Characterization of Hyperbrain Networks During Joint Piano Playing Using a Complex Dynamics Framework

Abstract

Social interaction is essential for human life, but we have little understanding of the neural mechanisms supporting it. Recent research has shown correlated activity between the brains of individuals (Goldstein et al. 2018; Müller et al. 2018; Dikker et al. 2017; Toppi et al. 2016) using the novel technique of Electroencephalography Hyperscanning, which allows us to record multiple persons’ electrical brain activity at the same time. Interpretation of this data, however, is still unclear: does common activity reflect social interaction or is it just a by-product of shared perception? Furthermore, there is no unifying framework on how to analyze these novel data. Although we did not find evidence for synchronous brain activity between pianists playing duets using a complex dynamics framework, we were able to differentiate music pieces with ambiguous leadership roles from those with clear leadership roles using multivariate statistical approaches (graph theory). Furthermore, ambiguous leadership network characteristics correlated with participants’ perceptions of the quality of their performances. This thesis also contributes to this field by expanding previously proposed frameworks (Duan et al. 2015) to include a complex dynamics approach and thoroughly discussing issues in hyperbrain analysis. By standardizing the protocols, interpretations, and data analysis approaches of data from EEG hyperscanning, we can better elucidate what this synchrony means, effectively helping us move the field of single-person social neuroscience towards a two person neuroscience (Dumas 2011; Schilbach et al. 2013). This has profound implications at several levels, including the quantification of high level social constructs, such as empathy or joint attention, to clinical research, where these statistics can be used as diagnosis tools for the socially impaired brain.