Mechanisms underlying the energy optimization of human walking

对许多人来说，走路是一件每天都会做的事情，不用多想，相对轻松。人们可以优雅地调整自己的步态，以适应不断变化的地形、任务要求和对身体的限制。然而，神经系统如何在所有可能的运动中导航，以迅速达到似乎是最佳协调策略，在很大程度上是未知的。 我们最近确定，实时选择步态的一个重要目标是能量最小化，因为人们倾向于以最大限度减少代谢能量消耗的方式行走。在这里，我们感兴趣的是了解这种能量优化是如何由神经系统完成的。为此，我们将结合人类行走实验和计算强化学习模型，系统地研究能量优化过程。具体地说，我们将调查三个因素如何促进或破坏一个人优化步态的能力，包括i)步态的可变性(即步态协调能力的变化)；ii)步态的能量结果的显著程度(即一个人可以轻松地感觉到能量消耗的变化)；以及iii)步态的能量结果的恒定性(即步行的能量结果改变的频率)。在每种情况下，我们都会扰乱走路的人或环境，以夸大其中一个因素，然后研究走路的人收敛到能量最佳步态的速度有多快，或者说有多快。最后，我们将建立一个系统，利用这种理解来增强步行人的优化过程，从而描述步态的变化，使他们能够更快地收敛到能量最优解。 这项拟议的研究将为人类神经系统如何在行走过程中如此优雅地进行能量优化提供基本和基础的理解。这对加拿大的研究和创新有许多广泛的影响，包括：了解伴随衰老或受伤的步态适应；设计考虑神经系统内部目标的康复计划；设计与用户协同工作以增强运动的辅助技术；以及加快运动技能的学习。此外，该研究项目将为跨学科的本科生和研究生提供丰富的培训环境，同样做好准备，通过在学术界或行业的职业生涯，造福于加拿大研究和创新的未来。

For many people, walking is something done every day, without much thought and with relative ease. People can gracefully adapt their gait to changing terrains, task demands, and constraints on their body. Yet, how the nervous system navigate the expanse of all possible movements, to rapidly arrive at what appears to be an optimal coordination strategy, is largely unknown. We have recently determined that an important objective in the real-time selection of our gait is energy minimizationpeople tend to walk in ways that minimize their metabolic energy expenditure. Here, we are interested in understanding how this energy optimization is accomplished by the nervous system. To do so, we will combine human walking experiments with a computational reinforcement learning model, to systematically investigate the energy optimization process. Specifically, we will investigate how three factors can facilitate or disrupt one's ability to optimize their gait, including i) variability in gait (i.e. the stride to stride changes in walking coordination); ii) saliency of the energetic consequences of gait (i.e. the ease with which one can sense changes in energy expenditure); and iii) constancy of the energetic consequences of gait (i.e. how frequently the energetic consequences of walking are changing). In each case, we will perturb the walking person or environment to exaggerate one of these factors and then study how rapidly, or if at all, the walking person can converge on the energy optimal gait. Finally, we will build a system that leverages this understanding to augment a walking persons optimization processprescribing changes to gait that will allow them to more rapidly converge on the energy optimal solution. The proposed research will provide a basic and foundational understanding of how the human nervous system so elegantly performs energy optimization during walking. This has a number of broad implications for Canadian research and innovation, including: understanding gait adaptations that accompany ageing or injury; designing rehabilitation programs that consider the nervous systems internal objectives; designing assistive technologies that work in concert with the user to augment movement; and expediating locomotor skill learning. In addition, this research program will provide a rich training environment for cross-disciplinary undergraduate and graduate students, equally prepared to benefit the future of Canadian research and innovation through careers in either academia or industry.