Machine Learning to Determine Risk Factors for Myopia Progression in Primary School Children: The Anyang Childhood Eye Study

Artificial intelligence in myopia in children: current trends and future directions

PURPOSE OF REVIEW

Myopia is one of the major causes of visual impairment globally, with myopia and its complications thus placing a heavy healthcare and economic burden. With most cases of myopia developing during childhood, interventions to slow myopia progression are most effective when implemented early. To address this public health challenge, artificial intelligence has emerged as a potential solution in childhood myopia management.

RECENT FINDINGS

The bulk of artificial intelligence research in childhood myopia was previously focused on traditional machine learning models for the identification of children at high risk for myopia progression. Recently, there has been a surge of literature with larger datasets, more computational power, and more complex computation models, leveraging artificial intelligence for novel approaches including large-scale myopia screening using big data, multimodal data, and advancing imaging technology for myopia progression, and deep learning models for precision treatment.

SUMMARY

Artificial intelligence holds significant promise in transforming the field of childhood myopia management. Novel artificial intelligence modalities including automated machine learning, large language models, and federated learning could play an important role in the future by delivering precision medicine, improving health literacy, and allowing the preservation of data privacy. However, along with these advancements in technology come practical challenges including regulation and clinical integration.

Bibliometric analysis of hotspots and trends of global myopia research

AIM

To gain insights into the global research hotspots and trends of myopia.

METHODS

Articles were downloaded from January 1, 2013 to December 31, 2022 from the Science Core Database website and were mainly statistically analyzed by bibliometrics software.

RESULTS

A total of 444 institutions in 87 countries published 4124 articles. Between 2013 and 2022, China had the highest number of publications (n=1865) and the highest H-index (61). Sun Yat-sen University had the highest number of publications (n=229) and the highest H-index (33). Ophthalmology is the main category in related journals. Citations from 2020 to 2022 highlight keywords of options and reference, child health (pediatrics), myopic traction mechanism, public health, and machine learning, which represent research frontiers.

CONCLUSION

Myopia has become a hot research field. China and Chinese institutions have the strongest academic influence in the field from 2013 to 2022. The main driver of myopic research is still medical or ophthalmologists. This study highlights the importance of public health in addressing the global rise in myopia, especially its impact on children's health. At present, a unified theoretical system is still needed. Accurate surgical and therapeutic solutions must be proposed for people with different characteristics to manage and intervene refractive errors. In addition, the benefits of artificial intelligence (AI) models are also reflected in disease monitoring and prediction.

Development and validation of predictive models for myopia onset and progression using extensive 15-year refractive data in children and adolescents

BACKGROUND

Global myopia prevalence poses a substantial public health burden with vision-threatening complications, necessitating effective prevention and control strategies. Precise prediction of spherical equivalent (SE), myopia, and high myopia onset is vital for proactive clinical interventions.

METHODS

We reviewed electronic medical records of pediatric and adolescent patients who underwent cycloplegic refraction measurements at the Eye & Ear, Nose, and Throat Hospital of Fudan University between January 2005 and December 2019. Patients aged 3-18 years who met the inclusion criteria were enrolled in this study. To predict the SE and onset of myopia and high myopia in a specific year, two distinct models, random forest (RF) and the gradient boosted tree algorithm (XGBoost), were trained and validated based on variables such as age at baseline, and SE at various intervals. Outputs included SE, the onset of myopia, and high myopia up to 15 years post-initial examination. Age-stratified analyses and feature importance assessments were conducted to augment the clinical significance of the models.

RESULTS

The study enrolled 88,250 individuals with 408,255 refraction records. The XGBoost-based SE prediction model consistently demonstrated robust and better performance than RF over 15 years, maintaining an R2 exceeding 0.729, and a Mean Absolute Error ranging from 0.078 to 1.802 in the test set. Myopia onset prediction exhibited strong area under the curve (AUC) values between 0.845 and 0.953 over 15 years, and high myopia onset prediction showed robust AUC values (0.807-0.997 over 13 years, with the 14th year at 0.765), emphasizing the models' effectiveness across age groups and temporal dimensions on the test set. Additionally, our classification models exhibited excellent calibration, as evidenced by consistently low brier score values, all falling below 0.25. Moreover, our findings underscore the importance of commencing regular examinations at an early age to predict high myopia.

