Learning the neurobiology of social behavior from data: Four networks underlying social cognition

4203 

Abstract Submission 

Wednesday, June 28 & Thursday, June 29 

Daniel Alcalá-López¹, Jonathan Smallwood², Elizabeth Jefferies², Frank Van Overwalle³, Kai Vogeley⁴, Rogier Mars⁵, Angie Laird⁶, Peter Fox⁷, Simon Eickhoff⁸, Danilo Bzdok⁹

¹RWTH, Aachen, Deutschland, ²The University of York, York, United Kingdom, ³Department of Psychology, Vrije Universiteit Brussel,, Belgium, Brussels, Belgium, ⁴Institute of Neuroscience and Medicine (INM-3), Research Center Jülich, Jülich, Germany, ⁵Donders Institute, Nijmegen, Netherlands, ⁶Florida International University, Miami, FL, ⁷Research Imaging Institute, UTHSCSA, San Antonio, TX, ⁸Research Center Jülich, INM-1, Jülich, Germany, ⁹RWTH Aachen University, Aachen, Germany

Daniel Alcalá-López    -  Lecture Information | Contact Me
RWTH
Aachen, Deutschland

Complex social interactions probably emerge from different yet interacting neural systems. However, the field of social neuroscience has been fragmented into highly specialized, rarely cross-referenced topics ranging from basic face recognition to abstract mentalizing about others' intentions. The present study hence provides a first systematic reconciliation by computing a data-driven brain atlas of social cognitive capacities.

We comprehensively summarized previously published quantitative meta-analyses on social-affective phenomena to provide a formal, observer-independent definition of the social brain topography (Fig. 1A). The selected coordinate-based meta-analyses involved more than 20.000 participants across almost 4.000 imaging studies. The 36 generated regions (Fig. 1B) served as seed regions in subsequent analyses to identify commonalities and differences in brain connectivity. The social brain was thus characterized by i) meta-analytic connectivity modeling (MACM), evaluating task-dependent coactivation, ii) resting-state physiological fluctuations (RSFC), evaluating correlations in task-free brain states, and iii) functional association profiling with BrainMap (Fig. 1C).

First, network clustering of each region's connectivity maps across MACM and RSFC (Fig. 2A) allowed the seeds to be grouped based on how similar their connectional profiles were. This yielded a functional segregation of the social brain into four main subsystems (Fig. 2B): i) a visual-sensory sub-network including the bilateral fusiform gyri, posterior part of the superior temporal sulci, and middle temporal visual areas V5; ii) a limbic sub-network consisting on the bilateral amygdala, hippocampus, and nucleus accumbens as well as the rostral portion of the anterior cingulate and ventromedial prefrontal cortices; iii) a intermediate-level sub-network that was composed of the anterior part of the mid-cingulate cortex and bilateral anterior insulae, often described as the "saliency network", as well as the bilateral inferior frontal gyri, supramarginal gyri, and supplementary motor areas, the suspected mirror-neuron system; and iv) a higher-level sub-network including the dorsomedial prefrontal cortex, medial frontal pole, posterior cingulate cortex, and precuneus, as well as the bilateral temporo-parietal junction areas, middle temporal gyri, and temporal poles, known as "default-mode network".
Second, we report a global trend for left-lateralization in the social brain seeds during tasks, but not at rest. These functional asymmetries converged to connectivity targets along the surface of the frontal and temporal lobes such as the inferior frontal gyrus, especially when seeding from the DMN-related regions.
Third, we observed pronounced overlapping MACM results between seeds in the putative mirror-neuron system, believed to be involved in embodied simulation of others' purposeful body movements (i.e., action observation and imitation), and those in the saliency network, believed to be involved in embodied simulation of others' affective states (i.e., empathy). These connectional configurations differed greatly from the connectivity profile of the default-mode seeds, believed to be involved in abstract emulation of others' mind states (i.e., theory of mind).

We determined a comprehensive social brain atlas and characterized its connectional profile across many tasks (MACM) and at rest (RSFC). We found networked configurations at different levels of the neural processing hierarchy, from sensory lower-level to abstract, higher-level systems. As the main conclusion, no brain region or network was exclusively devoted to social processes. Finally, nodes of the putative mirror-neuron system were extensively cross-connected with embodied simulation systems (cf. empathy) rather than abstract emulation systems (cf. theory of mind). These first steps will help reintegrate the currently isolated research agendas in the social and affective neurosciences.

Databasing and Data Sharing ²

Social Cognition ¹

Limbic Systems

Meta- Analysis

Social Interactions