Guiding multiple muscles and joints without redundancy problems: the principle of minimal interaction

这项研究计划的目的是减少理解神经系统以连贯和特定任务的方式控制多个肌肉和身体自由度的能力的基本原理的差距。在NSERC前期研究的基础上，我们成功验证并进一步发展了电机控制的阈值控制理论。应用于涉及整个身体的动作，这一理论表明，多个骨骼肌的肌电图（EMG）活动来自身体的实际配置（Q）与其内部几何图像或由大脑根据任务需求修改的参考（R）配置的比较。这些配置之间的差异可以被认为是影响身体所有骨骼肌的EMG活动的全局因素。局部生物力学和神经生理学因素（例如，自体反射和异体反射）结合全局因素使肌电模式多样化。还假设，为了产生运动动作，神经系统移动适当效应器（例如手）的阈值（虚拟）位置。作为响应，出现的神经和肌肉活动往往会在由内部和外部约束定义的限制中最小化效应器的实际位置和虚拟位置之间的差异。神经肌肉系统的每个元件可以或可以不参与动作，这取决于其减小效应器的虚拟位置和实际位置之间的差异的能力。这些想法提供了一个解决方案，不适定的冗余问题，在控制多个肌肉和身体部分。目前的研究建议的目的是测试这些新的想法，生产的适用性，第一，不同的手腕手势，第二，一系列的动作与从坐到站和走的过渡。测试将涉及这些行为的实验分析和数学建模。本研究不仅有助于了解运动控制的基本原理，而且有助于设计假肢和机器人学中的多关节假肢。

The aim of this research program is to diminish the gap in the understanding of the basic principles underlying the ability of the nervous system to control multiple muscles and degrees of freedom of the body in a coherent and task-specific way. Based on the previous research funded by NSERC, we successfully verified and further developed the threshold control theory of motor control. Applied to actions involving the whole body, this theory suggests that the electromyographic (EMG) activity of multiple skeletal muscles emerges from the comparison of the actual configuration (Q) of the body with its internal geometric image or referent (R) configuration modified by the brain according to task demands. The difference between these configurations can be considered as a global factor influencing the EMG activity of all skeletal muscles of the body. Local biomechanical and neurophysiological factors (e.g., autogenic and heterogenic reflexes) combined with the global factor diversify muscular EMG patterns. It was also assumed that, to produce a motor action, the nervous system shifts the threshold (virtual) position of an appropriate effector (e.g. the hand). In response, the emerging neural and muscular activity tends to minimize the difference between the actual and the virtual position of the effector, in the limits defined by internal and external constrains. Each element of the neuromuscular system may or may not be involved in the action depending on its ability to diminish the difference between the virtual and the actual position of the effector(s). These ideas offer a solution to the ill-posed redundancy problem in the control of multiple muscles and body segments. The present research proposal is designed to test the applicability of these new ideas to the production, first, of different hand-wrist gestures and, second, to a sequence of actions associated with the transition from sitting to standing and walking. Testing will involve both experimental analysis and mathematical modeling of these actions. The project may help not only in the understanding of the basic principles of motor control but also in designing muti-joint artificial limbs in prosthetics and robotics.