Distinct neural temporal architectures encode rapid social expressions and sustained internal mood states

Area of Science:

• Neuroscience
• Computational Psychiatry
• Affective Science

Background:

• Affective processing spans rapid facial signals and sustained moods, but underlying neural computations are unclear.
• Understanding temporal brain architectures for emotions is key for psychiatric biomarkers.
• Individual differences in affective processing remain poorly understood.

Purpose of the Study:

• Investigate distinct neural computations for facial expressions and mood states.
• Determine if facial expressions reliably indicate mood across individuals.
• Explore individual variability in neural mood encoding.

Main Methods:

• Multimodal approach: automated facial expression analysis and intracranial EEG.
• Data collected over multiple days from 16 epilepsy patients (2,037 contacts).
• Analyzed 1,396 smiles, 3,746 neutral expressions, and 336 mood assessments.

Main Results:

• Aperiodic activity in lateral temporal cortex accurately decoded facial expressions across participants.
• Facial expressions did not consistently predict mood states.
• Limbic low-gamma power encoded mood in a subset of individuals with expression-mood correlations.
• Neural features for expression and mood decoding were computationally independent.

Conclusions:

• Distinct neural mechanisms underlie facial expression and mood processing.
• Individual differences exist in how neural activity represents mood.
• Findings offer a framework for understanding neural mood representation and objective affective measurement.