Date of Graduation
School of Medicine
Lee E Fisher
Lee E Fisher
James W Lewis
Humans can perform complicated movements like writing or running without giving them much thought. The scientific understanding of principles guiding the generation of these movements is incomplete. How the nervous system ensures stability or compensates for injury and constraints – are among the unanswered questions today. Furthermore, only through movement can a human impose their will and interact with the world around them. Damage to a part of the motor control system can lower a person’s quality of life. Understanding how the central nervous system (CNS) forms control signals and executes them helps with the construction of devices and rehabilitation techniques. This allows the user, at least in part, to bypass the damaged area or replace its function, thereby improving their quality of life.
CNS forms motor commands, for example a locomotor velocity or another movement task. These commands are thought to be processed through an internal model of the body to produce patterns of motor unit activity. An example of one such network in the spinal cord is a central pattern generator (CPG) that controls the rhythmic activation of synergistic muscle groups for overground locomotion. The descending drive from the brainstem and sensory feedback pathways initiate and modify the activity of the CPG. The interactions between its inputs and internal dynamics are still under debate in experimental and modelling studies. Even more complex neuromechanical mechanisms are responsible for some non-periodic voluntary movements. Most of the complexity stems from internalization of the body musculoskeletal (MS) system, which is comprised of hundreds of joints and muscles wrapping around each other in a sophisticated manner. Understanding their control signals requires a deep understanding of their dynamics and principles, both of which remain open problems.
This dissertation is organized into three research chapters with a bottom-up investigation of motor control, plus an introduction and a discussion chapter. Each of the three research chapters are organized as stand-alone articles either published or in preparation for submission to peer-reviewed journals. Chapter two introduces a description of the MS kinematic variables of a human hand. In an effort to simulate human hand motor control, an algorithm was defined that approximated the moment arms and lengths of 33 musculotendon actuators spanning 18 degrees of freedom. The resulting model could be evaluated within 10 microseconds and required less than 100 KB of memory. The structure of the approximating functions embedded anatomical and functional features of the modelled muscles, providing a meaningful description of the system. The third chapter used the developments in musculotendon modelling to obtain muscle activity profiles controlling hand movements and postures. The agonist-antagonist coactivation mechanism was responsible for producing joint stability for most degrees of freedom, similar to experimental observations. Computed muscle excitations were used in an offline control of a myoelectric prosthesis for a single subject. To investigate the higher-order generation of control signals, the fourth chapter describes an analytical model of CPG. Its parameter space was investigated to produce forward locomotion when controlled with a desired speed. The model parameters were varied to produce asymmetric locomotion, and several control strategies were identified. Throughout the dissertation the balance between analytical, simulation, and phenomenological modelling for the description of simple and complex behavior is a recurrent theme of discussion.
Sobinov, Anton, "Description of motor control using inverse models" (2019). Graduate Theses, Dissertations, and Problem Reports. 4073.