CONCLUSIONS

The XGBoost predictive models exhibited high accuracy in predicting SE, onset of myopia, and high myopia among children and adolescents aged 3-18 years. Our findings emphasize the importance of early and regular examinations at a young age for predicting high myopia, thereby providing valuable insights for clinical practice.

Development of a risk score for myopia: A cohort study conducted among school-aged children in China

PURPOSE

To evaluate the myopia risk in school-aged children one year after lifting a pandemic-related lockdown and develop a tool to identify high-risk groups.

METHODS

In total, 38,079 children without myopia from 38 schools were included. The outcomes were myopia incidence and progression in 1 year after the COVID-19 lockdown was lifted, both obtained by the spherical equivalent refraction (SER). We separated the population into an exploratory (75%) and a validation sample (25%) to construct the risk score model.

RESULTS

In total, 9811 (29.57%) students became myopic, and the overall myopia progression was 0.22 ± 0.62 D. Even less myopia progression was noted in the pre-myopia group at baseline (All: P = 0.045, Boy: P = 0.005). The risk score model included seven predictors: gender, grade, SER at baseline, residence, parental myopia, eye discomfort symptoms, and online courses. The model had a score range of 0-46 and an optimal cutoff of 34. The area under the receiver operating curve of the model was 0.726 (0.719-0.732) for the exploratory sample and 0.731 (0.720-0.742) for the validation sample.

CONCLUSIONS

The risk score can serve as a practical tool for classifying the risk of myopia in school-aged children.

Insights into artificial intelligence in myopia management: from a data perspective

Given the high incidence and prevalence of myopia, the current healthcare system is struggling to handle the task of myopia management, which is worsened by home quarantine during the ongoing COVID-19 pandemic. The utilization of artificial intelligence (AI) in ophthalmology is thriving, yet not enough in myopia. AI can serve as a solution for the myopia pandemic, with application potential in early identification, risk stratification, progression prediction, and timely intervention. The datasets used for developing AI models are the foundation and determine the upper limit of performance. Data generated from clinical practice in managing myopia can be categorized into clinical data and imaging data, and different AI methods can be used for analysis. In this review, we comprehensively review the current application status of AI in myopia with an emphasis on data modalities used for developing AI models. We propose that establishing large public datasets with high quality, enhancing the model's capability of handling multimodal input, and exploring novel data modalities could be of great significance for the further application of AI for myopia.

Associations between meeting 24-hour movement guidelines and myopia among school-aged children: A cross-sectional study

BACKGROUND

The Canadian 24-hour movement behavior (24-HMB) guidelines recommend an adequate level of physical activity (PA), a limited amount of screen time (ST), and a sufficient sleep duration (SLP) to promote the healthy development of children. Although the positive effects of adhering to the 24-HMB guidelines have been established for several health parameters, less is known about how adherence to the 24-HMB guidelines relates to the myopia risk (i.e., inability to see distant objects properly). Thus, this study investigated associations between meeting 24-HMB guidelines and myopia risk in school-aged children.

METHOD

Using a questionnaire survey, this cross-sectional study was conducted among parents of school-aged children (5-13 years) in China from 15th September to 15th October 2022, with a total of 1423 respondents with complete data for analysis. Parents reported their child's time spent in moderate-to-vigorous-intensity physical activity (MVPA), SLP, and ST. Multiple logistic regression analyses were performed to examine the associations between measures of PA, ST, and SLP alone and in combination, and the occurrence of myopia.

RESULTS

A relatively low percentage of the children being included in the current study (4.92%) met all 24-HMB guidelines, while 32.46% had myopia. Girls had a significantly higher risk of myopia compared to boys (OR = 1.3, 1.002 to 1.68, p = 0.049). Children of parents without myopia had a lower risk of myopia (OR = 0.45, 0.34-0.59, p < 0.001). Children who lived in urban areas (OR = 1.83, 95% CI 1.33 to 2.52, p < 0.001) or towns (OR = 1.60, 1.03 to 2.47, p = 0.04) had a significantly higher risk of myopia compared to those living in rural areas. Meeting SLP guidelines (OR = 0.50, 95% CI 0.31 to 0.82, p < 0.01), meeting ST + SLP guidelines (OR = 0.47, 95% CI 0.32-0.69, <0.001), and meeting all three guidelines were associated with significantly lower risk of myopia (OR = 0.40, 95% CI 0.20-0.82, p = 0.01). Meeting more 24-HMB guidelines was associated with a reduced risk of myopia.

