Related Experiment Videos
Probabilistic decision making by slow reverberation in cortical circuits
1Volen Center for Complex Systems, Brandeis University, Waltham, MA 02254, USA. xjwang@brandeis.edu
Neuron
|December 7, 2002
Summary
This study models perceptual decision-making in the brain, revealing how recurrent excitation and feedback inhibition create attractor dynamics for visual choices. NMDA receptor-mediated excitation may enable slow sensory integration and categorical decision formation in cortical networks.
Area of Science:
- Neuroscience
- Computational Neuroscience
- Systems Neuroscience
Background:
- Physiological studies have identified neural correlates of perceptual decision-making in primate cortex.
- Understanding the synaptic mechanisms underlying these neural processes is crucial.
Purpose of the Study:
- To elucidate the synaptic mechanisms of decision-making using a biophysically realistic cortical network model.
- To investigate how neural network dynamics contribute to perceptual choices in a visual discrimination task.
Main Methods:
- Developed a biophysically realistic cortical network model for a visual discrimination experiment.
- Simulated attractor dynamics driven by slow recurrent excitation and feedback inhibition.
- Compared model outputs to physiological data, psychometric functions, and reaction times.
Main Results:
- The model successfully replicated decision-correlated neural activity observed in primates.
- Attractor dynamics amplified input differences, leading to binary choices.
- The model accounted for the animal's psychometric function and reaction times.
Conclusions:
- Recurrent excitation, potentially mediated by NMDA receptors, is a candidate mechanism for integrating sensory stimuli over time.
- This mechanism supports the formation of categorical choices in decision-making neocortical networks.
- The model provides a framework for understanding the cellular basis of perceptual decision-making.