Myopia is a leading cause of visual impairment and has raised significant international concern in recent decades with rapidly increasing prevalence and incidence worldwide. Accurate prediction of ...future myopia risk could help identify high-risk children for early targeted intervention to delay myopia onset or slow myopia progression. Researchers have built and assessed various myopia prediction models based on different datasets, including baseline refraction or biometric data, lifestyle data, genetic data, and data integration. Here, we summarize all related work published in the past 30 years and provide a comprehensive review of myopia prediction methods, datasets, and performance, which could serve as a useful reference and valuable guideline for future research.
To assess the performance of a deep learning algorithm for detecting referable glaucomatous optic neuropathy (GON) based on color fundus photographs.
A deep learning system for the classification of ...GON was developed for automated classification of GON on color fundus photographs.
We retrospectively included 48 116 fundus photographs for the development and validation of a deep learning algorithm.
This study recruited 21 trained ophthalmologists to classify the photographs. Referable GON was defined as vertical cup-to-disc ratio of 0.7 or more and other typical changes of GON. The reference standard was made until 3 graders achieved agreement. A separate validation dataset of 8000 fully gradable fundus photographs was used to assess the performance of this algorithm.
The area under receiver operator characteristic curve (AUC) with sensitivity and specificity was applied to evaluate the efficacy of the deep learning algorithm detecting referable GON.
In the validation dataset, this deep learning system achieved an AUC of 0.986 with sensitivity of 95.6% and specificity of 92.0%. The most common reasons for false-negative grading (n = 87) were GON with coexisting eye conditions (n = 44 50.6%), including pathologic or high myopia (n = 37 42.6%), diabetic retinopathy (n = 4 4.6%), and age-related macular degeneration (n = 3 3.4%). The leading reason for false-positive results (n = 480) was having other eye conditions (n = 458 95.4%), mainly including physiologic cupping (n = 267 55.6%). Misclassification as false-positive results amidst a normal-appearing fundus occurred in only 22 eyes (4.6%).
A deep learning system can detect referable GON with high sensitivity and specificity. Coexistence of high or pathologic myopia is the most common cause resulting in false-negative results. Physiologic cupping and pathologic myopia were the most common reasons for false-positive results.
We provide a standardized set of terminology, definitions, and thresholds of myopia and its main ocular complications.
Critical review of current terminology and choice of myopia thresholds was done ...to ensure that the proposed standards are appropriate for clinical research purposes, relevant to the underlying biology of myopia, acceptable to researchers in the field, and useful for developing health policy.
We recommend that the many descriptive terms of myopia be consolidated into the following descriptive categories: myopia, secondary myopia, axial myopia, and refractive myopia. To provide a framework for research into myopia prevention, the condition of "pre-myopia" is defined. As a quantitative trait, we recommend that myopia be divided into myopia (i.e., all myopia), low myopia, and high myopia. The current consensus threshold value for myopia is a spherical equivalent refractive error ≤ -0.50 diopters (D), but this carries significant risks of classification bias. The current consensus threshold value for high myopia is a spherical equivalent refractive error ≤ -6.00 D. "Pathologic myopia" is proposed as the categorical term for the adverse, structural complications of myopia. A clinical classification is proposed to encompass the scope of such structural complications.
Standardized definitions and consistent choice of thresholds are essential elements of evidence-based medicine. It is hoped that these proposals, or derivations from them, will facilitate rigorous, evidence-based approaches to the study and management of myopia.
Purpose
To determine the association between retinal blood vessel flow and geometric parameters and the risk of diabetic retinopathy (DR) progression through a 2-year prospective cohort study.
...Methods
Patients with type 2 diabetes mellitus (T2DM) were recruited from a diabetic registry between November 2017 and March 2019. All participants underwent standardized examinations at the baseline and 2-year follow-up visit, and the presence and severity of DR were assessed based on standard seven-field color fundus photographs. They also underwent swept-source optical coherence tomography angiography (OCTA) imaging to obtain measurements of foveal avascular zone area, blood vessel density (VD), fractal dimension (FD), blood vessel tortuosity (BVT) in the superficial capillary plexus (SCP) and deep capillary plexus (DCP).
Results
A total of 233 eyes of 125 patients were included, and 40 eyes (17.17%) experienced DR progression within 2 years. DR progression was significantly associated with lower baseline VD (odds ratio OR 2.323 per SD decrease; 95% confidence interval CI 1.456–3.708;
P
< 0.001), lower FD (OR, 2.484 per SD decrease; 95% CI 1.268–4.867;
P
= 0.008), and higher BVT (OR, 2.076 per SD increase; 95% CI 1.382–3.121;
P
< 0.001) of the DCP after adjusting for confounding factors. The addition of OCTA metrics improved the predictive ability of the original model for DR progression (area under the curve AUC from 0.725 to 0.805;
P
= 0.022).
Conclusions
OCTA-derived VD, FD and BVT in the DCP were independent predictors of DR progression and showed additive value when added to established risk models predicting DR progression.
