AI Eye Tracking Screens for ADHD
A new study demonstrates that AI-based eye tracking technology may offer a non-clinical tool for ADHD screening in children. This technology could provide quicker, less invasive ways to identify ADHD symptoms compared to traditional assessment methods.
This new screening method, detailed in the journal *Frontiers in Psychiatry*, involved 112 children with ADHD and 325 typically developing children. The AI, running on a standard tablet, tracked eye movements during simple tasks like focusing on a single point and shifting the gaze toward or away from a target. The study, led by researcher Xiaolu Chen, found statistically significant differences in the eye movement patterns between the two groups. Children with ADHD tended to have shorter fixation durations and made more corrective eye movements during tasks that required inhibitory control. This objective, data-driven approach contrasts sharply with traditional ADHD assessments, which often rely on subjective behavioral checklists and interviews with parents and teachers. These established methods can be hampered by the overlap of ADHD symptoms with other conditions like anxiety or learning disabilities, and are also subject to observer bias. While this specific study demonstrated it could "reliably discriminate" between the groups, other research into AI-based ADHD screening has shown high precision. A separate Korean study using a machine learning model and portable eye-tracking achieved a 76.3% accuracy in identifying ADHD. The ultimate goal of these technologies is to create a more accessible and objective tool for early ADHD screening. By moving assessments out of specialized labs and onto common devices like tablets, researchers hope to speed up the identification process, allowing for earlier intervention.
