Serious Games and Artificial Intelligence: A New Paradigm for ADHD and Dyslexia Assessment

This article introduces a workshop project published by Meril Titus on GitHub, exploring the new direction of combining serious games and artificial intelligence for the assessment of neurodevelopmental disorders such as ADHD and dyslexia. This paradigm collects behavioral data through gamified tools and uses AI to analyze cognitive patterns, aiming to address the limitations of traditional assessments and provide more objective and accessible evaluation solutions.

ADHD and dyslexia affect millions of people worldwide, but traditional assessments have four major limitations: 1. Subjectivity bias (relying on subjective reports); 2. Environmental constraints (significant differences between clinical settings and daily life); 3. Accessibility issues (difficult to obtain in areas with scarce professional resources); 4. Stigma barriers (patient anxiety affects the authenticity of results).

Serious games embed assessment tasks into game scenarios, with advantages including reducing defensive psychology, rich contexts, high data density, and repeatability. Games for ADHD focus on sustained/selective attention, impulse control, and working memory; games for dyslexia focus on skills such as phonological processing, rapid naming, and visual word recognition, with dynamically adjusted difficulty for precise assessment.

AI transforms game data into insights: 1. Behavioral pattern recognition (analyzing reaction time variability, error patterns, eye movement trajectories); 2. Predictive modeling (predicting disorder risks from behavioral data); 3. Personalized assessment (adaptive difficulty adjustment); 4. Multimodal data fusion (integrating behavioral, physiological, and visual data).

This workshop project includes: 1. Workshop report (literature review, theoretical analysis, ethical considerations, etc.); 2. Presentation slides (visual materials for academic presentations); 3. Reference list (academic resources supporting the content). The project was published on June 6, 2026, with the original author being Meril Titus (Marian Engineering College).

Application scenarios include: 1. Early screening (deployment in schools/communities for early identification); 2. Reducing medical burden (concentrating professional resources); 3. Eliminating assessment barriers (lowering psychological thresholds and geographical gaps); 4. Research data accumulation (promoting scientific research on neurodevelopmental disorders).

Current challenges: 1. Validity verification (needs comparison with traditional assessments); 2. Cultural adaptability (game content needs to be adapted to different cultures); 3. Technical thresholds (equipment, digital literacy, etc., affect popularization); 4. Privacy ethics (protection of sensitive data); 5. Risk of over-reliance (cannot replace professional clinical judgment).

Future directions include VR integration, remote assessment, longitudinal tracking, and intervention integration. Conclusion: The combination of serious games and AI opens up new possibilities for assessment. Although it cannot replace professional judgment, it can serve as an auxiliary tool to help more people get timely support, reflecting the potential of technology for good.