Internal models underlying fingertip force control during object manipulation in humans.

Area of Science:

• Motor control and learning
• Human-robot interaction
• Biomechanics

Background:

• Skillful object manipulation relies on predictive fingertip force scaling.
• Internal models, formed through experience, underpin this predictive control.
• Generalizability of predictive control to the contralateral limb is not fully understood.

Purpose of the Study:

• To investigate the transfer of predictive force control from the dominant (right) to the non-dominant (left) hand.
• To determine if practice-based predictive control for object weight and center of mass (CoM) location generalizes across hands.
• To explore the hierarchical nature of predictive motor control.

Main Methods:

• Participants performed object lifting tasks with varying weights and CoM displacements.
• Force transducers measured fingertip forces during lifts.
• Predictive force scaling was assessed after practice with the right hand and subsequent single trial with the left hand.

Main Results:

• Predictive force scaling for object weight generalized to the contralateral hand.
• Precise force adjustments for CoM-induced tilting did not transfer to the contralateral hand.
• Force rates were higher for heavier objects (800g) regardless of hand.

Conclusions:

• Predictive control exhibits a dichotomy in generalizability: weight-based parameters transfer, while individuated digit control for CoM does not.
• This suggests a hierarchical control system where higher-level parameters (weight) are globalized, while lower-level parameters (CoM adjustments) are not.
• Findings have implications for understanding motor learning and designing assistive robotic devices.