retinopathy; the prevalence period group of more than 11 years had a 26.025-fold higher risk than the newly diagnosed group (p < 0.001; 95% CI, 10.840–62.482). The risk of diabetic retinopathy increased with the hemoglobin A1c (HbA1c) level; the risk was 5.973-fold higher in the group with HbA1c above 11.0% (p < 0.001; 95% CI, 2.984–11.956) compared with the group with HbA1c < 6.0%. The risk of diabetic retinopathy was 2.050-fold greater with insulin injections (p = 0.003; 95% CI, 1.284–3.275). CONCLUSIONS: The risk of diabetic retinopathy was higher in patients with a longer duration of diabetes, a high HbA1c level, and in those using insulin. These findings highlight the need for early education and ophthalmologic examinations for at-risk patients.
Body Mass Index ; Classification ; Diabetic Retinopathy ; Diagnosis ; Education ; Fasting ; Glucose ; Health Behavior ; Humans ; Insulin ; Multivariate Analysis ; Nutrition Surveys ; Photography ; Prevalence ; Risk Factors

Artificial Intelligence ; Deep Learning ; Diabetic Retinopathy ; diagnosis ; Humans ; Image Processing, Computer-Assisted ; Machine Learning ; Mass Screening ; Melanoma ; diagnosis ; Referral and Consultation ; Skin Neoplasms ; diagnosis ; Triage

OBJECTIVE: Biomarkers of attention deficit hyperactivity disorder (ADHD) are crucial for early diagnosis and intervention, in which the identification of biomarkers in other areas of the body that represent the immature brain of children with ADHD is necessary. The present study aimed to find biomarkers of ADHD in the retina and assessed the relationship between macular thickness of the retina and cortical thickness of the brain in children with ADHD. METHODS: Twelve children with ADHD and 13 control children were recruited for the study. To find ocular markers of ADHD, we investigated the correlation between clinical symptoms of ADHD assessed with the Korean ADHD Rating Scale (K-ARS), cortical thickness of the brain, and macular thickness measured with the mean thickness from the Early Treatment Diabetic Retinopathy Study (ETDRS). RESULTS: Children with ADHD showed increased macular thicknesses quantified as an ETDRS ring in both eyes, compared to control subjects. Moreover, the right inner ETDRS ring had a positive correlation with K-ARS scores. The ADHD group had an increased ratio of thickness of the right frontal lobe to that of the parietal cortex, compared with the control group. There were positive correlations between the means of the inner ETDRS ring (right) and the left paracentral/right isthmus cingulate thicknesses in the control group. However, there were negative correlations between the means of the inner ETDRS ring (right) and the left frontal pole/right pars triangularis thicknesses in the ADHD group. The results of both groups were at the uncorrected level. CONCLUSION: The different patterns of macular thickness might represent the immature cortical thickness of the brain in children with ADHD.
Attention Deficit Disorder with Hyperactivity ; Biomarkers ; Brain ; Broca Area ; Child ; Diabetic Retinopathy ; Diagnosis ; Early Diagnosis ; Frontal Lobe ; Humans ; Magnetic Resonance Imaging ; Parietal Lobe ; Pilot Projects ; Retina

PURPOSE: To investigate longitudinal changes in the thicknesses of the peripapillary retinal nerve fiber layer (pRNFL) and the macular ganglion cell-inner plexiform layer (mGCIPL) in patients with diabetic retinopathy 3 years after panretinal photocoagulation (PRP). METHODS: We retrospectively reviewed the medical records of 60 eyes of 35 patients who were diagnosed with diabetic retinopathy and treated with PRP. The pRNFL and mGCIPL thicknesses were measured by optical coherence tomography at baseline, and then at 1, 3, 6, 9, 12, 24, and 36 months after PRP. RESULTS: The pRNFL and mGCIPL thicknesses (average and all sections) at 1 year after PRP increased significantly from baseline (p < 0.05, respectively). The average pRNFL and mGCIPL thicknesses showed a tendency to decrease continuously from 2 years after PRP (p < 0.05, respectively). There was no statistically significant difference in the average thicknesses of the pRNFL and the mGCIPL between pre-PRP (92.27 ± 7.76 µm, and 85.00 ± 4.80 µm, respectively) and 3 years after PRP (93.93 ± 7.49 µm, and 81.87 ± 14.00 µm, respectively) (p = 0.121, and p = 0.622, respectively). CONCLUSIONS: Although the pRNFL and the mGCIPL thicknesses increased at 1 year after PRP, there was no statistical difference in the average thicknesses of the pRNFL and the mGCIPL between pre-PRP and 3 years after PRP. These results should be considered with respect to the diagnosis and progression of glaucoma in patients with diabetic retinopathy who undergo PRP.
Diabetic Retinopathy ; Diagnosis ; Ganglion Cysts ; Glaucoma ; Humans ; Light Coagulation ; Medical Records ; Nerve Fibers ; Retinaldehyde ; Retrospective Studies ; Tomography, Optical Coherence

