Moving beyond average interbrain synchrony: dynamic two-brain states during collaboration

A recent study by Naudszus, Moffat, and Cross introduces a state-based analysis framework that captures these temporal changes by jointly analysing within-brain and between-brain functional connectivity during collaborative drawing. Instead of focusing solely on average synchrony, the proposed approach identifies recurring connectivity patterns that evolve throughout the interaction, providing a more detailed view of collaborative neural dynamics.

Study Design and Methodology

The study involved 122 participants forming 61 dyads: 30 same-generation dyads consisting of two younger adults and 31 intergenerational dyads consisting of one younger and one older adult. All participants were previously unacquainted and completed six weekly sessions including one individual and two collaborative drawing tasks. Brain activity was recorded simultaneously using two Cortivision Photon Cap fNIRS systems positioned over the bilateral inferior frontal gyrus (IFG) and temporoparietal junction (TPJ), enabling synchronized hyperscanning recordings throughout the experiment.

To characterise neural interactions over time, the researchers analysed dynamic functional connectivity using overlapping 15-second sliding windows rather than averaging connectivity across the entire task. Kernel functions were used to estimate similarity between neural signals, while HbO and HbR connectivity matrices were combined into block-diagonal matrices and clustered using Riemannian geometry-based k-means. Following hyperparameter optimisation, each time window was assigned to one of the identified two-brain states, allowing the researchers to quantify their occurrence, duration and coverage throughout the interaction.

The clustering pipeline consisted of five key components
(Source: https://www.sciencedirect.com/science/article/pii/S000169182601053X?via%3Dihub#f0005)

Seven Recurring Two-Brain States

The analysis identified seven recurring connectivity states representing different configurations of within-brain and between-brain functional connectivity. Three of these states showed significant differences between real interacting dyads and pseudo dyads, indicating that they reflected genuine interaction-related neural dynamics rather than coincidental similarity between independently recorded participants.

Among these, State G showed the clearest relationship with collaborative interaction. During collaborative drawing, intergenerational dyads remained in this state significantly longer than same-generation dyads. The state was characterised by relatively low-to-medium between-brain connectivity together with prominent within-brain connectivity between the bilateral IFG of one member of the dyad. The authors suggest that this asymmetric connectivity pattern may be consistent with leader-follower dynamics, while emphasising that additional behavioural measures would be needed to directly verify this interpretation.

Relationship with Collaborative Behaviour

Behavioural analyses showed that intergenerational dyads engaged in more overt turn-taking than same-generation dyads, while longer durations of State G were associated with less turn-taking, corresponding to more simultaneous drawing. Neither turn-taking behaviour nor the occurrence, duration or coverage of any two-brain state changed significantly across the six-week study.

Conclusion

The proposed state-based framework provides a way to analyse how within-brain and between-brain functional connectivity evolve during social interaction. By identifying recurring two-brain states, the approach captures temporal changes that are not accessible through conventional average connectivity measures. According to the authors, this methodology may support future fNIRS hyperscanning studies investigating collaborative behaviour in more naturalistic settings.

Source

Based on: Naudszus, M., Moffat, D. C., & Cross, E. S. (2026). Two-brain states characterize within- and between-brain connectivity during intergenerational collaborative drawing. Acta Psychologica, 252, 107252:

Learn more about Photon Cap in hyperscanning research

The study utilized fNIRS technology to capture brain activity during naturalistic interaction. If you’re interested in how Photon Cap supports this type of research, check out our previous articles:

Cardiac synchrony in same-generation and intergenerational dyads: a longitudinal fNIRS study
or Hyperscanning with the Photon Cap: The Role of Neural Synchrony During Social Interactionu