Spatiotemporal neural network dynamics for the processing of dynamic facial expressions

Area of Science:

• Neuroscience
• Cognitive Science
• Psychology

Background:

• Facial expressions of emotion trigger complex psychological responses.
• The precise timing and brain network interactions during this process are not well understood.

Purpose of the Study:

• To investigate the temporal dynamics and brain activity patterns during the observation of dynamic facial expressions.
• To elucidate the neural mechanisms underlying the rapid processing of emotional facial cues.

Main Methods:

• Magnetoencephalography (MEG) was used to record brain activity in participants passively viewing dynamic facial expressions (fear, happiness) and control stimuli (mosaics).
• Source-reconstruction analysis with functional magnetic resonance imaging (fMRI) data and dynamic causal modeling (DCM) were employed.

Main Results:

• Dynamic facial expressions elicited greater activation in bilateral occipital and temporal cortices compared to mosaics between 150-200 ms and later.
• The right inferior frontal gyrus showed increased activity for dynamic faces versus mosaics at 300-350 ms.
• Dynamic causal modeling indicated activation of dual visual and visual-motor routes, with feedforward and feedback connections showing superior influence before and after 200 ms, respectively.

Conclusions:

• Hierarchical, bidirectional neural network dynamics operate within milliseconds to process dynamic facial expressions.
• These findings provide insights into the rapid neural computations underlying social-emotional perception.