Paper Key : IRJ************884
Author: Ranjith Venkatachala,Suresh Mariyappa
Date Published: 31 Oct 2023
Abstract
Attention-deficithyperactivity disorder (ADHD) is a prevalent neurodevelopmental disorder with significant impacts on daily functioning. This paper explores the potential of smartphone data for early ADHD detection. We begin with a comprehensive literature review on ADHD and detection methods, emphasizing the limitations of subjective assessments. Smartphone data, collected through various sensors, offers real-time and objective measurements, presenting a promising avenue for early detection. The study employs a cross-sectional design, recruiting participants meeting DSM-5 criteria, and collects smartphone data using the AWARE application. Data preprocessing, feature selection, and classification algorithms are applied to identify significant predictors for ADHD detection. The authors further illustrate that, by employing machine learning models that are not specific to individual subjects, they can predict eating events with an AUROC (Area Under the Receiver Operating Characteristics Curve) of 0.74. The results show correlations between specific smartphone data variables and ADHD symptoms. High application usage correlates with hyperactivity, while location data relates to attention symptoms. Accelerometer and gyroscope data correlate with motor hyperactivity. This research underscores the potential of smartphone data for enhancing ADHD diagnosis and timely interventions.
DOI LINK : 10.56726/IRJMETS45695 https://www.doi.org/10.56726/IRJMETS45695