Evaluating efficacy of 0.125% atropine using a myopia progression machine learning model.

Area of Science:

• Ophthalmology
• Pediatric Ophthalmology
• Artificial Intelligence in Medicine

Background:

• Childhood myopia is a growing public health concern.
• Accurate prediction of myopia progression is crucial for effective intervention.
• Machine learning (ML) models offer potential for predicting disease trajectories.

Purpose of the Study:

• To assess the utility of an ML model in predicting natural myopia progression in children.
• To evaluate the inhibitory effects of 0.125% atropine on childhood myopia progression using the ML model as a comparator.
• To determine the myopia suppression rate attributed to 0.125% atropine treatment.

Main Methods:

• Retrospective cohort study involving 397 children treated with 0.125% atropine eye drops.
• Participants were grouped by treatment duration (6, 12, 18, 24, 30 months).
• Comparison of actual spherical equivalent (SE) with ML-predicted SE to calculate myopia suppression rate.

Main Results:

• Treatment with 0.125% atropine resulted in significantly less myopic outcomes than predicted by the ML model, except in the 6-month group.
• The mean myopia suppression rate was 53.5%, indicating effective inhibition of progression.
• The ML model predicted a more advanced myopia progression compared to the actual treatment outcomes.

Conclusions:

• 0.125% atropine treatment demonstrates efficacy in suppressing myopia progression in children.
• ML models can serve as valuable tools for predicting myopia progression and evaluating treatment effectiveness.
• This approach aids in understanding the natural course of myopia and the impact of interventions.