Explore the vast range of titles available.

To evaluate the performance of machine learning (ML) algorithms and to compare them with logistic regression for the prediction of risk of cardiovascular diseases (CVDs), chronic kidney disease (CKD), diabetes (DM), and hypertension (HTN) and in a prospective cohort study using simple clinical predictors. We conducted analyses in a population-based cohort study in Asian adults (n = 6,762). Five different ML models were considered—single-hidden-layer neural network, support vector machine, random forest, gradient boosting machine, and k-nearest neighbor—and were compared with standard logistic regression. The incidences at 6 years of CVD, CKD, DM, and HTN cases were 4.0%, 7.0%, 9.2%, and 34.6%, respectively. Logistic regression reached the highest area under the receiver operating characteristic curve for CKD (0.905 [0.88, 0.93]) and DM (0.768 [0.73, 0.81]) predictions. For CVD and HTN, the best models were neural network (0.753 [0.70, 0.81]) and support vector machine (0.780 [0.747, 0.812]), respectively. However, the differences with logistic regression were small (less than 1%) and nonsignificant. Logistic regression, gradient boosting machine, and neural network were systematically ranked among the best models. Logistic regression yields as good performance as ML models to predict the risk of major chronic diseases with low incidence and simple clinical predictors. •Low-dimensional settings include low number of events and predictors.•In such settings, logistic regression yields as good performance as ML models.•ML techniques may not be warranted in such cases.

BackgroundThe impacts of social restrictions for COVID-19 on children’s vision and lifestyle remain unknown.AimsTo investigate myopia incidence, spherical equivalent refraction (SER) and lifestyle changes among schoolchildren during the COVID-19 pandemic.MethodsTwo separate longitudinal cohorts of children aged 6–8 years in Hong Kong were included. The COVID-19 cohort was recruited at the beginning of the COVID-19 outbreak, whereas the pre-COVID-19 cohort was recruited before the COVID-19 pandemic. All children received ocular examinations, and answered a standardised questionnaire relating to their lifestyle, including time spent on outdoor activities and near work, both at baseline and at follow-up visits.ResultsA total of 1793 subjects were recruited, of whom 709 children comprised the COVID-19 cohort with 7.89±2.30 months of follow-up, and 1084 children comprised the pre-COVID-19 cohort with 37.54±3.12 months of follow-up. The overall incidence was 19.44% in the COVID-19 cohort, and 36.57% in pre-COVID-19 cohort. During the COVID-19 pandemic, the change in SER and axial length was –0.50±0.51 D and 0.29±0.35 mm, respectively; the time spent on outdoor activities decreased from 1.27±1.12 to 0.41±0.90 hours/day (p<0.001), while screen time increased from 2.45±2.32 to 6.89±4.42 hours/day (p<0.001).ConclusionsWe showed a potential increase in myopia incidence, significant decrease in outdoor time and increase in screen time among schoolchildren in Hong Kong during the COVID-19 pandemic. Our results serve to warn eye care professionals, and also policy makers, educators and parents, that collective efforts are needed to prevent childhood myopia—a potential public health crisis as a result of COVID-19.

The vascularity of the choroid has been implicated in the pathogenesis of various eye diseases. To date, no established quantifiable parameters to estimate vascular status of the choroid exists. Choroidal vascularity index (CVI) may potentially be used to assess vascular status of the choroid. We aimed to establish normative database for CVI and identify factors associated with CVI in healthy eyes. In this population-based study on 345 healthy eyes, choroidal enhanced depth imaging optical coherence tomography scans were segmented by modified image binarization technique. Total subfoveal choroidal area (TCA) was segmented into luminal (LA) and stromal (SA) area. CVI was calculated as the proportion of LA to TCA. Linear regression was used to identify ocular and systemic factors associated with CVI and subfoveal choroidal thickness (SFCT). Subfoveal CVI ranged from 60.07 to 71.27% with a mean value of 65.61 ± 2.33%. CVI was less variable than SFCT (coefficient of variation for CVI was 3.55 vs 40.30 for SFCT). Higher CVI was associated with thicker SFCT, but not associated with most physiological variables. CVI was elucidated as a significant determinant of SFCT. While SFCT was affected by many factors, CVI remained unaffected suggesting CVI to be a more robust marker of choroidal diseases.