CONCLUSIONS

Our data suggest that adhering to SLP, ST + SLP, and ST + SLP + PA guidelines is associated with the risk of myopia. Future research investigating dose-response associations, and potential mechanisms, is necessary to achieve a more nuanced understanding of the observed associations.

Artificial intelligence-aided diagnosis and treatment in the field of optometry

With the rapid development of computer technology, the application of artificial intelligence (AI) to ophthalmology has gained prominence in modern medicine. As modern optometry is closely related to ophthalmology, AI research on optometry has also increased. This review summarizes current AI research and technologies used for diagnosis in optometry, related to myopia, strabismus, amblyopia, optical glasses, contact lenses, and other aspects. The aim is to identify mature AI models that are suitable for research on optometry and potential algorithms that may be used in future clinical practice.

Machine learning to analyze the factors influencing myopia in students of different school periods

Purpose

We aim to develop myopia classification models based on machine learning algorithms for each schooling period, and further analyze the similarities and differences in the factors influencing myopia in each school period based on each model.

Design

Retrospective cross-sectional study.

Participants

We collected visual acuity, behavioral, environmental, and genetic data from 7,472 students in 21 primary and secondary schools (grades 1-12) in Jiamusi, Heilongjiang Province, using visual acuity screening and questionnaires.

Methods

Machine learning algorithms were used to construct myopia classification models for students at the whole schooling period, primary school, junior high school, and senior high school period, and to rank the importance of features in each model.

Results

The main influencing factors for students differ by school section, The optimal machine learning model for the whole schooling period was Random Forest (AUC = 0.752), with the top three influencing factors being age, myopic grade of the mother, and Whether myopia requires glasses. The optimal model for the primary school period was a Random Forest (AUC = 0.710), with the top three influences being the myopic grade of the mother, age, and extracurricular tutorials weekly. The Junior high school period was an Support Vector Machine (SVM; AUC = 0.672), and the top three influencing factors were gender, extracurricular tutorial subjects weekly, and whether can you do the "three ones" when reading and writing. The senior high school period was an XGboost (AUC = 0.722), and the top three influencing factors were the need for spectacles for myopia, average daily time spent outdoors, and the myopic grade of the mother.

Conclusion

Factors such as genetics and eye use behavior all play an essential role in students' myopia, but there are differences between school periods, with those in the lower levels focusing on genetics and those in the higher levels focusing on behavior, but both play an essential role in myopia.

Artificial intelligence and digital solutions for myopia

Myopia as an uncorrected visual impairment is recognized as a global public health issue with an increasing burden on health-care systems. Moreover, high myopia increases one's risk of developing pathologic myopia, which can lead to irreversible visual impairment. Thus, increased resources are needed for the early identification of complications, timely intervention to prevent myopia progression, and treatment of complications. Emerging artificial intelligence (AI) and digital technologies may have the potential to tackle these unmet needs through automated detection for screening and risk stratification, individualized prediction, and prognostication of myopia progression. AI applications in myopia for children and adults have been developed for the detection, diagnosis, and prediction of progression. Novel AI technologies, including multimodal AI, explainable AI, federated learning, automated machine learning, and blockchain, may further improve prediction performance, safety, accessibility, and also circumvent concerns of explainability. Digital technology advancements include digital therapeutics, self-monitoring devices, virtual reality or augmented reality technology, and wearable devices - which provide possible avenues for monitoring myopia progression and control. However, there are challenges in the implementation of these technologies, which include requirements for specific infrastructure and resources, demonstrating clinically acceptable performance and safety of data management. Nonetheless, this remains an evolving field with the potential to address the growing global burden of myopia.

