What is it about?
Smartphones are ubiquitous for many populations and can function as human sensors that generate a wealth of behavioral data. Existing studies have used smartphones for mental health prediction—for example, depression. To the best of our knowledge, this is the first study that explores the feasibility of using smartphone data for ADHD symptom prediction. In this pilot study, we constructed a family of machine learning prediction models using the features calculated from SMS data collected from college students on the Android platform. Our results indicate that specific ADHD symptoms could be predicted fairly accurately using the smartphone SMS data.
Featured Image
Photo by Manu Del Moral on Unsplash
Why is it important?
ADHD is an impairing neurodevelopmental disorder characterized by age-inappropriate difficulties with inattention and hyperactivity/impulsivity (APA, 2013). It is associated with chronic academic impairment that often persists in post-secondary educational settings (Barkley et al., 2008). Current diagnostic methods rely on self-report, which may introduce recall and other biases, and collateral reports may suffer from their own sources of bias. Our work begins to explore the feasibility of using smartphone sensing data to better understand ADHD in adults and for potential clinical applications.
Perspectives
This work exemplifies the great potential of utilizing smartphone sensing data for understanding ADHD in adults. The behavioral features derived using smartphone data may strongly predict ADHD symptoms. In the future, diagnostic tools designed using these machine learning models might be extremely helpful for clinicians in making treatment decisions.
Dr. Shweta Ware
University of Richmond
Read the Original
This page is a summary of: Predicting ADHD Symptoms Using Smartphone Sensing Data, September 2022, ACM (Association for Computing Machinery),
DOI: 10.1145/3544793.3563430.
You can read the full text:
Contributors
The following have contributed to this page
