Evidence for sparse synergies in grasping actions.

Area of Science:

• Neuroscience
• Biomechanics
• Motor Control

Background:

• Hand actions are orchestrated by neural synergies, not individual muscles.
• Synergy-based control offers potential for intrinsic sparsity and shared representations across actions.

Purpose of the Study:

• To evaluate three hypotheses of hand-action optimal-control: sparse-combination (SC), sparse-elements (SE), and double-sparsity (DS).
• To determine the most effective model for representing hand actions based on synergy and degree-of-freedom usage.

Main Methods:

• Analysis of hand kinematic data from six human subjects.
• Subjects performed nine distinct reach-to-grasp actions.
• Evaluation of hypotheses using computational modeling of action representation.

Main Results:

• The double-sparsity (DS) hypothesis was best supported by the experimental data.
• Hand actions are represented by a combination of a substantial number of synergies.
• Each synergy involves a restricted set of degrees-of-freedom, and distinct synergies are employed for different tasks.

Conclusions:

• The findings support a double-sparsity model for optimal hand-action control.
• This suggests a neural strategy that balances the number of synergies with the complexity of their constituent degrees-of-freedom.
• This representation allows for efficient and adaptable control of diverse hand movements.