Prevalence Patterns and Onset Prediction of High Myopia for Children and Adolescents in Southern China via Real-World Screening Data: Retrospective School-Based Study

BACKGROUND

Patients with high myopia have an increased lifetime risk of complications. The prevalence patterns of high myopia in children and adolescents in southern China are unclear. Early identification of high-risk individuals is critical for reducing the occurrence and development of high myopia and avoiding the resulting complications.

OBJECTIVE

This study aimed to determine the prevalence of high myopia in children and adolescents in southern China via real-world screening data and to predict its onset by studying the risk factors for high myopia based on machine learning.

METHODS

This retrospective school-based study was conducted in 13 cities with different gross domestic products in southern China. Through data acquisition and filtering, we analyzed the prevalence of high myopia and its association with age, school stage, gross domestic product, and risk factors. A random forest algorithm was used to predict high myopia among schoolchildren and then assessed in an independent hold-out group.

RESULTS

There were 1,285,609 participants (mean age 11.80, SD 3.07, range 6-20 years), of whom 658,516 (51.2%) were male. The overall prevalence of high myopia was 4.48% (2019), 4.88% (2020), and 3.17% (2021), with an increasing trend from the age of 11 to 17 years. The rates of high myopia increased from elementary schools to high schools but decreased at all school stages from 2019 to 2021. The coastal and southern cities had a higher proportion of high myopia, with an overall prevalence between 2.60% and 5.83%. Age, uncorrected distance visual acuity, and spherical equivalents were predictive factors for high myopia onset in schoolchildren. The random forest algorithm achieved a high accuracy of 0.948. The area under the receiver operator characteristic curve (AUC) was 0.975. Both indicated sufficient model efficacy. The performance of the model was validated in an external test with high accuracy (0.971) and a high AUC (0.957).

CONCLUSIONS

High myopia had a high incidence in Guangdong Province. Its onset in children and adolescents was well predicted with the random forest algorithm. Efficient use of real-world data can contribute to the prevention and early diagnosis of high myopia.

Artificial intelligence technology for myopia challenges: A review

Myopia is a significant global health concern and affects human visual function, resulting in blurred vision at a distance. There are still many unsolved challenges in this field that require the help of new technologies. Currently, artificial intelligence (AI) technology is dominating medical image and data analysis and has been introduced to address challenges in the clinical practice of many ocular diseases. AI research in myopia is still in its early stages. Understanding the strengths and limitations of each AI method in specific tasks of myopia could be of great value and might help us to choose appropriate approaches for different tasks. This article reviews and elaborates on the technical details of AI methods applied for myopia risk prediction, screening and diagnosis, pathogenesis, and treatment.

Deep Learning-Based Mental Health Model on Primary and Secondary School Students’ Quality Cultivation

The purpose was to timely identify the mental disorders (MDs) of students receiving primary and secondary education (PSE) (PSE students) and improve their mental quality. Firstly, this work analyzes the research status of the mental health model (MHM) and the main contents of PSE student-oriented mental health quality cultivation under deep learning (DL). Secondly, an MHM is implemented based on big data technology (BDT) and the convolutional neural network (CNN). Simultaneously, the long short-term memory (LSTM) is introduced to optimize the proposed MHM. Finally, the performance of the MHM before and after optimization is evaluated, and the PSE student-oriented mental health quality training strategy based on the proposed MHM is offered. The results show that the accuracy curve is higher than the recall curve in all classification algorithms. The maximum recall rate is 0.58, and the minimum accuracy rate is 0.62. The decision tree (DT) algorithm has the best comprehensive performance among the five different classification algorithms, with accuracy of 0.68, recall rate of 0.58, and F1-measure of 0.69. Thus, the DT algorithm is selected as the classifier. The proposed MHM can identify 56% of students with MDs before optimization. After optimization, the accuracy is improved by 0.03. The recall rate is improved by 0.19, the F1-measure is improved by 0.05, and 75% of students with MDs can be identified. Diverse behavior data can improve the recognition effect of students' MDs. Meanwhile, from the 60th iteration, the mode accuracy and loss tend to be stable. By comparison, batch_size has little influence on the experimental results. The number of convolution kernels of the first convolution layer has little influence. The proposed MHM based on DL and CNN will indirectly improve the mental health quality of PSE students. The research provides a reference for cultivating the mental health quality of PSE students.

