Mechanisms of Energy Optimization in Human Locomotion

点击翻译按钮获取中文摘要

Why do we move the way we do? Why do we prefer to shift our gaze with a particular timing, reach to a cup using particular muscles, and walk with a particular step width? Perhaps the most general answer we have for these questions is that people prefer movements that are, in some sense, optimal. This optimization principle underlies theories on the control of reaching, grasping, gaze and gait. The principle is not just general, but also useful-it has provided insight into healthy and pathological behaviour, as well as the functions of different brain areas. Remarkably, how the nervous system performs this optimization is largely unknown. Our long-term vision is to use both experiments and theory to understand how the nervous system learns optimal movements, and then apply our findings to accelerate motor learning. We have three short-term objectives. Our first objective is to further understand the algorithms that the nervous system uses to minimize energetic cost. Here we seek to answer one of the important mysteries from our recent work-how the nervous system reliably initiates optimization. We will also determine how the nervous system optimizes control across many gait variables, including step frequency and step width, as well as the limits to this multi-dimensional optimization. Our second objective is to determine how the nervous system senses energetic cost. Here we will quantify the role of feedback from ergoreceptors, which directly sense by-products of muscle metabolism and are known to play a role in controlling ventilation and sensing fatigue. We will also determine the role of indirect energetic cost sensing, arising from muscle proprioceptor feedback. The third objective is to leverage our understanding of the nervous system's learning mechanisms to develop methods for accelerating energy optimization. Here we will focus on learning to gain assistance from powered exoskeletons as a model system. We will manipulate experience with new controllers, the complexity of the controller, and the frequency with which the controllers are changed. The fundamental impact of this work is due, in part, to uncovering the mechanisms that underlie the most general principle we have for human walking-that we prefer to walk in ways that minimize energetic cost. This new understanding has relevance not just to the particular objective of energy minimization and the particular task of walking, but also for how the nervous system optimizes for other objectives, such as accuracy and stability, in other tasks, such as reaching and eye movements. One application of this work is to powered exoskeletons and prostheses where the acceleration of motor learning can be immediately applied to increase the effectiveness of this technology at improving users' lives. Accelerating motor learning in novel conditions will also be useful for the design of running shoes, the customization of rehabilitation programs, and the optimization of athletic training.