Manufacture of a next-generation artificial limb system with sensory function that can be driven voluntarily using a neuro-machine interface

我们构建了一个假肢系统的原型，有了这个系统，受试者就能感觉到身体的感觉，就像他们用自己的手触摸物体一样，也能根据针肌电图信息自主地移动假肢。我们通过对感觉神经纤维的微电刺激来唤起被试的躯体感觉，并试图利用相关周围神经记录的运动神经活动信息来控制假肢系统的手和手指的运动。然而，我们所研制的筛电极的性能还不足以稳定地记录运动神经信号，我们使用相关肌肉的针肌电信号代替运动神经信号来控制假肢系统的运动。评估实验由健康受试者采用主从模式，采用人手形状的机械手系统进行；机器人手上的传感器检测到的压力或振动刺激被传送到计算机上，并确定刺激感觉神经纤维的电脉冲频率。然后这些电脉冲通过钨微电极输出到相关的感觉神经纤维。使用与感觉系统相同的钨微针电极测量相关肌肉的针肌电信息，并根据针肌电信息确定手指关节的抓握力和位置。结果表明，该系统工作满意。开发能够测量运动神经纤维活动的神经电极，并能够刺激具有足够数量通道的感觉神经纤维，仍然是创建临床可接受的系统的迫切需求。

We constructed a prototype of an artificial limb system with which subjects were able to feel somatic sensations exactly as if they were touching the object with their own hands and were also able to move the artificial arm voluntarily exactly as intended using needle EMG information.We used microelectrical stimulation of the sensory nerve fibers in order to evoke somatic sensations in the subjects and we attempted to control the movement of the hand and fingers of the artificial limb system using information of the motor nerve activities recorded from the concerned peripheral nerves. However, the performance of the sieve electrode which we developed was not yet sufficient for steadily recording motor nerve signals, we used needle EMG signals of the concerned muscles instead of motor nerve signals in order to control the movement of the artificial limb system.The experiment for evaluation was performed by a healthy subject in a master-slave format using a robot hand system with the shape of human hand; pressure or vibration stimuli detected by the sensor on the robot hand were transferred to a computer and the frequency of the electrical pulses for stimulating the sensory nerve fiber was determined. These electrical pulses were then output to the concerned sensory nerve fiber via a tungsten microelectrode.Needle EMG information was measured from the concerned muscles using the same tungsten micro-needle electrodes as were used with the sensory system and the grasping power and the positions of the finger joints were determined from the needle EMG information.The results showed that the system worked satisfactorily. Development of nerve electrodes that can measure activity of motor nerve fibers and can stimulate sensory nerve fibers with a sufficient number of channels remains an urgent demand for the creation of a system acceptable for clinical use.