[Research progress on the development of myopia prediction models and their predictive performance].

Area of Science:

• Ophthalmology
• Public Health
• Biomedical Engineering

Background:

• Myopia incidence is rising globally, particularly in China, necessitating effective prevention and control strategies.
• Early prediction and intervention are vital for mitigating myopia progression and its associated societal costs.
• Existing prediction models vary in approach, highlighting the need for comprehensive reviews.

Purpose of the Study:

• To review recent advancements in myopia prediction models and evaluate their performance.
• To explore diverse predictive factors including ocular characteristics, environmental influences, and genetic markers.
• To examine the role of artificial intelligence in enhancing myopia prediction accuracy.

Main Methods:

• Review of literature on myopia prediction models, encompassing refractive power and ocular biological characteristics (e.g., axial length).
• Analysis of models incorporating environmental factors (e.g., age of onset, parental myopia, education, living conditions).
• Inclusion of genetic factor-based models (parental myopia, SNPs from GWAS) and artificial intelligence applications (machine learning, deep learning).

Main Results:

• Models utilizing refractive power and ocular biometrics offer valuable predictive insights.
• Environmental and genetic factors significantly correlate with myopia development in school-aged children.
• AI algorithms demonstrate potential for predicting axial length growth and refractive power changes with high accuracy.

Conclusions:

• Advances in myopia prediction models provide a foundation for more precise and personalized interventions.
• Integrating diverse data sources (ocular, environmental, genetic, AI) can improve predictive accuracy.
• Enhanced myopia prediction is key to effective, individualized prevention and control strategies.