The encoding of alternatives in multiple-choice decision making.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Decision Neuroscience

Background:

• Research has advanced understanding of binary decision-making neural mechanisms.
• Current research is shifting towards decision-making processes involving multiple alternatives.

Purpose of the Study:

• To investigate how the brain handles and encodes decision-making with multiple alternatives.
• To present a biophysically realistic spiking neuron model for multi-alternative decision tasks.

Main Methods:

• Developed a minimal, biophysically realistic spiking neuron model.
• Tested the model on a random-dot motion-discrimination task with varying numbers of alternatives (2 and 4).
• Ensured network parameters and inputs were independent of the number of choice alternatives.

Main Results:

• The model successfully accounted for experimental data on both cellular and behavioral levels.
• Network performance was independent of the number of choice alternatives without top-down regulation.
• Increased neuronal representation per choice positively correlated with choice-number-independent decision-making capacity.

Conclusions:

• A pooled, multineuron representation of choice alternatives offers a physiological advantage.
• The proposed model provides a framework for understanding neural computations in multi-alternative decision-making.