Music in behavioral health: from digital biomarker to therapeutics

Music is a fascinating medium which is closely related to our emotions, personality, and even mental health. What music we choose to listen to can not just be indicative on how we feel right now, but it can even help us regulate our emotional states. In this paper, we argue that with the increased pervasiveness of digital music services there is potential for these novel data traces to aid the development of personalized mental health biomarkers for monitoring, diagnosis, and prognosis. These biomarkers can aid the care team to adjust treatment or intervene. Further, we argue that these personalized models can be the basis for introducing music recommendations with the aim to improve the mental health states of the user in times of need -- the use of music as a therapeutic tool.