Perceived ambiguity of social interactions increases coupling between frontal and temporal nodes of the social brain

Matthew Ainsworth, Jérôme Sallet, Olivier Joly, Diana Kyriazis, Nikolaus Kriegeskorte, John Duncan, Urs Schüffelgen, Matthew F. S. Rushworth, Andrew H. BellbioRxiv 2020 · DOI

TL;DR

This study reveals that the brain's "social network" dynamically adjusts its connectivity based on the perceived ambiguity of social interactions. When faced with unclear social cues, specific frontal and temporal brain regions, particularly the cingulate gyrus, show increased coordination, suggesting a heightened need for cognitive processing to interpret these situations.

What Problem Does This Paper Address?

While we know that a network of brain regions is involved in social behavior, inferring intentions, and controlling social interactions, how these regions interact dynamically during complex social scenarios remains largely unknown. Specifically, how the brain's social network connectivity changes when processing different types of social interactions – from clear-cut aggressive or affiliative encounters to more ambiguous ones – is not well understood. This paper aims to bridge this gap by investigating how functional connectivity within this social brain network is modulated by the context and ambiguity of observed social interactions.

Methods

The researchers employed functional Magnetic Resonance Imaging (fMRI) to record brain activity in three macaque monkeys while they viewed video clips of social interactions. They used a progressive four-stage analysis approach to examine dynamic connectivity:

  1. Global Connectivity: Assessed overall connectivity within a defined "putative social network" (comprising 8 bilateral regions including temporal, parietal, frontal, and subcortical areas).
  2. Non-Social Stimuli Response: Examined how connectivity changed in response to non-social visual stimuli.
  3. Social Behavior Response: Investigated changes in connectivity in response to different social behaviors (aggressive, affiliative, ambiguous).
  4. Specific Behavior Analysis: Focused on identifying which specific behaviors elicited the most significant changes in pairwise connections.

Key techniques included:

  • Phase Synchrony: A measure of signal coherence between brain regions, used to assess dynamic connectivity. This method is less sensitive to signal amplitude variations compared to correlation-based measures.
  • General Linear Model (GLM): Used for initial whole-brain analysis to identify brain regions activated by viewing social stimuli and to create regressors for video features.
  • Repeated Measures ANOVA: Employed to assess how network connectivity was modulated by the type of social interaction.
  • Degree Centrality and Eigenvector Centrality: Network metrics used to quantify the importance and interconnectedness of individual brain regions within the network.

Datasets

The study used video clips of social interactions between macaques (conspecifics and Bonnet macaques). These clips were sourced from nature documentaries and a local breeding center, featuring one or two+ actors engaged in various social behaviors. The videos were categorized as aggressive, affiliative, or ambiguous. No new benchmarks were introduced; existing fMRI data collection and analysis pipelines were utilized.

Key Insights and Ideas

  • Ambiguity is Key: The most striking finding is that ambiguous social interactions, rather than clear aggressive or affiliative ones, significantly increase connectivity within the social brain network. This suggests that the brain expends more effort to interpret uncertain social situations.
  • Frontal-Temporal Coordination: The study highlights the crucial role of fronto-temporal connectivity, which is significantly modulated by social context. This suggests a dynamic interplay between regions involved in perception and higher-level cognitive processing.
  • Cingulate Gyrus as a Hub for Uncertainty: The increased connectivity involving the cingulate gyrus during ambiguous interactions points to its potential role in processing uncertainty and coordinating cognitive resources for social inference.
  • Dynamic Nature of Social Cognition: The research underscores that social brain networks are not static but actively reconfigure their connections based on the nature of the social stimuli.

Key Results

  • Viewing social interactions activated several brain regions, including the superior temporal sulcus (STS), parietal cortex, medial and lateral prefrontal cortex (PFC), and caudate nucleus.
  • Frontal areas were primarily recruited when monkeys viewed scenes with multiple actors.
  • Viewing aggressive or affiliative behavior led to a modest increase in temporo-temporal and premotor-temporal connectivity, respectively.
  • Viewing ambiguous interactions resulted in a pronounced increase in connectivity within the cingulate-cingulate, temporo-temporal, and cingulate-temporal pathways.
  • The cingulate gyrus and premotor cortex showed higher degree and centrality scores, indicating their importance in socially modulated connections.

Limitations and Future Work

  • Passive Viewing: The study involved passive viewing of social interactions, meaning the monkeys were not required to make decisions or predictions. This limits the ability to infer what specific information is being processed between frontal and temporal regions.
  • Technical Limitations of fMRI: fMRI has limited temporal resolution, making it difficult to reveal the precise nature of information transfer between brain regions on a millisecond-by-millisecond basis.

Future work suggested by the authors includes:

  • Behavioral Components: Incorporating behavioral tasks where animals must make judgments or predictions about ambiguous social situations to better understand the cognitive demands.
  • Advanced Techniques: Utilizing methods with higher temporal resolution (e.g., electrophysiology) to investigate how information is communicated between nodes in the social network and how neural representations are updated in real-time.
  • Interactions with Other Brain Areas: Further exploring the role of striatal and caudal interactions in social cognition.

Summary updated

Abstract

ABSTRACT Social behaviour is coordinated by a network of brain regions, including those involved in the perception of social stimuli and those involved in complex functions like inferring perceptual and mental states and controlling social interactions. The properties and function of many of these regions in isolation is relatively well understood but little is known about how these regions interact whilst processing dynamic social interactions. To investigate whether social network connectivity is modulated by social context we collected fMRI data from monkeys viewing “affiliative”, “aggressive”, or “ambiguous” social interactions. We show activation relating to the perception of social interactions along both banks of the superior temporal sulcus, parietal, medial and lateral PFC and caudate nucleus. Within this network we demonstrate that fronto-temporal connectivity are significantly modulated by social context. Crucially, we link the observation of specific behaviours to changes in connectivity within our network. Viewing aggressive or affiliative behaviour was associated with a limited increase in temporo-temporal and premotor-temporal connectivity respectively. By contrast, viewing ambiguous interactions was associated with a pronounced increase in cingulate-cingulate, temporo-temporal, and cingulate-temporal connectivity. We hypothesise that this widespread network synchronisation occurs when cingulate and temporal areas coordinate their activity when more difficult social inferences are made.
BibTeX
@misc{ainsworth2020perceived,
  author = {Matthew Ainsworth and Jérôme Sallet and Olivier Joly and Diana Kyriazis and Nikolaus Kriegeskorte and John Duncan and Urs Schüffelgen and Matthew F. S. Rushworth and Andrew H. Bell},
  title = {{Perceived ambiguity of social interactions increases coupling between frontal and temporal nodes of the social brain}},
  year = {2020},
  doi = {10.1101/2020.03.11.987792},
  howpublished = {bioRxiv},
}
Figure