Direct observation of the neural computations underlying a single decision.

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Perceptual decision-making involves flexible cognitive processes.
• Neurons in parietal and prefrontal cortex are hypothesized to accumulate noisy evidence over time.
• Neural recordings suggest activity rises to a termination bound, with stochasticity causing variability.

Purpose of the Study:

• To elucidate the drift-diffusion signal on individual decisions in perceptual decision-making.
• To investigate the neural basis of choice and reaction time variability.
• To identify neuronal populations representing evidence accumulation.

Main Methods:

• Simultaneous neural recordings from hundreds of neurons in the lateral intraparietal cortex of monkeys.
• Monkeys performed a random dot motion direction decision task.
• Analysis of population activity to derive a drift-diffusion signal.

Main Results:

• A single scalar quantity from weighted population activity represents drift-diffusion.
• This signal directly supports the hypothesis that it underlies choice and reaction time variability.
• Specific neuronal subsets were identified as representing the integral and integrand of noisy evidence.

Conclusions:

• The identified drift-diffusion signal provides a mechanism for flexible cognitive processes in decision-making.
• This neural architecture explains variability in choice and reaction times.
• The findings offer a parsimonious explanation for evidence accumulation at the single-neuron level.