Purpose Amblyopia is a leading cause of vision impairment among children and young adults. Individual studies showed significant variations in the prevalence of amblyopia in different regions and age groups. This study is to estimate the global prevalence of amblyopia by pooling its prevalence from the previous studies and to project the number of people affected through 2040. Methods We performed a systematic review and meta-analysis on the prevalence of amblyopia using data published before 20 October 2018. We estimated the prevalence rate of amblyopia and its 95% CI globally and by subgroups (eg, region and age). The prevalence data were applied to United Nations World Population Prospects to derive the projected number with amblyopia through 2040. Results A meta-analysis of 60 studies (1 859 327 subjects) showed that the pooled prevalence rate of amblyopia was 1.44% (95% CI 1.17% to 1.78%). Prevalences in Europe (2.90%) and North America (2.41%) were higher than in Asia (1.09%) and Africa (0.72%). The highest prevalence was found in subjects over 20 years old (3.29%). There was no difference in the prevalence between genders. We estimated 99.2 (95% CI 71.7 to 146.1) million people with amblyopia in 2019 worldwide, increasing to 175.2 (95% CI 81.3 to 307.8) million by 2030 and 221.9 (95% CI 83.7 to 429.2) million by 2040. Conclusions The amblyopia is becoming a significant vision problem worldwide. It is of great importance to design and implement amblyopia screening, treatment and related public health strategies.

Purpose To determine the prevalence of amblyopia and strabismus in preschool children in Eastern China. Methods The Nanjing Pediatric Vision Project was a cross-sectional, population-based cohort study conducted in preschool children aged 36–72 months from 2011 to 2012 in Yuhua District, Nanjing, China, using an age-stratified random sampling procedure. A questionnaire totalling 117 items was sent to be completed by the corresponding parents or legal guardians of each eligible child. Comprehensive eye examinations, including visual acuity, anterior segment examination, autorefraction, cover test and ocular motility, were conducted. Postcycloplegic refraction and fundus examinations were performed if necessary. Results Amblyopia was present in 68 children (1.20%, 95% CI 0.92% to 1.48%), with no statistical differences in gender (p=0.903) and age (p=0.835). Among these, 27 had bilateral amblyopia and 41 had unilateral amblyopia, including 26 anisometropic without strabismus, 26 binocular refractive, 7 strabismic, 6 combined strabismic/anisometropic and 3 deprivation amblyopia. Strabismus was found in 320 children (5.65%, 95% CI 5.05% to 6.25%), including 43 with concomitant esotropia, 259 with concomitant exotropia (1 esotropia at near but exotropia at distance), 8 with microtropia (3 esotropia and 5 exotropia), 10 with pure vertical strabismus (3 dissociated vertical deviation and 7 oblique muscle dysfunction) and 1 with Type I Duane syndrome. The prevalence of strabismus had no statistical differences in gender (p=0.952), but had significant statistical differences between different age groups (p=0.021). Conclusions The prevalence of amblyopia and strabismus in preschool children in Eastern China were 1.20% and 5.65%, respectively. The refractive and strabismus are the main factors associated with amblyopia.

Optical coherence tomography (OCT) is a promising high-speed, non-invasive imaging modality providing high-resolution retinal scans. However, a variety of external factors such as light occlusion and patient movement can seriously degrade OCT image quality, which complicates manual retinopathy detection and computer-aided diagnosis. As such, this study first presents an OCT image quality assessment (OCT-IQA) system, capable of automatic classification based on signal completeness, location, and effectiveness. Four CNN architectures (VGG-16, Inception-V3, ResNet-18, and ResNet-50) from the ImageNet classification task were used to train the proposed OCT-IQA system via transfer learning. The ResNet-50 with the best performance was then integrated into the final OCT-IQA network. The usefulness of this approach was evaluated using retinopathy detection results. A retinopathy classification network was first trained by fine-tuning Inception-V3 model. The model was then applied to two test datasets, created randomly from the original dataset, one of which was screened by the OCT-IQA system and only included high quality images while the other was mixed by high and low quality images. Results showed that retinopathy detection accuracy and area under curve (AUC) were 3.75% and 1.56% higher, respectively, for the filtered data (compared with the unfiltered data). These experimental results demonstrate the effectiveness of the proposed OCT-IQA system and suggest that deep learning could be applied to the design of computer-aided systems (CADSs) for automatic retinopathy detection.

Backgrounds: Myopia has become a global public health problem. Children with early onset of myopia are at particular risk of complications associated with myopia. Younger children and children with greater initial myopic refractive errors are at a greater risk of myopia progression. Therefore, it is essential to identify subjects at high risk of developing myopia to facilitate myopia prevention in the early stage, especially during the preschool period. The purpose of this study was to determine whether premyopia and myopia in preschool children can be predicted by easily obtainable parameters. Methods: Data was collected in a population-based cohort. Comprehensive examinations included height, weight, refraction, axial length (AL), and corneal radius of curvature (CR), with a follow-up of 2 years. Parental myopia history was obtained from a questionnaire. Myopia was defined as spherical equivalent (SE) ≤ − 0.50 D. Premyopia was defined as − 0.50 D < SE ≤ + 0.75 D. Multivariate linear regression models were fitted to determine the associations between these parameters at baseline and future SE. To predict premyopia and myopia, Cox proportional hazard regression analysis coupled with a nomogram was used. Results: A total of 830 children (433 boys and 397 girls) were included (40.83 ± 3.43 months old at baseline). A significantly negative relationship was observed in the multivariate analysis between baseline AL, AL/CR, two myopic parents, and the future SE after adjusting for age and gender (coefficient = − 0.291, coefficient = − 5.791, coefficient = − 0.273, respectively, both p < 0.001). Higher baseline AL, AL/CR (hazard ratio (HR) = 4.916, HR = 2.979, respectively, comparing the top quartile with the bottom quartile, both p < 0.001) and two myopic parents (HR = 1.756, compared to no myopic parents, p = 0.001) were associated with a higher risk of future onset of premyopia. From the nomogram, AL/CR was found to have the most enormous effect on survival. Different baseline AL and AL/CR values (both Log Rank p < 0.001) had different survival curves. Conclusions: AL and AL/CR could be used as obtainable indicators for identifying subjects at high risk of developing premyopia and myopia in young preschool children.