Detecting Anxiety Through Song: Mapping Physiological Indicators to Music
Dr. Elaine Biddiss, Research Scientist, University of Toronto/Holland Bloorview Kids Rehabilitation Hospital; Elizabeth Han, MASc Candidate, University of Toronto/Holland Bloorview Kids Rehabilitation Hospital (Main Contact)
Abstract:
Introduction/Background:In pediatric health care, there are many situations in which children experience anxiety, such as during surgery and hospitalization. For children with profound disabilities who cannot communicate through traditional pathways, anxiety may go unnoticed. To improve quality of care for these children, there is a need for a medium through which caregivers can effectively monitor anxiety. Previous research in adults has shown that physiological signals can provide insights into emotional states, especially when multiple signals are integrated and viewed as a whole. However, physiological signals can be complex and difficult to interpret visually, particularly for community-based caregivers who may not have medical training. Currently, there is interest in music as a means to represent physiological signals: it has been shown that music effectively provides biofeedback for management of stress and blood pressure. Thus, we will present the physiological indicators of anxiety through music in hopes that it will be a more intuitive way for caregivers to monitor a child's anxiety.
Objective: This work is part of the long-term goal to provide children, as well as their caregivers, with a medium to easily recognize the physiological indicators of anxiety. To this purpose, we will elucidate the physiological correlates of anxiety in able-bodied children and translate them into musical elements. We call these physiological-musical mappings biosongs. A prototype of the biosongs has been developed previously by Stefanie Blain and Pierre Duez at Holland Bloorview, which will be tested and optimized in a later stage of this research study.
The immediate objective was to determine the most significant physiological features correlated with anxiety in children. These results will inform the mapping of biosignals to musical elements.
Target Population: Although this work was performed with able-bodied children, our results may inform the extension of anxiety-detection via biosongs to children with profound disabilities and their caregivers. As the care of these children is increasingly assumed by community-based care teams, the need for communication mediums that are acceptable and intuitive to individuals from diverse backgrounds and expertise, is growing. Such tools may provide cues to guide caregivers’ decisions, actions and interactions with these children, thus improving the quality of care and quality of life. These tools may also be extended to palliative care of elderly individuals, self management of chronic disease, and patient monitoring.
Methods: We collected physiological signals (electrocardiogram, respiration, skin temperature, blood volume pulse, and electrodermal activity) from 16 able-bodied children aged 8-12 as they performed a time-limited anagram task. We used non-invasive, wearable sensors and an existing signal capture system that has been well used with children at Holland Bloorview (Procomp Infiniti, Thought Technology Inc.). Physiological features were extracted from the data using MATLAB software and signal processing techniques. ANOVA will be used to determine the physiological feature set that is most significantly correlated with anxiety in children.
Results and Conclusions: At the time of this report, physiological data collection has been completed for 6 children out of 16 total. This poster will report preliminary patterns observed.