Multivariate model for cooperation: bridging social physiological compliance and hyperscanning

Abstract

The neurophysiological analysis of cooperation has evolved over the past 20 years, moving towards the research of common patterns in neurophysiological signals of people interacting. Social Physiological Compliance (SPC) and Hyperscanning represent two frameworks for the joint analysis of autonomic and brain signals respectively. Each of the two approaches allows to know about a single layer of cooperation according to the nature of these signals: SPC provides information mainly related to emotions, and Hyperscanning that related to cognitive aspects. In this work, after the analysis of the state of the art of SPC and Hyperscanning, we explored the possibility to unify the two approaches creating a complete neurophysiological model for cooperation considering both affective and cognitive mechanisms. We synchronously recorded electrodermal activity, cardiac and brain signals of 14 cooperative dyads. Time series from these signals were extracted and Multivariate Granger Causality was computed. The results showed that only when subjects in a dyad cooperate there is a statistically significant causality between the multivariate variables representing each subject. Moreover, the entity of this statistical relationship correlates with the dyad’s performance. Finally, given the novelty of this approach and its exploratory nature, we provided its strengths and limitations.

Estimation of Granger causality through Artificial Neural Networks: applications to physiological systems and chaotic electronic oscillators

“…Then, for each window, the PSD was integrated within four different frequency bands to obtain time series representative of the δ (0.5–3 Hz), θ (3–8 Hz), α (8–12 Hz) and β (12–25 Hz) brain wave amplitudes. The use of these frequency bands was motivated by studies which relate increasing levels of fatigue or alertness with higher PSD of the δ , θ and α processes and lower PSD of the β process ( Sciaraffa et al, 2020 ; Tran et al, 2007 ; Trejo et al, 2007 ). The brain time series extracted in this way was synchronous with those obtained resampling at 1 Hz the three cardiovascular time series using spline interpolation ( Zanetti et al, 2019 ).…”

Being ‘in sync’—is interactional synchrony the key to understanding the social brain?

“…It seems instead that this merely facilitates attention or positive engagement with a social or non-social stimulus. Indeed, the wish to cooperate with ( Reinero et al , this issue ; Sciaraffa et al , this issue ), to deceive or to detect deception in an unknown partner ( Pinti et al , this issue ) also prompts interactional synchronizing. Moreover, the mere exposure to regularly repeating sounds amplifies relevant sound frequencies in the brain and benefits the cyclic allocation of attention irrespective of modality ( Schirmer et al , this issue ).…”

“…Some of these approaches, featuring in the current issue, include simple cross-correlations of original or wavelet-transformed time series ( Kruppa et al , this issue ; Pan et al , this issue ) as well as indices of phase locking ( Heggli et al , this issue ) or shared changes in the power of certain frequencies characterizing the time series ( Schirmer et al , this issue ; Zamm et al , this issue ). These as well as other analytical strategies including Granger causality ( Sciaraffa et al , this issue ) differently address the problem of how to define synchrony and whether and how delays between corresponding changes in the time series of interest (e.g. leads/lags) should be considered and mathematically modeled.…”

“…Towards this end, the present collection of papers makes an initial effort by providing insightful reviews of study approaches and findings ( Hoehl et al , this issue ; Jiang et al , this issue ; Levy et al , this issue ; Misaki et al , this issue ), empirically linking different synchrony measures ( Pan et al , this issue ; Sciaraffa et al , this issue ) and by introducing a new python-based software package that can help with standardizing data analysis ( Ayrolles et al , this issue ). However, more work is needed.…”

Technologically-Assisted Communication Attenuates Inter-Brain Synchrony

“…These include; (a) homolog connectivity of same-area-same-hemisphere, such as temporal-to-temporal (Djalovski et al, 2021; Kinreich et al, 2017), central-to-central (Djalovski et al, 2021), and frontal-to-frontal connectivity (Azhari et al, 2019; Cui et al, 2012; Kruppa et al, 2021; Pan et al, 2017; Reindl et al, 2018; Wang et al, 2020); (b) cross-hemisphere same-region linkage, such as left temporal to right temporal connectivity; and (c) non-homolog multi-region linkage of same or different hemisphere, such as frontal-to-temporal (Pérez et al, 2017), frontal-to-parietal (Piva et al, 2017), central-to-temporal (Endevelt-shapira et al, 2021; Pérez et al, 2017), central-to-parieto-occipital and centro-parietal and parieto-occipital connectivity (Dumas et al, 2010). Notably, most studies reported right-hemisphere connectivity of homolog or non-homolog regions (Cui et al, 2012; Dumas et al, 2010; Endevelt-shapira et al, 2021; Jahng et al, 2017; Noah et al, 2020; Pan et al, 2017; Sciaraffa et al, 2021), suggesting that the right hemisphere, which matures early (Geschwind and Galaburda, 1985) and is implicated in non-verbal affective processing (Borod et al, 1998), may be particularly sensitive to two-brain communication. Such multiple areas of neural linkage underscore the richness of cross-brain possibilities afforded by naturalistic co-present interactions that may reflect distinct mechanisms triggered by different social goals.…”