Convolutional neural networks have recently been applied to ophthalmic diseases; however, the rationale for the outputs generated by these systems is inscrutable to clinicians. A visualization tool ...is needed that would enable clinicians to understand important exposure variables in real time.
To systematically visualize the convolutional neural networks of 2 validated deep learning models for the detection of referable diabetic retinopathy (DR) and glaucomatous optic neuropathy (GON).
The GON and referable DR algorithms were previously developed and validated (holdout method) using 48 116 and 66 790 retinal photographs, respectively, derived from a third-party database (LabelMe) of deidentified photographs from various clinical settings in China. In the present cross-sectional study, a random sample of 100 true-positive photographs and all false-positive cases from each of the GON and DR validation data sets were selected. All data were collected from March to June 2017. The original color fundus images were processed using an adaptive kernel visualization technique. The images were preprocessed by applying a sliding window with a size of 28 × 28 pixels and a stride of 3 pixels to crop images into smaller subimages to produce a feature map. Threshold scales were adjusted to optimal levels for each model to generate heat maps highlighting localized landmarks on the input image. A single optometrist allocated each image to predefined categories based on the generated heat map.
Visualization regions of the fundus.
In the GON data set, 90 of 100 true-positive cases (90%; 95% CI, 82%-95%) and 15 of 22 false-positive cases (68%; 95% CI, 45%-86%) displayed heat map visualization within regions of the optic nerve head only. Lesions typically seen in cases of referable DR (exudate, hemorrhage, or vessel abnormality) were identified as the most important prognostic regions in 96 of 100 true-positive DR cases (96%; 95% CI, 90%-99%). In 39 of 46 false-positive DR cases (85%; 95% CI, 71%-94%), the heat map displayed visualization of nontraditional fundus regions with or without retinal venules.
These findings suggest that this visualization method can highlight traditional regions in disease diagnosis, substantiating the validity of the deep learning models investigated. This visualization technique may promote the clinical adoption of these models.
It is unclear whether breakfast consumption and breakfast composition are independently associated with changes in cognition over a long-term period in older adults.
We aimed to examine the ...associations between energy and macronutrient intakes at breakfast and cognitive declines.
We included 2935 participants aged 55–93 y at baseline from the China Health and Nutrition Survey in our analysis. Cognition was assessed in 1997, 2000, 2004, 2006, and 2015. Dietary intake was assessed using weighing methods in combination with 24-h food records.
Breakfast contributed to 25.9% of total energy intake of the day and percentages of breakfast energy intake from protein, fat, and carbohydrates were 12.8%, 11.5%, and 75.7%, respectively. During a median follow-up of 9 y, the β values for changes in global cognitive z-scores for Quintile 5 of protein and fat intakes at breakfast, with Quintile 1 as the reference, were 0.13 (95% CI: 0.01–0.25) and 0.17 (95% CI: 0.04–0.30), respectively. Substitution of 5% energy from carbohydrates with equivalent energy from protein (β, 0.06; 95% CI: 0.01–0.11) or fat (β, 0.05; 95% CI: 0.02–0.08) at breakfast was positively associated with the change in the global cognitive z-score. Energy intake at breakfast was not significantly associated with the global cognitive z-score. Similar results were found for the verbal memory z-score. The positive association of breakfast fat intake and the inverse association of breakfast carbohydrate intake with cognitive declines were stronger in urban residents.
Higher intakes of protein and fat and lower intake of carbohydrates at breakfast were associated with a lower rate of cognitive decline in older adults. Substitution of carbohydrates with protein or fat intake at breakfast may help to delay or prevent cognitive declines.
To develop a deep learning (DL) model that predicts age from fundus images (retinal age) and to investigate the association between retinal age gap (retinal age predicted by DL model minus ...chronological age) and mortality risk.
A total of 80 169 fundus images taken from 46 969 participants in the UK Biobank with reasonable quality were included in this study. Of these, 19 200 fundus images from 11 052 participants without prior medical history at the baseline examination were used to train and validate the DL model for age prediction using fivefold cross-validation. A total of 35 913 of the remaining 35 917 participants had available mortality data and were used to investigate the association between retinal age gap and mortality.
The DL model achieved a strong correlation of 0.81 (p<0·001) between retinal age and chronological age, and an overall mean absolute error of 3.55 years. Cox regression models showed that each 1 year increase in the retinal age gap was associated with a 2% increase in risk of all-cause mortality (hazard ratio (HR)=1.02, 95% CI 1.00 to 1.03, p=0.020) and a 3% increase in risk of cause-specific mortality attributable to non-cardiovascular and non-cancer disease (HR=1.03, 95% CI 1.00 to 1.05, p=0.041) after multivariable adjustments. No significant association was identified between retinal age gap and cardiovascular- or cancer-related mortality.
Our findings indicate that retinal age gap might be a potential biomarker of ageing that is closely related to risk of mortality, implying the potential of retinal image as a screening tool for risk stratification and delivery of tailored interventions.