Predicting risk of overweight or obesity in Chinese preschool-aged children using artificial intelligence techniques

OBJECTIVES

We adopted the machine-learning algorithms and deep-learning sequential model to determine and optimize most important factors for overweight and obesity in Chinese preschool-aged children.

METHODS

This is a cross-sectional survey conducted in 2020 at Beijing and Tangshan. Using a stratified cluster random sampling strategy, children aged 3-6 years were enrolled. Data were analyzed using the PyCharm and Python.

RESULTS

A total of 9478 children were eligible for inclusion, including 1250 children with overweight or obesity. All children were randomly divided into the training group and testing group at a 6:4 ratio. After comparison, support vector machine (SVM) outperformed the other algorithms (accuracy: 0.9457), followed by gradient boosting machine (GBM) (accuracy: 0.9454). As reflected by other 4 performance indexes, GBM had the highest F1 score (0.7748), followed by SVM with F1 score at 0.7731. After importance ranking, the top 5 factors seemed sufficient to obtain descent performance under GBM algorithm, including age, eating speed, number of relatives with obesity, sweet drinking, and paternal education. The performance of the top 5 factors was reinforced by the deep-learning sequential model.

CONCLUSIONS

We have identified 5 important factors that can be fed to GBM algorithm to better differentiate children with overweight or obesity from the general children, with decent prediction performance.

Novel Uses and Challenges of Artificial Intelligence in Diagnosing and Managing Eyes with High Myopia and Pathologic Myopia

Myopia is a global health issue, and the prevalence of high myopia has increased significantly in the past five to six decades. The high incidence of myopia and its vision-threatening course emphasize the need for automated methods to screen for high myopia and its serious form, named pathologic myopia (PM). Artificial intelligence (AI)-based applications have been extensively applied in medicine, and these applications have focused on analyzing ophthalmic images to diagnose the disease and to determine prognosis from these images. However, unlike diseases that mainly show pathologic changes in the fundus, high myopia and PM generate even more data because both the ophthalmic information and morphological changes in the retina and choroid need to be analyzed. In this review, we present how AI techniques have been used to diagnose and manage high myopia, PM, and other ocular diseases and discuss the current capacity of AI in assisting in preventing high myopia.

Corneal morphology correlates with choriocapillaris perfusion in myopic children

AIMS

The present study investigated the difference in choriocapillaris (CC) perfusion between different AL/K ratio groups with similar spherical equivalent refraction (SER) and analyzed factors affecting CC perfusion.

METHODS

This cross-sectional study included 129 children with low-to-moderate myopia. Axial length (AL), average K-reading (Ave-K), and SER were measured. Choroidal vascularity, including the total choroidal area (TA), choroidal luminal area (LA), stromal area (SA), choroidal vascularity index (CVI), CC flow voids (FVs), and FVs%, was obtained using optical coherence tomography angiography.

RESULTS

Participants with similar SER were divided into two groups (high AL/K ratio, n = 57; low AL/K ratio, n = 72). The high AL/K group had lower LA, TA, and CVI (P < 0.01) and lower FVs (inner ring and fovea, P < 0.05) and FVs% (outer ring, inner ring, and fovea, P < 0.05). The AL/K ratio and FVs% were negatively correlated in the outer ring (r =  - 0.174, P < 0.05) and inner ring (r =  - 0.174, P < 0.05). The Ave-K and inner FVs (r = 0.178, P < 0.05), outer FVs% (r = 0.175, P < 0.05), and inner FVs% (r = 0.196, P < 0.05) were positively correlated. In stepwise multiple regression for the outer ring, the horizontal CVI was related to FVs (β = 0.175, P < 0.05), and the vertical CVI was related to FVs% (β = 0.232, P < 0.01). Independent risk factors associated with inner FVs area were vertical CVI (β = 0.329; P < 0.001) and SER (β =  - 0.196, P < 0.05); FVs% was also associated with vertical CVI (β = 0.360, P < 0.01) and SER (β =  - 0.196, P < 0.05).

CONCLUSION

With a similar SER, myopic eyes with a higher AL/K ratio maintained more CC perfusion and lower CVI, which may indicate rapid myopic progression. Low K-reading eyes had more CC perfusion and less CVI, which may explain the relatively poor myopia control efficacy in the clinic.