Effectiveness of Music-Based Respiratory Biofeedback in Reducing Stress during Visually Demanding Tasks
MetadataShow full item record
Biofeedback techniques have shown to be effective to manage stress and improve task performance. Biofeedback generally can be divided into two steps (i) measuring physiological functions (e.g. respiration, heart rate) via sensors and (ii) conveying the physiological signals to the user to improve self-awareness. Current systems require costly and invasive sensors to measure physiology, which are not comfortable and are not readily accessible to the general population. Additionally, current feedback mechanisms may be physically unpleasant or may hinder multitasking, especially in visually-demanding environments. To overcome these problems, we developed two tools: a music-based biofeedback tool that uses music as the medium of feedback, and a tool to measure breathing rate using a smartphone camera. The music biofeedback tool encourages slow breathing by adjusting the quality of the music in response to the user’s breathing rate. This intervention combines the benefits of biofeedback and music to help users regulate their stress response while performing a visual task (driving a car simulator). We evaluate the intervention on a 2×2 design with music and auditory biofeedback as independent variables. Our results indicate that music-biofeedback leads to lower arousal (as measured by electrodermal activity and heart rate variability) than music alone, auditory biofeedback alone, and a control condition. Music biofeedback also reduces driving errors when compared to the other three conditions. While our results suggest that the music-based biofeedback tool is useful and enjoyable, it still requires expensive physiological sensors which are intrusive in nature. Hence, we present a second tool to measure breathing rate in real-time via smartphone camera, which makes it easily accessible given the pervasiveness of smartphones. Our algorithm measures breathing rate by obtaining the photoplethysmographic signal and performing spectral analysis using Goertzel algorithm. We validated the method under a range of controlled breathing rate conditions, and our results show a high degree of agreement between our estimates and ground truth measurements obtained via standard respiratory sensors. These results show that it is possible to accurately compute breathing rate in real-time using a smartphone.
Bhandari, Rhushabh Rajendra (2015). Effectiveness of Music-Based Respiratory Biofeedback in Reducing Stress during Visually Demanding Tasks. Master's thesis, Texas A & M University. Available electronically from