Predicting the psycho-physiological effects of pre-selected music� A feasibility Study (Preprint)

BACKGROUND The use of music therapy in healthcare, including mental health and general wellbeing has become common in many parts of the world, and is usually supported by clinical validation and supervision. In the last decades, research into music therapy and music medicine has advanced rapidly in step with advances in music neuroscience, brain imaging, neuroendocrinology and increasingly easy access to physiological measurements through a new generation of sensors. X-System has developed algorithms predicated upon these developments. These algorithms are intended to model areas and functions of the brain relating to the processing of music, in order to predict the differing neurophysiological effects of different music, and to stream music to achieve desired states of mind and body. X-System has been able to verify these results by both subjective measures and physiological measures including, heart rate and galvanic skin conductance. However, predictions of valence, leading to predictions of mood and emotion, are important for work in mental health but have been difficult to verify physiologically. In this context, Heart Rate Variability (HRV) has proved to be a very promising way forward. OBJECTIVE Objectives: This study had two main objectives 1) Understanding the impact of different types of music on heart rate variability as a means of understanding the effects on the subject, and 2) Exploring the potentials of the X-System algorithm to predict psycho-physiological effects of music by comparing the output data to HRV data, and where possible offering validation of existing algorithms METHODS Six participants listened to four such pre-selected pieces (two “happy” and two “sad”, according to internet-sourced popular curations) and one self-selected song in a random order whilst ECG data was recorded continuously throughout. After each song, participants were asked to fill a valence questionnaire. The pre-selected songs were analysed by X-system software to predict the effects of songs on the subjects, and the results were compared to the corresponding HRV data. RESULTS Results showed that song 4 (happy song), had the largest significant impact on HRV (P=.05), in particular the total power domain where an average increase of 574 ms2 was seen. Promising potential correlations were identified between X-System predictions for both arousal and valence on the one hand and subjective responses of participants on the other. Both arousal and valence, X-System predictions and subjective rankings correlated with HR and HRV rankings for three out of four songs. In tThere were some tentative, identifiable correlations between the X-System predictions, subjective responses and HRV data that suggest that further work should be undertaken, directed towards improving X-System mood predictions and developing treatments for autonomic and mood disorders based on music streaming. CONCLUSIONS There were some tentative, identifiable correlations between the X-System predictions, subjective responses and HRV data that suggest that further work should be undertaken, directed towards improving X-System mood predictions and developing treatments for autonomic and mood disorders based on music streaming.