Electronic External Control of the Neuromuscular System

(1)基于对电刺激下神经肌肉系统反应的分析，建立了神经肌肉系统和骨骼系统的模型。在此基础上，开发了fes诱导截瘫患者运动的基本仿真系统。采用动态优化的方法，在仿真系统上自动生成站立激励模式。(2)验证了肌电模式线性插值和人工神经网络可以用于刺激模式生成。此外，我们还开发了一个系统，可以让患者在没有医学知识的情况下修改刺激模式，并在控制正常人手腕角度的实验中验证了它的能力。(3)将基于汉明编码的纠错检测电路（ECC）应用于可植入FES实验系统中，以抑制外部电磁噪声的影响。在初步的可植入FES系统上验证了ECC对增产数据传输可靠性的提高。(4)建议采用以下两种方法控制FES预防尿失禁：1)左右阴部神经刺激脉冲相位变换180度，在短时间内提高尿道闭锁力。2)长时间交替刺激其中一条阴部神经，尿道闭锁力低。(5)分析外部机械刺激引起的神经动作电位波形，提取信息用于FES应用中的反馈控制。发现聚类分析对估计引起神经信号的刺激是有用的。(6)基于耳后肌、舌部运动和点头等人工神经网络识别头部运动的FES系统控制指令获取方法初步确定具有临床应用的可行性。

(1)Models of the neuromuscular and the skeletal systems were developed based on results of analyzing responses of the neuromuscular system during electrical stimulation. Then, a basic simulation system of FES-induced motions for the paraplegics was developed. A stimulation pattern for standing up was made up automatically on the simulation system using the dynamic optimization.(2)It was verified that linear interpolation of EMG patterns and artificial neural network could be used for stimulation pattern generation. In addition, a system that patients could modify stimulation patterns without medical knowledge was developed and its capabilities were verified in experiments of controlling normal subjects' wrist angles.(3)The error detecting and correcting circuit (ECC) based on the Hamming coding was applied to an experimental implantable FES system for rejecting influence of external electromagnetic noise. Improvement of the reliability of transmitted stimulation data by using the ECC was verified on the preliminary implantable FES system.(4)It was suggested that the following two methods were useful in FES control for preventing the incontinence : 1) 180 degrees shift of the phase between stimulus pulses applying to the left and right pudendal nerves for high urethral closing force in short time. 2) stimulation to one of the pudendal nerves alternately for long time low closing force of the urethra.(5)Waveforms of nerve action potentials elicited by external mechanical stimulation were analyzed for extracting information for feed back control in FES application. Cluster analysis was found to be useful for estimating the stimulation that elicited the nerve signals.(6)Methods of getting control commands for FES system, which were based on the posterior auricular muscle, the tongue movement, and head movement identification by artificial neural network e.g. nodding, were confirmed to have fundamental feasibility of using in clinical application.