BrainGaze: A new Digital Tool to Support the Diagnosis of Adult ADHD

Introduction: Visuospatial attention deficits are a key characteristic of ADHD. BrianGaze is an innovative technology that leverages eye-tracking to capture changes in the angle of eye vergence during an attention task. This angle, formed between lines of sight when the eyes converge on a focal point, reflects shifts in visuospatial attention. During fixation, its amplitude is modulated by the ability of a peripheral stimuli to draw attention.
Methods: One hundred and eight adults (18-65 years old) with ADHD were compared to age-matched clinical controls with conduct problems but not diagnosed with ADHD (n=38). On a computer, participants were asked to discriminate cartoon images of a tadpole from a fish while fixating on a central point surrounded by eight possible target locations, with or without informative cues on where the stimuli were presented (cue/no cue conditions). The eye tracking was calibrated for each participant before the cognitive task.
Results: No difference in performance emerged between groups except for reaction time. In both cue conditions, the modulation of angle of eye vergence was weaker in individuals with ADHD. A machine learning model based on vergence signal data and blinded to ADHD status, differentiated the two groups with 79% accuracy, a false positive rate of 25%, and a false negative rate of 20.55%.
Conclusion: BrianGaze is an innovative technology that can support the diagnosis of ADHD in adults based on changes in a digital biomarker (eye vergence) during attention tasks.