Longitudinal Digital Phenotyping for Early Cognitive-Motor Screening

2026-03-26

AI Summary

This study addresses the limitations of traditional cognitive-motor developmental assessments, which rely on subjective and static methods and struggle to enable early, objective, and continuous screening. The authors propose an AI-driven longitudinal digital phenotyping framework that leverages years of tablet-based interaction data from children. Using unsupervised learning techniques—including t-SNE for dimensionality reduction and K-Means++ for clustering—the framework analyzes digital biomarkers derived from six cognitive-motor tasks. Three stable developmental trajectories—low, medium, and high—were identified, with over 90% of children in the low-performing group showing persistent deficits without intervention, underscoring the high continuity of early unaddressed impairments. This approach establishes an objective data foundation and a novel paradigm for scalable early screening and personalized intervention in child development.

Technology Category

• Cognitive Modeling & Cognitive Systems: Computational Creativity
• Humans and AI: Other Foundations of Human Computation & AI
• Search and Optimization: Learning to Search

Application Category

• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success

Abstract

Early detection of atypical cognitive-motor development is critical for timely intervention, yet traditional assessments rely heavily on subjective, static evaluations. The integration of digital devices offers an opportunity for continuous, objective monitoring through digital biomarkers. In this work, we propose an AI-driven longitudinal framework to model developmental trajectories in children aged 18 months to 8 years. Using a dataset of tablet-based interactions collected over multiple academic years, we analyzed six cognitive-motor tasks (e.g., fine motor control, reaction time). We applied dimensionality reduction (t-SNE) and unsupervised clustering (K-Means++) to identify distinct developmental phenotypes and tracked individual transitions between these profiles over time. Our analysis reveals three distinct profiles: low, medium, and high performance. Crucially, longitudinal tracking highlights a high stability in the low-performance cluster (>90% retention in early years), suggesting that early deficits tend to persist without intervention. Conversely, higher-performance clusters show greater variability, potentially reflecting engagement factors. This study validates the use of unsupervised learning on touchscreen data to uncover heterogeneous developmental paths. The identified profiles serve as scalable, data-driven proxies for cognitive growth, offering a foundation for early screening tools and personalized pediatric interventions.

Problem

• cognitive-motor development
• early screening
• digital phenotyping
• longitudinal monitoring
• developmental trajectories

Innovation

• longitudinal digital phenotyping
• unsupervised learning
• developmental trajectories
• touchscreen biomarkers
• cognitive-motor screening