Bayesian models of mentalizing.

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Psychiatry

Background:

• Mentalizing, the ability to infer others' mental states, is crucial for social interaction.
• Facial emotion perception relies on anticipating dynamic emotional expressions.
• Electroencephalography (EEG) offers insights into the neural dynamics of mentalizing.

Purpose of the Study:

• To model the neural dynamics of facial emotion perception using Bayesian probabilistic models.
• To investigate how Bayesian models can dynamically update beliefs based on incoming emotional information.
• To assess the efficacy of Bayesian models in predicting neural activity during the appraisal of facial expressions.

Main Methods:

• Utilized dynamic EEG recordings to capture neural activity.
• Applied Bayesian probabilistic models to analyze neural data from neutral to emotional facial expressions.
• Compared Bayesian model predictions against actual neural dynamics, including event-related potentials (ERPs).

Main Results:

• A reproducible model of neural dynamics in facial expression appraisal was derived from grand mean ERPs across subjects.
• One Bayesian model accurately predicted individual subject dynamics for four out of five participants.
• Demonstrated the utility of probabilistic modeling in capturing dynamic belief updating during emotional perception.

Conclusions:

• Bayesian probabilistic models offer a powerful framework for modeling the neural dynamics of mentalizing and emotion perception.
• These models provide a more dynamic approach compared to conventional methods for understanding belief updating.
• Future research can leverage Bayesian formalism for more detailed, single-trial level modeling of neural dynamics.