Differential temporal dynamics in motor imagery shaped by agent type and action duration

Area of Science:

• Cognitive Neuroscience
• Human-Robot Interaction
• Motor Control

Background:

• Motor imagery (MI) involves mentally simulating actions.
• The temporal dynamics of MI can be distorted.
• Embodied simulation theories suggest MI relies on shared neural resources with action execution.

Purpose of the Study:

• To investigate how observing robotic versus human actions influences the temporal dynamics of motor imagery.
• To determine if action duration and agent type (human vs. robot) affect temporal distortions in MI.
• To test the predictions of embodied simulation theories in the context of human-robot interaction.

Main Methods:

• Participants observed videos of goal-directed actions performed by a human or a NAO robot.
• After observation, participants performed motor imagery until an auditory stop signal.
• A temporal judgment task assessed the accuracy of the motor imagery stop point, comparing it to presented frames.

Main Results:

• A systematic temporal bias towards 'After' incorrect frames was observed in motor imagery.
• This 'After' bias was significantly reduced when imagining robotic actions compared to human actions.
• An increase in 'Before' errors for longer robotic actions suggested a perceived slowing of motor imagery for less biologically plausible movements.

Conclusions:

• The temporal characteristics of motor imagery are modulated by the observed agent's movement plausibility and action duration.
• Findings challenge simplistic embodied simulation accounts by highlighting differences in MI based on agent type.
• Motor imagery simulation is sensitive to the specific kinematic and dynamic properties of observed actions.