Hyperscanning fNIRS data analysis using multiregression dynamic models: an illustration in a violin duo

Area of Science:

• Social Neuroscience
• Neuroimaging
• Computational Neuroscience

Background:

• Interpersonal neural synchronization (INS) is crucial for understanding brain-to-brain influence.
• Existing methods for analyzing brain connectivity are insufficient for complex intersubject interactions.
• Novel statistical inference methods are needed to study INS.

Purpose of the Study:

• To propose a novel two-step network estimation method for analyzing fNIRS hyperscanning data.
• To investigate interpersonal neural synchronization in a real-world scenario.
• To develop tools for understanding brain-to-brain associations.

Main Methods:

• Utilized a two-step network estimation approach combining Tabu search local method and global maximization.
• Employed partial conditional directed acyclic graph (DAG) and multiregression dynamic models.
• Applied the method to fNIRS hyperscanning data from two individuals playing the violin together.

Main Results:

• The proposed approach successfully estimated probabilistic network representations of brain activity.
• Time-varying parameters provided insights into brain-to-brain associations between the violinists.
• Confirmed a leading-following dynamic between the players based on ROI causal relations.

Conclusions:

• The study introduces new analytical tools for the field of social neuroscience.
• The developed methods offer new perspectives on intersubject interactions and neural synchronization.
• The findings highlight the dynamic and evolving nature of brain influence between individuals.