Stereotyped goal-directed manifold dynamics in the insular cortex.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Machine Learning

Background:

• The insular cortex plays a crucial role in homeostasis, emotions, and cognition.
• Understanding insular cortex population-level information processing remains a challenge.

Purpose of the Study:

• To investigate insular cortex population activity patterns during goal-directed behaviors using unsupervised machine learning.
• To characterize the neuronal manifold of the insular cortex and its dynamics.

Main Methods:

• Applied unsupervised machine learning techniques to analyze neuronal population activity in mice.
• Recorded insular cortex activity during goal-directed behaviors and self-paced consumption.
• Compared activity patterns across different motivational states and behavioral contexts.

Main Results:

• The insular cortex activity manifold demonstrated remarkable consistency across animals and motivational states.
• Highly stereotyped activity dynamics were observed during rewarded trials, allowing robust prediction of outcomes.
• Stereotyped patterns were identified as task-dependent reward anticipation, distinct from sensory or valence-related activity.

Conclusions:

• Insular cortex activity exhibits a consistent neuronal manifold supporting goal-directed reward anticipation.
• These findings reveal a core computational principle in the insular cortex.
• This understanding may elucidate the insular cortex's role in pathologies characterized by aberrant motivations.