Open Access
Hu, Xiaogang
Graduate Program:
Master of Science
Document Type:
Master Thesis
Date of Defense:
May 12, 2008
Committee Members:
  • Karl Maxim Newell, Thesis Advisor
  • Dagmar Sternad, Thesis Advisor
  • Motor control
  • Motor learning
  • Variability
Our motor system has the ability to reliably achieve the task goal while revealing substantial variability. The inherent system variability and perturbations from external environment continuously challenge the central nervous system (CNS) during the control of movement. The general purpose of this study is to investigate the role of variability (both internal and external to the system) in the control of movement and motor learning. A virtual throwing task was adopted as a model task where different movement executions lead to the same task performance. Subjects manipulated a lever arm with a single-joint elbow movement and released a virtual ball that traversed a concentric force field to hit a target. From basic mechanics, release angle and velocity at ball release fully determined the ball trajectory and hitting success. Based on this task a nonlinear solution manifold representing the set of all solutions to the task was calculated. Experiment 1 tested whether subjects were sensitive to motor variability and find solutions in the solution manifold with least error-sensitivity and whether subjects preferred solutions with minimum velocity. Results showed that with practice solutions became aligned with the solution manifold and concentrated at the most error-tolerant locations. The individually chosen velocities at release covered a wide range suggesting that subjects did not gravitate to minimum velocity. Even with prolonged practice, subjects still had substantial variability in error-sensitive dimension. Experiment 2 investigated whether the control of movement accuracy and variability could benefit from the enhancement of potentially perturbing information. Based on stochastic resonance (SR), detection of sensory signals can be enhanced in the presence of externally added noise. Error-dependent noise was added to the angle of ball release; this essentially enhanced the result variability of throws via visual feedback to the subjects. In addition, the threshold signaling successful hits was lowered. Results showed that subjects significantly improved accuracy and reduced motor variability with enhanced error information compared with controls. This improved performance was retained for a prolonged period even after removal of the enhanced error information. These results showed that the CNS can alter its motor strategy to compensate for perturbations from the internal system and external environment. The control of movement can be enhanced by external noise via SR. Experiment 3 examined different magnitudes of noise to examine whether SR was present. Four different levels of error-dependent noise were added to the angle of ball release. The results showed that an optimal level of noise best enhanced the control of movement accuracy and variability comparing with three other noise levels. This work indicated that with externally added optimal level of noise, the CNS could adapt and further decrease its internal noise. Taken together, the three studies provided insight on how the CNS selects control strategies in the present of sensorimotor variability and perturbations from the environment. It extended our understanding on the role of intrinsic and extrinsic variability during the control of movement. It suggested possible intervention protocols for motor learning and rehabilitation.