Osu R, Gomi H
Kawato Dynamic Brain Project, ERATO, Japan Science and Technology Corporation, Kyoto, 619-0288, Japan.
J Neurophysiol. 1999 Apr;81(4):1458-68. doi: 10.1152/jn.1999.81.4.1458.
Stiffness properties of the musculo-skeletal system can be controlled by regulating muscle activation and neural feedback gain. To understand the regulation of multijoint stiffness, we examined the relationship between human arm joint stiffness and muscle activation during static force control in the horizontal plane by means of surface electromyographic (EMG) studies. Subjects were asked to produce a specified force in a specified direction without cocontraction or they were asked to keep different cocontractions while producing or not producing an external force. The stiffness components of shoulder, elbow, and their cross-term and the EMG of six related muscles were measured during the tasks. Assuming that the EMG reflects the corresponding muscle stiffness, the joint stiffness was predicted from the EMG by using a two-link six-muscle arm model and a constrained least-square-error regression method. Using the parameters estimated in this regression, single-joint stiffness (diagonal terms of the joint-stiffness matrix) was decomposed successfully into biarticular and monoarticular muscle components. Although biarticular muscles act on both shoulder and elbow, they were found to covary strongly with elbow monoarticular muscles. The preferred force directions of biarticular muscles were biased to the directions of elbow monoarticular muscles. Namely, the elbow joint is regulated by the simultaneous activation of monoarticular and biarticular muscles, whereas the shoulder joint is regulated dominantly by monoarticular muscles. These results suggest that biarticular muscles are innervated mainly to control the elbow joint during static force-regulation tasks. In addition, muscle regulation mechanisms for static force control tasks were found to be quite different from those during movements previously reported. The elbow single-joint stiffness was always higher than cross-joint stiffness (off-diagonal terms of the matrix) in static tasks while elbow single-joint stiffness is reported to be sometimes as small as cross-joint stiffness during movement. That is, during movements, the elbow monoarticular muscles were occasionally not activated when biarticular muscles were activated. In static tasks, however, monoarticular muscle components in single-joint stiffness were increased considerably whenever biarticular muscle components in single- and cross-joint stiffness increased. These observations suggest that biarticular muscles are not simply coupled with the innervation of elbow monoarticular muscles but also are regulated independently according to the required task. During static force-regulation tasks, covariation between biarticular and elbow monoarticular muscles may be required to increase stability and/or controllability or to distribute effort among the appropriate muscles.
肌肉骨骼系统的刚度特性可通过调节肌肉激活和神经反馈增益来控制。为了理解多关节刚度的调节机制,我们通过表面肌电图(EMG)研究,在水平面静态力控制过程中,研究了人体手臂关节刚度与肌肉激活之间的关系。受试者被要求在不进行协同收缩的情况下,沿指定方向产生指定的力,或者在产生或不产生外力的同时保持不同程度的协同收缩。在任务过程中,测量了肩部、肘部的刚度分量及其交叉项,以及六条相关肌肉的肌电图。假设肌电图反映了相应的肌肉刚度,通过使用双连杆六肌肉手臂模型和约束最小二乘误差回归方法,从肌电图预测关节刚度。利用该回归中估计的参数,单关节刚度(关节刚度矩阵的对角项)成功分解为双关节和单关节肌肉成分。尽管双关节肌肉作用于肩部和肘部,但发现它们与肘部单关节肌肉强烈共变。双关节肌肉的首选力方向偏向于肘部单关节肌肉的方向。也就是说,肘关节由单关节和双关节肌肉的同时激活来调节,而肩关节主要由单关节肌肉调节。这些结果表明,在静态力调节任务中,双关节肌肉主要受神经支配以控制肘关节。此外,发现静态力控制任务的肌肉调节机制与先前报道的运动过程中的机制有很大不同。在静态任务中,肘关节单关节刚度总是高于交叉关节刚度(矩阵的非对角项),而在运动过程中,肘关节单关节刚度有时据报道与交叉关节刚度一样小。也就是说,在运动过程中,当双关节肌肉被激活时,肘部单关节肌肉偶尔不会被激活。然而,在静态任务中,每当单关节和交叉关节刚度中的双关节肌肉成分增加时,单关节刚度中的单关节肌肉成分就会显著增加。这些观察结果表明,双关节肌肉不仅与肘部单关节肌肉的神经支配简单耦合,而且还根据所需任务进行独立调节。在静态力调节任务中,双关节和肘部单关节肌肉之间的共变可能是为了增加稳定性和/或可控性,或者在适当的肌肉之间分配力量。