Diabetic kidney disease (DKD) and diabetic retinopathy (DR) are major diabetic microvascular complications, contributing significantly to morbidity, disability, and mortality worldwide. The kidney and the eye, having similar microvascular structures and physiological and pathogenic features, may experience similar metabolic changes in diabetes. This study aimed to use machine learning (ML) methods integrated with metabolic data to identify biomarkers associated with DKD and DR in a multiethnic Asian population with diabetes, as well as to improve the performance of DKD and DR detection models beyond traditional risk factors. We used ML algorithms (logistic regression [LR] with Least Absolute Shrinkage and Selection Operator and gradient-boosting decision tree) to analyze 2772 adults with diabetes from the Singapore Epidemiology of Eye Diseases study, a population-based cross-sectional study conducted in Singapore (2004-2011). From 220 circulating metabolites and 19 risk factors, we selected the most important variables associated with DKD (defined as an estimated glomerular filtration rate <60 mL/min/1.73 m ) and DR (defined as an Early Treatment Diabetic Retinopathy Study severity level ≥20). DKD and DR detection models were developed based on the variable selection results and externally validated on a sample of 5843 participants with diabetes from the UK biobank (2007-2010). Machine-learned model performance (area under the receiver operating characteristic curve [AUC] with 95% CI, sensitivity, and specificity) was compared to that of traditional LR adjusted for age, sex, diabetes duration, hemoglobin A , systolic blood pressure, and BMI. Singapore Epidemiology of Eye Diseases participants had a median age of 61.7 (IQR 53.5-69.4) years, with 49.1% (1361/2772) being women, 20.2% (555/2753) having DKD, and 25.4% (685/2693) having DR. UK biobank participants had a median age of 61.0 (IQR 55.0-65.0) years, with 35.8% (2090/5843) being women, 6.7% (374/5570) having DKD, and 6.1% (355/5843) having DR. The ML algorithms identified diabetes duration, insulin usage, age, and tyrosine as the most important factors of both DKD and DR. DKD was additionally associated with cardiovascular disease history, antihypertensive medication use, and 3 metabolites (lactate, citrate, and cholesterol esters to total lipids ratio in intermediate-density lipoprotein), while DR was additionally associated with hemoglobin A , blood glucose, pulse pressure, and alanine. Machine-learned models for DKD and DR detection outperformed traditional LR models in both internal (AUC 0.838 vs 0.743 for DKD and 0.790 vs 0.764 for DR) and external validation (AUC 0.791 vs 0.691 for DKD and 0.778 vs 0.760 for DR). This study highlighted diabetes duration, insulin usage, age, and circulating tyrosine as important factors in detecting DKD and DR. The integration of ML with biomedical big data enables biomarker discovery and improves disease detection beyond traditional risk factors.

According to a 2013–2014 survey of nearly 46,000 U.S. adults and youths, 28% of adults were current users of tobacco and 9% of youths had used tobacco in the previous 30 days. Approximately 40% of users used multiple products, with cigarettes plus e-cigarettes the most common combination. Smoking is responsible for more U.S. deaths annually than the acquired immunodeficiency syndrome, use of alcohol and illegal drugs, motor vehicle accidents, murders, and suicides combined. 1 With recent data suggesting higher smoking-attributable mortality than previously estimated, 2 the medical community is urged to make tobacco control a high priority. 3 The prevalence of current use of cigarettes has declined during the past 50 years, from 42% of adults in 1965 to less than 20% in 2014, 4 , 5 but disparities in cigarette smoking across demographic subgroups (particularly according to race or ethnic group, educational attainment, and socioeconomic status) have widened during the past . . .

To examine prevalence and risk factors of epiretinal membrane (ERM) in a large, contemporary, multiethnic Asian population. Combined analysis of three population-based studies of eye diseases, with a total of 9799 Chinese, Malays and Indians residing in the general communities of Singapore. A comprehensive ophthalmic examination, interviews and laboratory blood tests were performed to assess potential risk factors. Digital retinal photographs were used to assess ERM according to a standardised protocol. ERM was classified into cellophane macular reflex (CMR) and/or preretinal macular fibrosis (PMF), and also as primary or secondary (in eyes with other retinal pathology or a history of cataract surgery). The age-standardised and ethnicity-standardised prevalence was 12.1% for any ERM, 6.8% for CMR, 6.7% for PMF and 2.8% for bilateral ERM. ERM prevalence was higher in Chinese (13.0%) compared with Malays (7.9%) or Indians (8.7%). In multivariate analysis, significant factors associated with primary ERM were older age (OR 1.08 per year increase; p<0.01), Chinese ethnicity (OR 1.60 vs Indians; p<0.01; OR 1.39 vs Malays; p<0.01), smoking (OR 0.70; p=0.01), longer axial length (OR 1.07 per mm increase; p=0.03) and cataract (OR 0.64; p<0.01). Significant factors independently associated with secondary ERM were older age (OR 1.05; p<0.01), cataract surgery (OR 10.6; p<0.01) and diabetic retinopathy (OR 2.48; p<0.01). ERM is common in Asians, particularly among Chinese. Older age is the most consistent risk factor for any ERM, and previous cataract surgery and diabetic retinopathy are the strongest risk factors for secondary ERM.

