Developing Intelligent Garments for Real-Time Healthcare
ECE Associate Professor Kris Dorsey and Khoury/MIE Assistant Professor Megan Hofmann, in collaboration with Emory University, were awarded a $699,789 NSF grant for “Adaptive Intelligent Healthcare Garment: Advancing Real-Time Monitoring and Behavioral Interventions.”
Abstract Source: NSF
This research project aims to prototype and evaluate an intelligent garment featuring an adaptive splint that clinicians can dynamically control to address the treatment goals of individuals with neurodevelopmental disabilities who engage in self-injurious behavior. This novel device will facilitate sensorimotor interaction between the adaptive splint and the individual. Additionally, the garment will communicate with therapists to enhance their understanding of the individual’s responses to treatment decisions and to predict and analyze attempts at self-injurious behavior. Self-injurious behaviors in minimally verbal individuals with neurodevelopmental disorders can lead to significant physical, emotional, social, and economic challenges. Behavioral interventions currently represent the most established approach to managing self-injurious behaviors. Presently, interventions for severe cases involve individuals wearing rigid splints on their arms, which help mitigate injury risk while still allowing the behavior to occur. If successful, this research is anticipated to improve the quality of life for both patients and clinicians alike.
The adaptive garment prototype will feature an active split that can adjust its mechanical stiffness in real time with a clinician-in-the-loop feedback mechanism. This research initiative has several key interdisciplinary objectives: 1) to assess the effectiveness of the first active splint designed for behavioral interventions, drawing on design and actuation principles from soft robotics, which is capable of withstanding self-injurious behaviors; 2) to address the challenge of accurately and automatically measuring the frequency, intensity, and duration of self-injurious behaviors in therapeutic settings, utilizing models developed from retrospective data gathered through ambulatory accelerometry, electrodermal activity, and photoplethysmography via a wearable biosensor, in conjunction with video data; 3) to analyze and evaluate the interactions between individuals and therapists during behavioral intervention sessions that incorporate the garment; and 4) to examine how the garment impacts therapists’ decision-making processes. If successful, this research program will lay the foundation for healthcare garments capable of reasoning and modeling wearer behaviors to improve therapeutic outcomes.