Predicting Early Onset of Age-Related Macular Degeneration: A Machine Learning Approach

Area of Science:

• Ophthalmology
• Medical Informatics
• Machine Learning

Background:

• Age-related macular degeneration (AMD) is a leading cause of vision loss.
• Early detection of AMD is crucial for effective management and intervention.
• Predictive models can aid in identifying at-risk individuals for early diagnosis.

Purpose of the Study:

• To develop and validate a machine learning model for predicting early-onset AMD.
• To identify early patient comorbidities associated with the development of early-onset AMD.
• To assess the predictive performance of various machine learning algorithms.

Main Methods:

• Retrospective case-control study utilizing two datasets: a tertiary eye care center cohort and the All of Us Research Program.
• Unsupervised clustering (UMAP) and supervised machine learning models (GBDT, logistic regression, random forests) were employed.
• Model performance was evaluated using AUC, feature importance (Gini index, coefficients), and odds ratios for comorbidities.

Main Results:

• Machine learning models achieved approximately 76% accuracy in predicting early-onset AMD based on comorbidities diagnosed before age 55.
• Inflammatory joint disorders, hypertension, and dyslipidemia were identified as the most significant predictive features.
• Validation confirmed strong associations between essential hypertension, rheumatoid arthritis, hyperlipidemia, and early-onset AMD.

Conclusions:

• Machine learning models can effectively predict early-onset AMD using patient comorbidities.
• Identifying hypertension, dyslipidemia, and inflammatory joint disorders is key for early AMD detection.
• Screening for these comorbidities in ophthalmologic evaluations may facilitate earlier AMD diagnosis and intervention.