COVID-19 has led to massive disruptions in societal, economic and healthcare systems globally. While COVID-19 has sparked a surge and expansion of new digital business models in different industries, healthcare has been slower to adapt to digital solutions. The majority of ophthalmology clinical practices are still operating through a traditional model of ‘brick-and-mortar’ facilities and ‘face-to-face’ patient–physician interaction. In the current climate of COVID-19, there is a need to fuel implementation of digital health models for ophthalmology. In this article, we highlight the current limitations in traditional clinical models as we confront COVID-19, review the current lack of digital initiatives in ophthalmology sphere despite the presence of COVID-19, propose new digital models of care for ophthalmology and discuss potential barriers that need to be considered for sustainable transformation to take place.

Background/aimsTo evaluate the performances of deep learning (DL) algorithms for detection of presence and extent pterygium, based on colour anterior segment photographs (ASPs) taken from slit-lamp and hand-held cameras.MethodsReferable pterygium was defined as having extension towards the cornea from the limbus of >2.50 mm or base width at the limbus of >5.00 mm. 2503 images from the Singapore Epidemiology of Eye Diseases (SEED) study were used as the development set. Algorithms were validated on an internal set from the SEED cohort (629 images (55.3% pterygium, 8.4% referable pterygium)), and tested on two external clinic-based sets (set 1 with 2610 images (2.8% pterygium, 0.7% referable pterygium, from slit-lamp ASP); and set 2 with 3701 images, 2.5% pterygium, 0.9% referable pterygium, from hand-held ASP).ResultsThe algorithm’s area under the receiver operating characteristic curve (AUROC) for detection of any pterygium was 99.5%(sensitivity=98.6%; specificity=99.0%) in internal test set, 99.1% (sensitivity=95.9%, specificity=98.5%) in external test set 1 and 99.7% (sensitivity=100.0%; specificity=88.3%) in external test set 2. For referable pterygium, the algorithm’s AUROC was 98.5% (sensitivity=94.0%; specificity=95.3%) in internal test set, 99.7% (sensitivity=87.2%; specificity=99.4%) in external set 1 and 99.0% (sensitivity=94.3%; specificity=98.0%) in external set 2.ConclusionDL algorithms based on ASPs can detect presence of and referable-level pterygium with optimal sensitivity and specificity. These algorithms, particularly if used with a handheld camera, may potentially be used as a simple screening tool for detection of referable pterygium. Further validation in community setting is warranted.Synopsis/precisDL algorithms based on ASPs can detect presence of and referable-level pterygium optimally, and may be used as a simple screening tool for the detection of referable pterygium in community screenings.

A major complication of hypertension is microvascular damage and capillary rarefaction is a known complication of hypertensive end-organ damage which confers a higher risk of systemic disease such as stroke and cardiovascular events. Our aim was to study the effect of hypertension on the retinal microvasculature using non-invasive optical coherence tomography angiography (OCTA). We performed a case-control study of 94 eyes of 94 participants with systemic hypertension and 46 normal control eyes from the Singapore Chinese Eye Study using a standardized protocol to collect data on past medical history of hypertension, including the number and type of hypertensive medications and assessed mean arterial pressure. Retinal vascular parameters were measured in all eyes using OCTA. In the multivariate analysis adjusting for confounders, compared to controls, eyes of hypertensive patients showed a decrease in the macular vessel density at the level of the superficial [OR 0.02; 95% CI, 0 to 0.64; P 0.027] and deep venous plexuses [OR 0.03; 95% CI, 0 to 0.41; P 0.009] and an increase in the deep foveal avascular zone. This shows that hypertension is associated with reduced retinal vessel density and an increased foveal avascular zone, especially in the deep venous plexus, as seen on OCTA and there is a potential role in using OCTA as a clinical tool to monitor hypertensive damage and identifying at risk patients

Deep learning (DL), a subset of artificial intelligence (AI) based on deep neural networks, has made significant breakthroughs in medical imaging, particularly for image classification and pattern recognition. In ophthalmology, applying DL for glaucoma assessment with optical coherence tomography (OCT), including OCT traditional reports, two-dimensional (2D) B-scans, and three-dimensional (3D) volumetric scans, has increasingly raised research interests. Studies have demonstrated that using DL for interpreting OCT is efficient, accurate, and with good performance for discriminating glaucomatous eyes from normal eyes, suggesting that incorporation of DL technology in OCT for glaucoma assessment could potentially address some gaps in the current practice and clinical workflow. However, further research is crucial in tackling some existing challenges, such as annotation standardization (i.e., setting a standard for ground truth labelling among different studies), development of DL-powered IT infrastructure for real-world implementation, prospective validation in unseen datasets for further evaluation of generalizability, cost-effectiveness analysis after integration of DL, the AI “black box” explanation problem. This review summarizes recent studies on the application of DL on OCT for glaucoma assessment, identifies the potential clinical impact arising from the development and deployment of the DL models, and discusses future research directions.