Utilizing artificial intelligence-based eye tracking technology for screening ADHD symptoms in children.

Area of Science:

• Neuroscience
• Computer Science
• Pediatrics

Background:

• Attention-deficit/hyperactivity disorder (ADHD) is a prevalent neurodevelopmental disorder affecting children.
• Current ADHD screening methods often require specialized clinical settings.
• Objective eye-tracking metrics could offer an objective and accessible screening tool.

Purpose of the Study:

• To evaluate the efficacy of AI-based eye tracking on a tablet for screening ADHD symptoms in children.
• To develop and validate a convolutional neural network model for low-resolution eye gaze prediction.
• To compare eye movement patterns between children with ADHD and typically developing children.

Main Methods:

• Recruited 112 children with ADHD and 325 typically developing children.
• Developed an AI model for eye gaze prediction suitable for low-resolution tablet eye tracking.
• Administered fixation, prosaccade, and antisaccade tasks on a tablet to assess attention and inhibition.

Main Results:

• AI eye tracking demonstrated significant differences in performance between ADHD and typically developing groups across tasks.
• Age and diagnosis significantly impacted performance on fixation, prosaccade, and antisaccade tasks.
• Correlations indicated that higher ADHD symptom scores were associated with specific eye movement patterns, including shorter fixation duration and poorer accuracy.

Conclusions:

• AI-based eye tracking on a tablet can reliably differentiate eye movement patterns between children with and without ADHD.
• This technology presents a potential non-clinical solution for ADHD screening.
• Further validation in diverse settings is warranted to establish its clinical utility.