Distinguishing Dyslexia, Attention Deficit, and Learning Disorders: Insights from AI and Eye Movements

Area of Science:

• Ophthalmology
• Neuroscience
• Pediatrics

Background:

• Eye movement abnormalities are increasingly studied for neurological conditions.
• Differentiating between dyslexia, attention deficit, and learning difficulties in children is clinically significant.
• Objective diagnostic markers are needed to complement clinical assessments.

Purpose of the Study:

• To determine if eye movement patterns can distinguish between dyslexia, attention deficit, and school learning difficulties in children.
• To evaluate the efficacy of machine learning models in classifying these conditions based on eye movement data.
• To explore the diagnostic potential of saccade and vergence analysis.

Main Methods:

• Analysis of saccade and vergence eye movement data from 355 and 454 children, respectively.
• Utilized REMOBI and AIDEAL technologies for data acquisition and AIDEAL software for analysis.
• Trained machine learning models (logistic regression, random forest, SVM, neural networks) using GroupKFold cross-validation.

Main Results:

• Machine learning models accurately identified children with dyslexia based on eye movement data.
• Identification accuracy for attention deficit and learning difficulties was less distinct.
• Kruskal-Wallis test showed significant differences in saccade parameters (velocity, latency) for dyslexic children.

Conclusions:

• Specific eye movement patterns are associated with dyslexia, suggesting unique neurological underpinnings.
• Eye movement analysis shows promise as a diagnostic tool for refining the precision of diagnoses in pediatric neurodevelopmental conditions.
• Further research can leverage eye tracking to better understand and differentiate complex learning and attention disorders.