HEART.FM is a mobile app development initiative to facilitate tailored music-based interventions with physiological feedback to lower cardiovascular disease risk. Cardiovascular disease is the world’s leading cause of death and hypertension is its foremost risk factor. Although hypertensive drugs are effective, they have unpleasant side effects, leading to poor drug compliance. Music offers non-pharmacological, non-invasive, and pleasurable ways to alter heart rate and blood pressure with few, if any, side effects. HEART.FM uses portable/wearable technologies to provide physiological feedback to guide music interventions. App-guided music intervention has the advantage of scalability and can be readily deployed to large populations. Real-time data acquisition enables the app to adapt quickly to changing situations.
The HEART.FM mobile app registers and synchronises physiological and music data during music listening in order to personalise music selection to an individual listener’s physiological response in that moment. It not only records physiological data whilst users listen to pre-recorded music, it can also record physiological signals during live music. The mobile app is designed to allow users to see their own electrocardiographic (ECG) data captured from a wearable sensor and selected heart rate variability information during music listening. For research observation and for science communication, a desktop visualisation extension allows researchers, clinicians, and the general public to see connections within more complex combinations of physiological data and music information streams.
Our scientific approach is based on linking music expressivity – the acoustic variations introduced during music communication – to physiological response. Using the mobile app, we have created a sizable, ecologically valid music-physiology dataset – participants listen to whole pieces of music over a sustained period of time – to enable the building of mathematical models for individual and group cardiac responses to music. We analyse data not only from music listeners, but also from players, to better understand the heartbeat dynamics internally amongst ensemble musicians, and end-to-end from players to listeners and vice versa. Leveraging mathematical representations of music structures and expressivity, we are formalising knowledge about the mechanisms of cardiac response to music. This knowledge will allow us to more effectively use music in digital therapeutics to prevent arrhythmias and optimise blood pressure and reduce heart disease burden on society.
