Age-Based Developmental Biomarkers in Eye Movements: A Retrospective Analysis Using Machine Learning.

Area of Science:

• Neuroscience
• Developmental Psychology
• Ophthalmology

Background:

• Understanding age-related changes in eye movements is crucial for identifying developmental biomarkers.
• Previous research has not comprehensively mapped eye movement trajectories across the entire human lifespan.

Purpose of the Study:

• To identify developmental changes in eye movements across the human lifespan.
• To establish benchmarks for developmental biomarkers using eye movement data.
• To analyze how different types of eye movements vary with age.

Main Methods:

• Utilized a large dataset of 45,696 participants aged 6 to 80 years.
• Administered six standardized eye movement tests measuring fixations, saccades, pursuits, and vergence.
• Employed a semi-supervised, self-training machine learning classifier to define 12 distinct age groups.
• Conducted multiple analyses of variance (MANOVA) to assess the significance of age group effects on eye movement metrics.

Main Results:

• Significant developmental changes in eye movements were identified across all age groups (p < 0.001).
• Eye movement patterns showed similarities between very young (6-14 years) and elderly (69-80 years) individuals.
• Individuals in their 30s demonstrated the optimal eye movement metrics.

Conclusions:

• Eye movement characteristics evolve significantly across the human lifespan.
• The study provides a normative dataset for eye movement development, useful for clinical and research applications.
• Findings can inform health and wellness assessments and guide future research methodologies in aging and development.