Knowledge Bank

Mental stress has known detrimental effects on human health, however few algorithmic methods of reducing mental stress have been widely explored. While the act of listening to music has been shown to have beneficial effects for stress reduction, and furthermore, audio players have been designed to selectively choose music and other inputs with the intent of stress reduction, limited work has been conducted for real-time stress reduction with feedback control using physiological input signals such as heart rate or Heart Rate Variability (HRV). This thesis proposes a feedback controller that uses HRV signals from wearable sensors to perform real-time (< 1 second) modulations to music through tempo changes with the goal to regulate and reduce stress levels. A standardized, stress inducing test based on the popular Stroop test is also introduced, which has been shown to induce acute stress in subjects and can be used as a testing benchmark for controller design. Ultimately, a controller is presented that when used is not only able to maintain stress levels during stress-inducing inputs to a human but even provides de-stressing effects beyond baseline performance.