The high-pace rise in advanced computing and imaging systems has given rise to a new research dimension called computer-aided diagnosis (CAD) system for various biomedical purposes. CAD-based diabetic retinopathy (DR) can be of paramount significance to enable early disease detection and diagnosis decision. Considering the robustness of deep neural networks (DNNs) to solve highly intricate classification problems, in this paper, AlexNet DNN, which functions on the basis of convolutional neural network (CNN), has been applied to enable an optimal DR CAD solution. The DR model applies a multilevel optimization measure that incorporates pre-processing, adaptive-learning-based Gaussian mixture model (GMM)-based concept region segmentation, connected component-analysis-based region of interest (ROI) localization, AlexNet DNN-based highly dimensional feature extraction, principle component analysis (PCA)- and linear discriminant analysis (LDA)-based feature selection, and support-vector-machine-based classification to ensure optimal five-class DR classification. The simulation results with standard KAGGLE fundus datasets reveal that the proposed AlexNet DNN-based DR exhibits a better performance with LDA feature selection, where it exhibits a DR classification accuracy of 97.93% with FC7 features, whereas with PCA, it shows 95.26% accuracy. Comparative analysis with spatial invariant feature transform (SIFT) technique (accuracy—94.40%) based DR feature extraction also confirms that AlexNet DNN-based DR outperforms SIFT-based DR.
Classification ; Dataset ; Diabetic Retinopathy ; Diagnosis ; Passive Cutaneous Anaphylaxis

BACKGROUND: To analyze the results of court rulings about medical litigations related to cataract surgery in Korea. METHODS: We collected 50 anonymized judgements regarding cataract surgery between 2000 and 2016 and analyzed the reasons for the medical litigations, the court rulings, the reasons for compensation, and the amount claimed and finally awarded. RESULTS: Forty-eight litigations (96%) resulted from errors in treatment, and the reasons were as follows: endophthalmitis, dissatisfaction of visual outcome or ocular discomfort, bullous keratopathy or corneal opacity, retinal detachment, glaucoma or vitreous hemorrhage due to the progression of an underlying diabetic retinopathy, and others in order. Two litigations (4%) arose due to errors in diagnosis. Among the 50 final cases, 21 litigations (42%) were decided in favor of the plaintiff, and 29 litigations (58%) were decided against the plaintiff and dismissed. Ten cases awarded damages to the plaintiffs because of a violation of duty of care, and 11 cases awarded damages due to a violation of informed consent. When comparing cases with errors in diagnosis to cases with errors in treatment, there was no significant difference in the relative risk of a defendant's verdict (P = 0.503). The total amount of awarded damages was KRW 439,124,496 (USD 399,204), and the average amount was KRW 20,910,690 (USD 19,010). CONCLUSION: Nearly half of the cases were decided in favor of the plaintiff due to the violation of informed consent. This study's results will be helpful in understanding the results of medical litigations regarding cataract surgery and reducing future lawsuits.
Anonyms and Pseudonyms ; Awards and Prizes ; Cataract* ; Compensation and Redress ; Corneal Opacity ; Diabetic Retinopathy ; Diagnosis ; Endophthalmitis ; Glaucoma ; Informed Consent ; Jurisprudence* ; Korea* ; Ophthalmology ; Retinal Detachment ; Vitreous Hemorrhage