Neural dynamics encoding risky choices during deliberation reveal separate choice subspaces.

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Human decision-making integrates information from multiple brain areas.
• Decisions often occur under uncertainty, necessitating deliberation.
• Neural dynamics of deliberation remain largely unknown.

Purpose of the Study:

• To investigate the neural dynamics underlying human deliberation during decision-making.
• To identify latent brain states associated with choice under uncertainty.
• To explore neural representations of evidence accumulation.

Main Methods:

• Utilized intracranial electrophysiological recordings from hundreds of brain locations during behavior.
• Applied dynamic systems modeling to analyze high temporal resolution neural data.
• Examined neural dynamics during both risky and non-risky decision-making tasks.

Main Results:

• Neural dynamics for risky decisions converged to choice-specific subspaces.
• Subspace overlap decreased as the decision approached, suggesting evidence accumulation.
• These dynamics were specific to risky decisions, not those without risk.
• Identified a link between latent state models and regression analyses.

Conclusions:

• Deliberation and evidence accumulation in risky decisions are represented by shared neural dynamics.
• Findings provide novel insights into the neural computations governing human choice.
• Suggests a network-level representation of evidence accumulation during decision-making.