Machine Learning and Physiological Sensing in Health Psychology
Stress can cause major health problems and is a highly prevalent symptom in our post-industrialized, always-online society. In particular it is known as a cause for many negative long-term physical effects on the body like heart diseases, obesity, and depression. The high prevalence is particularly threatening as most people experience high loads of stress in their work environment and do not manage to perform proper post-work relaxation, either because they are not even aware of the symptoms or because they have no motivation. Therefore, it is important to understand the underlying mechanisms and identify possibilities/tools to cope with it and prevent negative health outcomes.
Recent developments in wearable technology have led to unobtrusive, sensor-packed, always-on devices that are able to continuously monitor biosignals to possibly detect, determine or even prevent stress or some of its negative outcomes. Additionally, the use of algorithms and machine learning techniques enable new possibilities for the analysis of physiological signals and psychological data. Thus, it might provide new insights into the underlying physiology of stress and identify possibilities/tools to cope with it and prevent negative health outcomes.
For this reason, this project explores different applications of machine learning, wearable computing and physiological computing to various topics in health psychology, such as the analysis of biological markers such as cortisol and amylase, the development of novel VR-based stress induction methods, or the exploration of different stress reduction techniques using biofeedback and VR.