Designing Biofeedback for Managing Stress
Mainly provoked by increasing stress-related health problems and driven by recent technological advances in human-computer interaction (HCI), the ubiquitous physiologically-relevant information will potentially transform the role of biofeedback from clinical treatment to a readily available tool for personal stress management. The primary motivation for this thesis is to bring biofeedback techniques closer to everyday use so that the average people can harness it more intuitively, effortlessly and comfortably.
The first part of the thesis aims to understand the current status of biofeedback technology in the context of stress management. We first decompose a biofeedback-assisted self-regulation process from the perspective of an information pipeline. Then we present an overview towards state of the art relevant to the biofeedback techniques for stress management and relaxation training. A systematic literature review is carried out to classify biofeedback systems for stress management, regarding bio-sensing technique, bio-data processing approach, biofeedback protocol, feedback modality, and evaluation approach. This systematic review helps us to identify the challenges and opportunities for biofeedback design.
In the second part of this thesis, we present four novel biofeedback interfaces that are developed with various HCI technologies, such as sonification, metaphorical visualization, shape-changing displays, and haptic interfaces. The ‘presentation mapping’ in these biofeedback displays embodies a similar design rationale around the idea of ‘natural coupling’. The user studies provide sufficient confirmation of the design rationales presented. The natural coupling in presentation mapping may help users associate the interface expressions with their physiological activities and specific meanings relevant to life and health. These associations make the biofeedback representations meaningful and further facilitate the users to understand and leverage the biofeedback information intuitively in their self-regulation and self-reflection.
In chapter 3, we apply the idea of nature coupling to the design of an audio interface for heart rate variability (HRV) biofeedback. The heart rate variability is presented by the changing rhythm of short melodies. In this design, the timing variations of heart rate are directly mapped to the rhythmic variations in MIDI notes. In chapter 4, we present StressTree and HeartBloom, two metaphorical visualizations of HRV data. We introduce the images of tree and flower as visual metaphors in the visualization of HRV data. The traditional IBI tachogram and HRV Poincaré plots are transformed into a common flower or tree image that can be understood intuitively. Besides visualizing the IBI dataset, the appearance of the flower-shape or tree-shape visualization also represents a health-related physiological meaning semantically. In chapter 5, we present LivingSurface, an interactive wall-like surface as a shape-changing display of biofeedback. LivingSurface aims at using the qualities of a physical change to enhance the interaction with digital biofeedback information. The expressivity of LivingSurface is explored to embody the interface with a sense of life. In chapter 6, we present Breathe with Touch, a tactile interface that simulates human breathing movements through the inflation and deflation of an airbag. The natural coupling between the user’s breathing behaviour and the interface’s action is designed to facilitate an automatic breathing regulation.
In the third part, we explore using ambient media for biofeedback display. The initial intention is simple: to turn the biofeedback system into ‘invisible background’, where the users can perceive their internal physiological states from an environment without computer screens. We utilize nature sounds and ambient lights for an engaging and comforting biofeedback interaction. In chapter 7 and 8, we design an auditory display with nature sounds. We conduct an experimental study to test the modulation of various nature sounds for calm information display. The study concludes with a simplified three-layer structure and a nature soundscape model, which is used in the design of BioSoundscape in chapter 8. BioSoundscape harnesses nature sounds to create a ‘calm’ nature soundscape that responds to the user’s physiological activities. It can not only serve as an ambient biofeedback display but also be integrated into an indoor acoustic environment as a natural augment. In chapter 9, we present DeLight, an ambient lighting biofeedback system. The intent of DeLight is not only to present biofeedback data but also offer a comforting environmental stimulus for relaxation training. In chapter 10, we integrate the BioSoundscape and DeLight together into a room-scale audio-visual biofeedback system: RESonance. It offers a lightweight solution for immersive biofeedback training. In chapter 11, the developed RESonance biofeedback system was applied in a multi-session biofeedback training with five PhD students and five young soccer players. We evaluate the effectiveness of biofeedback in skill-learning for stress coping, and also investigate the users’ learning curve with biofeedback.
In the last part of the thesis, we formulate our answers to the research questions and conclude our contributions to the research and design biofeedback for stress management. We indicate four future research directions in everyday biofeedback: Inherent Biofeedback, Adaptive Biofeedback, Casual Biofeedback and Peripheral Biofeedback. In conclusion, this thesis presents a design-driven exploration of biofeedback applications for managing stress. The design explorations cover a broad design space including data sonification, metaphorical visualization, shape-changing displays, tangible interactions, and ambient displays. In these explorations, the designs themselves become a resource for new knowledge through empirical research surrounding the evaluation of these works. We hope this work could be a starting point for initiating a new field of ‘Everyday Biofeedback’.