Cho Woorim, Barradas Victor R, Schweighofer Nicolas, Koike Yasuharu
School of Engineering, Tokyo Institute of Technology, Yokohama, Japan.
Institute of Innovative Research, Tokyo Institute of Technology, Yokohama, Japan.
Front Hum Neurosci. 2022 May 27;16:805452. doi: 10.3389/fnhum.2022.805452. eCollection 2022.
Muscle synergy analysis surface electromyography (EMG) is useful to study muscle coordination in motor learning, clinical diagnosis, and neurorehabilitation. However, current methods to extract muscle synergies in the upper limb suffer from two major issues. First, the necessary normalization of EMG signals is performed maximum voluntary contraction (MVC), which requires maximal isometric force production in each muscle. However, some individuals with motor impairments have difficulties producing maximal effort in the MVC task. In addition, the MVC is known to be highly unreliable, with widely different forces produced in repeated measures. Second, synergy extraction in the upper limb is typically performed with a multidirection reaching task. However, some participants with motor impairments cannot perform this task because it requires precise motor control. In this study, we proposed a new isometric rotating task that does not require precise motor control or large forces. In this task, participants maintain a cursor controlled by the arm end-point force on a target that rotates at a constant angular velocity at a designated force level. To relax constraints on motor control precision, the target is widened and blurred. To obtain a reference EMG value for normalization without requiring maximal effort, we estimated a linear relationship between joint torques and muscle activations. We assessed the reliability of joint torque normalization and synergy extraction in the rotating task in young neurotypical individuals. Compared with normalization with MVC, joint torque normalization allowed reliable EMG normalization at low force levels. In addition, the extraction of synergies was as reliable and more stable than with the multidirection reaching task. The proposed rotating task can, therefore, be used in future motor learning, clinical diagnosis, and neurorehabilitation studies.
肌肉协同分析中的表面肌电图（EMG）对于研究运动学习、临床诊断和神经康复中的肌肉协调很有用。然而，目前在上肢提取肌肉协同作用的方法存在两个主要问题。首先，EMG信号的必要归一化是通过最大自主收缩（MVC）进行的，这需要每块肌肉产生最大等长力。然而，一些有运动障碍的个体在MVC任务中难以产生最大努力。此外，已知MVC非常不可靠，重复测量时产生的力差异很大。其次，上肢的协同作用提取通常是通过多方向伸手任务进行的。然而，一些有运动障碍的参与者无法执行此任务，因为它需要精确的运动控制。在本研究中，我们提出了一种新的等长旋转任务，该任务不需要精确的运动控制或大力气。在这个任务中，参与者将由手臂端点力控制的光标保持在以指定力水平以恒定角速度旋转的目标上。为了放宽对运动控制精度的限制，目标被加宽和模糊化。为了在不需要最大努力的情况下获得用于归一化的参考EMG值，我们估计了关节扭矩和肌肉激活之间的线性关系。我们评估了年轻神经正常个体在旋转任务中关节扭矩归一化和协同作用提取的可靠性。与使用MVC进行归一化相比，关节扭矩归一化允许在低力水平下进行可靠的EMG归一化。此外，协同作用的提取与多方向伸手任务一样可靠且更稳定。因此，所提出的旋转任务可用于未来的运动学习、临床诊断和神经康复研究。
