Electromyography and ultrasound investigation of upper-arm muscles

我的研究方向是研究上肢肌肉的激活。重点放在多功能二头肌上，因为从生理上讲，它的运动单位根据手臂的位置在肌肉的不同区域被激活，从解剖学上讲，因为在它的内表面可以观察到多达6个单独支配的隔室。我们的目标是找出这些间隔是如何自愿收缩的，从而表现为肌肉中的“肌肉”。这一基本方面对使用现代假肢的上肢截肢者具有实际影响，这些假肢需要许多肌电(EMG)控制信号来充分受益于他们产生许多动作的能力。在我们的实验室中，10个表面电极对放置在10个正常受试者的二头肌上，当他们的二头肌收缩时，他们实验了不同的上肢和手的位置，进行了肌电记录。对于这些表面信号，将与二头肌组织内电偶极子的存在相关联。这些偶极子的位置将通过在有限元上臂模型上应用的反问题求解方法来完成。这个模型将利用从上臂磁共振图像中获得的信息来建立。为了在这个模型中包括隔室的存在，从人类二头肌获得的高分辨率磁性图像将被分割，以识别沿着其二头肌长度的隔室的位置和大小。当我们的手势改变时，皮肤下的组织会发生变形，这可以用超声波(US)设备成像。在获得的图像上，将使用半自动分割算法来识别二头肌的短头和长头、肱骨顶部以及头静脉和贵要静脉。在这些已识别的结构之间，将测量距离和角度位置，并测量二头肌的短头和长头的面积。皮肤下的变化将被用于分析表面肌电信号。在收缩的二头肌中，高僵硬应力区将通过超声弹性成像来识别，在类似于EMG记录的条件下，高僵硬应力区预计与电偶极子位置重叠。最近，从许多表面肌电信号中提取肌肉协同效应被证明对假肢控制是有用的。我们发现，肌肉协同作用也可以从一块肌肉中提取出来，比如二头肌。作为概念验证，将利用二头肌5个信号的肌肉协同作用，让一个小型人形机器人在线复制正常受试者产生的手臂动作。几种模式的融合、对肌肉间隔的探索和肌肉的协同作用可能会导致对肌电假体的更容易控制。

My research program is oriented toward the study of upper limb muscles activation. Focus is on the multifunctional biceps brachii because physiologically, its motor units are activated in different regions of the muscle depending on the arm position, and anatomically because of up to 6 individually innervated compartments can be observed on its inner surface. Our objective is to find how those compartments could voluntarily be contracted and thus behave as “muscles” within a muscle. This fundamental aspect has a practical impact for upper limb amputees who use modern prostheses requiring many electromyographic (EMG) control signals to fully benefit from their ability to produce many movements. In our lab, EMG recordings were done with 10 surface electrode pairs positioned across the biceps of 10 normal subjects who experimented different upper limb and hand positions while their biceps was contracted. To those surface signals, will be associated the presence of electric dipoles within the biceps tissue. Location of those dipoles will be done with an inverse problem solving method applied on a finite element upper arm model. This model will be established with information obtained on upper arm magnetic resonance images. To include the presence of compartments within this model, high resolution magnetic images obtained from human biceps will be segmented to identify location and size of the compartments all along the length of their biceps. As our hand posture changes, tissues under the skin are subjected to deformations which can be imaged with an ultrasound (US) equipment. On the obtained images a semi-automatic segmentation algorithm will be used to identify the short and long head of the biceps, the top of the humerus bone and the cephalic and basilic veins. Between those identified structures, distance and angular position will be measured and area of the short and long heads of the biceps measured. Those under the skin changes will be used in the analysis of surface EMG signals. Within a contracted biceps, zones of high stiffness stress will be identified with US elastography, Obtained in condition similar to those of EMG recording, high stiffness zones are expected to overlap the electric dipole positions. Recently, extraction of muscular synergies from many surface EMG signals proved to be useful to control a prosthesis. We found that muscle synergy could also be extracted from a single muscle such as the biceps. As a proof of concept, muscle synergies of 5 signals of the biceps will be used to have a small humanoid robot to on-line duplicate arm movements produced by normal subjects. Fusion of several modalities, exploration of muscle compartments and muscular synergy could lead to an easier control of myoelectric prostheses.