Rat movements reflect internal decision dynamics in an evidence accumulation task

Area of Science:

• Neuroscience
• Cognitive Science
• Animal Behavior

Background:

• Perceptual decision-making integrates sensory evidence, planning, and motor execution.
• The interplay between decision processes and motor execution remains incompletely understood.
• Some models posit decisions precede movement, while others suggest movements reflect ongoing decisions.

Purpose of the Study:

• To investigate the relationship between decision dynamics and motor actions leading to a choice.
• To develop and validate a novel model incorporating movement data into decision-making models.
• To enhance the predictive power of decision models by including movement parameters.

Main Methods:

• Utilized a free-response pulse-based evidence accumulation task with rats.
• Developed a motion-based drift diffusion model (mDDM) using video pose estimation.
• Constrained decision parameters trial-by-trial using movement variables.

Main Results:

• The mDDM provided a superior fit to rat choice behavior compared to traditional drift diffusion models.
• A period of head immobility preceding choice was observed and correlated with decision bounds.
• Stimuli presented during the immobility period had the most significant impact on choice selection.

Conclusions:

• Internal decision dynamics are reflected in observable movements.
• Integrating movement parameters into diffusion-to-bound models significantly improves their performance.
• The mDDM offers a more accurate description of animal choice behavior in evidence accumulation tasks.