Self-Organizing Neural Circuits for Sensory-Guided Motor Control

摘要

The reported projects developed mathematical models to explain how self-organizing neural circuits that operate under continuous or intermittent sensory guidance achieve flexible and accurate control of human movement. Neural models were developed for the control of visually guided arm/hand movements, saccadic eye movements, and limb gait transitions. These circuits generate movement trajectories, adapt movement execution on the fly to unforseen contingencies, and improve accuracy over time by learning to act in anticipation of predictable contingencies. The circuits meet behavioral, neurobiological, and design constraints. Thus, the proposed circuits have operating characteristics that match those documented for human performance and learning, such as voluntary control of speed and amplitude, transfer of learning, and learned recovery from damage to parts of a circuit. The circuits also exhibit stability, robustness, short-term flexibility, and long-term adaptability. The circuits also provide an integrative explanation of many neuroanatomical, neurophysiological, and biophysical observations. By satisfying.