Smartwatch Biomarker for Disruptive Behaviour
A feasibility study showing that a week of continuous smartwatch data can predict impending disruptive behavior in children — a direct pediatric proof point for anticipatory, wearable-based support.Rana Elmaghraby, Arjun P. Athreya, Jennifer L. Vande Voort, Julia Shekunov, Magdalena Romanowicz, Michelle Skime, Kyle Croarkin, Paul E. Croarkin
For children with severe disruptive behaviors, meltdowns and outbursts often seem to arrive without warning — making them hard to prevent. This feasibility study tested whether a consumer smartwatch, worn continuously for seven days, could capture physiological signals that precede those behaviors. Using machine learning on the collected data, the researchers identified predictive patterns that anticipated a child's behavioral state with roughly 81% accuracy, pointing toward a future where wearables could give caregivers and clinicians advance warning rather than after-the-fact reaction.
Link to article: