Related Experiment Videos
Adaptive representation of dynamics during learning of a motor task
1Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge 02139.
Summary
The central nervous system (CNS) adapts to new motor dynamics by building an internal model of external forces. This model generalizes across the workspace, indicating flexible motor learning.
Area of Science:
- Motor control
- Computational neuroscience
- Robotics
Background:
- The central nervous system (CNS) must adapt motor commands to changing environmental dynamics.
- Motor adaptation allows for flexible control of movements in novel conditions.
Purpose of the Study:
- To investigate how the CNS learns to control movements under altered dynamics.
- To understand the internal representation of learned motor behavior.
Main Methods:
- Subjects performed reaching movements using a robotic manipulandum imposing force fields.
- Motor adaptation was assessed by observing changes in hand trajectories over time.
- Aftereffects were measured upon sudden removal of the force field.
Main Results:
- Initial movements were distorted by the force field, but performance recovered with practice.
- Aftereffects upon force field removal mirrored initial distortions, suggesting an internal model.
- Adaptation generalized to untrained regions, indicating a broadly tuned internal model.
Conclusions:
- The CNS constructs an internal model of environmental forces for motor control.
- This model is represented by broadly tuned computational elements, not a simple lookup table.
- The model operates in an intrinsic coordinate system related to joints and muscles, enabling generalization.