Utilising pupillometry as a biomarker for diagnosing ADHD

Abstract Neurodivergent conditions, notably Attention-deficit Hyperactivity Disorder (ADHD) and Autism Spectrum Disorder (ASD), pose considerable challenges in unraveling their pathophysiology due to intricate genetic underpinnings and diverse symptom presentations. The locus coeruleus-norepinephrine (LC-NE) system, central to noradrenergic signaling crucial for cognitive processes like sustained attention, has emerged as a focal point for understanding these conditions. However, the LC's small size and deep brainstem location pose significant hurdles to its cellular, molecular, and physiological investigation. Given these challenges, this study explores pupillometry as a non-invasive, objective measure of cognitive load reflecting LC-NE activity. The aim is to establish pupillometry as a novel biomarker and potentially an endophenotype for ADHD. Leveraging an elastic-net logistic model alongside meticulous five-fold cross-validation and hyperparameter tuning, our research yields promising evidence for future research, achieving an Area under the Curve (AUC) score of 0.969, accompanied by a precision of 100% and a recall of 86%. These findings underscore the efficacy of pupillometry in discerning dysfunction within the LC-NE system, thereby enabling accurate diagnosis of ADHD in pediatric populations.