Linking perception, cognition, and action: psychophysical observations and neural network modelling

Area of Science:

• Cognitive Neuroscience
• Computational Neuroscience
• Neural Networks

Background:

• Perception, decision-making, and movement planning are increasingly recognized as interconnected brain processes.
• The precise neural mechanisms underlying this integration remain a subject of ongoing scientific investigation.

Purpose of the Study:

• To investigate how categorization difficulty and movement parameters influence decision-making and response times.
• To develop and validate a biologically plausible neural network model that integrates sensory and motor information for decision tasks.

Main Methods:

• Human subjects performed a temporal categorization task, classifying intervals as short or long.
• Movement choices involved directing a cursor to targets with varying angular separations.
• A novel neural network model with mutually inhibiting populations, accumulation, and memory nodes was developed.

Main Results:

• Categorization difficulty significantly impacted performance, reaction time (RT), and movement time (MT).
• Target distance also influenced RT and MT, indicating the incorporation of motor planning into decisions.
• The proposed neural network model accurately replicated observed human performance, RT, and MT across conditions.

Conclusions:

• Decision-making processes inherently integrate perceptual information with motor planning considerations.
• The developed neural network provides a biologically plausible framework for understanding integrated sensory-motor decision-making.
• The model successfully predicts behavioral outcomes and offers a basis for future predictions on untested parameters.