DBIC Special Lecture: April 26, 2024
Ray Lee
Associate Professor of Radiology, Department of Radiology, University of Texas Health Science Center at San Antonio
Quantitative brain-to-brain interactions in eye gaze between parents and children: their brain networks, communication modes, and information contents
Abstract:
Social interaction studies have been going through some paradigm shift, from speculator approach to the second-person or interactor thesis to the communication algorithm. Although all of them capture significant aspects of social interaction, the essence of the interaction, which cannot be reduced to two independent agents’ behaviors, has not been translated to experimental design and data analysis. Consequentially, quantifying social interaction has been artificially convoluted or even arbitrary sometimes. In this study, a simple social interaction, eye gazing, is conceptually modeled, experimentally measured, and analytically computed based on the irreducible interaction thesis (IIT) – the interaction determines dyadic brain states; not dyadic brain states determine the interaction – during eye gazing. To test the power and scope of the proposed IIT-based approach, the eye gazing between parents (P) and children (C) was chosen to quantify the interaction between the developed and developmental brains.
A feature-space conceptual model of the parent-child gazing was substantiated with the spatiotemporal dynamic data acquisitions from the fMRI that scans dyads either within one scanner or in two separated scanners. By analytically distinguishing between “seeing eyes” and “seeing face without eyes”, a data-driven approach identified several communication modes in eye-to-eye interaction between P and C. The most dominant mode is the dyadic visual paths which is made of occipital pole, occipital fusiform, and lingual gyrus in both P and C, as well as precuneus, angular gyrus, and cingulate in P only. This mode contributes 90.8% of the explained variances and has bidirectional mutual information 1.55±0.4bit from C-->P and 0.85±0.52bit from P-->C. Another interesting mode suggests developmental brains’ external regulation mechanism which consists of most parietal components for P and most frontal components for C. This mode contributes 1.84% of the explained variances and has bidirectional mutual information 0.76±0.34bit C-->P and 1.66±0.33bit P-->C. It reveals dynamics that eye-to-eye interaction between P and C suppresses C’s own reflective processes and relies P for reflection.
In this interactor-based study, some dyadic social behaviors, such as developmental brains’ attachment and external regulation, may naturally emerge from the experimental data as the mathematical mode, which presents an experimental approach for quantitatively measuring social interaction.