ObjectivesTo explore associations between visual impairment (VI) and mortality in an adult population in urban China.MethodsThe Liwan Eye Study was a population-based prevalence survey conducted in ...Guangzhou, Southern China. The baseline examination was carried out in 2003. All baseline participants were invited for the 10-year follow-up visit. VI was defined as the visual acuity of 20/40 or worse in the better-seeing eye with habitual correction if worn. Correctable VI was defined as the VI correctable to 20/40 or better by subjective refraction, and non-correctable VI was defined as the VI correctable to worse than 20/40. Mortality rates were compared using the log-rank test and Cox proportional hazards regression models.ResultsOf the 1399 participants (mean age: 65.3 ± 9.93 years; 56.4% female) with available baseline visual acuity measurement, 320 participants (22.9%) had VI. After 10 years, 314 (22.4%) participants died. Visually impaired participants had a significantly increased 10-year mortality compared with those without VI (40.0% vs. 17.2%, P < 0.05). After adjusting for age, gender, income, educational attainment, BMI, history of diabetes and hypertension, both VI (HR, 1.55; 95% CI, 1.14–2.11) and non-correctable VI (HR, 2.72; 95% CI, 1.86–3.98) were significantly associated with poorer survival, while correctable VI (HR, 0.99; 95% CI, 0.66–1.49) was not an independent risk factor for 10-year mortality.ConclusionsOur findings that VI, particularly non-correctable VI, predicting poorer survival may imply the underlying mechanism behind VI-mortality association and reinforce the importance of preventing and treating disabling ocular diseases to prevent premature mortality in the elderly.
To examine annual changes in refraction and ocular components around the onset of myopia in Chinese children.
Longitudinal study.
Twins aged 7 to 15 years in 2006 from the Guangzhou Twin Registry.
...Participants underwent eye examinations annually from 2006 to 2010. Years were defined (+1 or -1) relative to the examination at which the onset of myopia was first identified.
Annual change in spherical equivalent refraction (SER) and ocular biometry.
Children who were not myopic at the first examination and myopic in at least 1 subsequent examination from 2006 to 2010 were included in the analysis. Annual change in SER increased slowly from 4 years before the first detection of myopia to 2 years before myopia onset (-0.25 to -0.4 diopter D). The rate of progression was the highest during the year of onset (-0.92 D). After the first detection of myopia, the rate of progression decreased to -0.71 D in the following year and kept decreasing. Annual change in axial length showed a similar, but inverse, shape to that of SER. Annual change in lens power did not change significantly around the onset of myopia.
Before the onset of myopia, axial elongation and progression accelerate. After a myopic refraction is established, axial elongation and progression decrease. We suggest that the increases before myopia may be due to increased intensity of study and decreased time outdoors. In contrast, the rapid slowing after the onset of myopia may represent an inhibitory effect of myopic defocus on eye growth.
Abstract
This study investigated the diagnostic performance, feasibility, and end-user experiences of an artificial intelligence (AI)-assisted diabetic retinopathy (DR) screening model in real-world ...Australian healthcare settings. The study consisted of two components: (1) DR screening of patients using an AI-assisted system and (2) in-depth interviews with health professionals involved in implementing screening. Participants with type 1 or type 2 diabetes mellitus attending two endocrinology outpatient and three Aboriginal Medical Services clinics between March 2018 and May 2019 were invited to a prospective observational study. A single 45-degree (macula centred), non-stereoscopic, colour retinal image was taken of each eye from participants and were instantly screened for referable DR using a custom offline automated AI system. A total of 236 participants, including 174 from endocrinology and 62 from Aboriginal Medical Services clinics, provided informed consent and 203 (86.0%) were included in the analysis. A total of 33 consenting participants (14%) were excluded from the primary analysis due to ungradable or missing images from small pupils (n = 21, 63.6%), cataract (n = 7, 21.2%), poor fixation (n = 2, 6.1%), technical issues (n = 2, 6.1%), and corneal scarring (n = 1, 3%). The area under the curve, sensitivity, and specificity of the AI system for referable DR were 0.92, 96.9% and 87.7%, respectively. There were 51 disagreements between the reference standard and index test diagnoses, including 29 which were manually graded as ungradable, 21 false positives, and one false negative. A total of 28 participants (11.9%) were referred for follow-up based on new ocular findings, among whom, 15 (53.6%) were able to be contacted and 9 (60%) adhered to referral. Of 207 participants who completed a satisfaction questionnaire, 93.7% specified they were either satisfied or extremely satisfied, and 93.2% specified they would be likely or extremely likely to use this service again. Clinical staff involved in screening most frequently noted that the AI system was easy to use, and the real-time diagnostic report was useful. Our study indicates that AI-assisted DR screening model is accurate and well-accepted by patients and clinicians in endocrinology and indigenous healthcare settings. Future deployments of AI-assisted screening models would require consideration of downstream referral pathways.
PDF
