ADVISOR:
Lena Ting - Biomedical Engineering, GA Tech

COMMITTEE MEMBERS:
Tom Burkholder – Applied Physiology, GA Tech
Daniel Goldman – Physics, GA Tech
Karen Liu – Computer Science, GA Tech
Randy Trumbower – Rehabilitation Medicine, Emory

Interactions between the neural and musculoskeletal systems are a prerequisite for the production of robust movement. In spite of this, the neural control and musculoskeletal structure underlying biological movements are typically studied independently, with little attention paid to how changes in one may affect the other. Understanding these interactions may be critical to improving current rehabilitation technologies and therapy methods. As an example, balance disorders are multifactorial in nature and identifying whether biomechanical or neural changes are the source of instability remains an unanswered question.

I have used a combined experimental and modeling approach to understand neural and biomechanical interactions governing human balance control. I developed a simple four-bar linkage model with delayed feedback to investigate frontal-plane standing balance. Using methods from time-delay systems I present evidence from this model that biomechanical structure is important for behavioral function and show that neural control and biomechanical structure co-vary for stable human balance. Predictions from the model were tested experimentally to dissociate the effects of inertia and postural configuration on balance. In addition, I applied robust control methods to solve the difficult problem of comparing the relative performance between neuromechanical systems that differ in parameter values and predicted a common mechanism to explain changes in neural control across biomechanical contexts.

In the future, the analytical tools and simulation methods I have developed can be generalized to investigate changes in neuromechanical interactions of various deficits in biomechanics (ACL rupture, amputation) and neural control (Parkinson’s disease, stroke).

Furthermore, this approach can be used to explain how neural control and biomechanical structure relate to the diversity of animal form and function, as well as suggest biomimetic control policies for robotics.