正常生理与帕金森症病理运动控制的计算与神经模型研究

发布时间：2018-05-16 12:38

本文选题：运动控制 + 虚拟手臂模型　；参考：《上海交通大学》2015年博士论文

【摘要】：研究生理和病理运动控制,是我们认识感觉运动系统,尤其是大脑脊髓中枢神经系统的结构和生理功能的重要内容,也是为运动功能障碍患者提供运动功能补偿和康复的基础。神经运动控制的研究方法分实验观测分析与计算模型仿真两种:实验研究对感觉运动系统中各层次回路的辨识和行为功能认识,是建立计算模型的基础;而计算模型仿真结果,又可以提供在实验条件下无法观测的系统状态的神经信息。因此,实验分析和模型仿真方法在认识中枢神经系统感知运动功能中起到互补的作用。本论文探讨的科学问题是,在姿态和运动双重控制的理论框架下,大脑和脊髓中各神经回路在整合感觉信息、控制运动执行过程中的模块化功能。本论文的工作突破了现阶段研究方法的瓶颈,将感觉运动系统各子系统部分的大量数据和已知信息整合成为多尺度计算模型,从系统层面研究生理和病理运动的神经控制机制。本论文在前期工作的基础上,完善了多尺度虚拟手臂模型的仿真平台,并验证了虚拟手臂模型的真实性和有效性;利用虚拟手臂模型,分析并提出了肌梭伽马神经支配的中枢非线性编码假设;搭建人体上肢运动实验采集分析平台,并分析了帕金森病肢体震颤中,肌群的协同发放与震颤幅度的相关关系,为使用虚拟手臂模型进一步探索帕金森病震颤的神经病理机制提供了实验数据。本论文的主要研究内容和成果包括以下四个方面:1.在前期工作基础上开发多尺度整合的虚拟手臂模型。我们将前期工作中开发的上肢肌骨系统生物力学模型、虚拟肌肉、肌梭、高尔基腱器官等模块,和新增的脊髓反射回路和脊髓固有神经元网络模型模块整合起来,移植到SIMULINK建模仿真平台中,优化参数,统一接口,整合为一个具有真实生理特征的虚拟手臂系统模型。我们采用了模块化的模型结构,使其具有模块易于更新替换,拓展性强,可根据需要定制化和个性化等特点。该模型在模拟正常和病理神经运动行为,研究神经运动控制机制中有巨大潜力,并为康复应用的开发和临床前评估提供了一个可靠且高效的仿真平台和环境。2.验证虚拟手臂模型神经力学行为的真实性和有效性,定量评估前馈和反馈控制对维持上肢端点姿态稳定性的贡献。手刚度反映了手的阻抗特性,手的运动变异性描述了固有神经噪声影响下的手的稳定范围,这两个神经力学行为指标可以定量评价姿态控制下的手稳定性。将系统模型仿真的神经力学行为(手刚度与变异性)与人体实验测量数据比较,确认虚拟手臂模型的真实性和有效性。由于实验测量手段的局限性,现阶段我们还无法在系统层面准确测量脊髓反射回路对上肢端点的姿态稳定性的贡献,因而在不同的运动控制理论中,对本体感觉反馈在运动控制中的作用有不同解释。虚拟手臂模型使我们可以通过设计开环和闭环条件下的仿真实验,定量评估多关节系统中,闭环感觉反馈对系统阻抗特性和稳定性的影响。结果表明,在稳态的闭环感觉反馈作用下,手刚度椭圆的面积增大35.75±16.99%(均值±标准差,下同),手变异性椭圆的面积缩小49.41±21.19%,显示出本体感觉反馈在脊髓层面的闭环姿态控制中,对抗内部神经噪声和外界环境扰动,维持多关节肢体稳定性的显著贡献。3.研究肌梭运动神经控制中可能的关节角度编码模式,及其对外周本体感觉传入中关节角度编码的影响。实验证据表明下行的肌梭运动神经发放活动与关节运动轨迹相关,以某种形式编码了关节角度信息作为中枢对外周姿态控制的参考信号。同时肌梭运动神经起到维持不同姿态下肌梭梭内肌纤维张力的作用,调节感觉末梢的灵敏度以控制本体感觉发放频率中对关节角度的线性编码。由于人体正常运动状态下的肌梭运动神经元发放模式无法通过电生理技术手段测量,我们尚不了解肌梭运动神经中的关节角度编码模式。基于虚拟手臂模型仿真实验平台,我们提出并检验了三种可能的肌梭运动控制编码模态假设。各假设的正确性通过检验假定的肌梭运动神经编码模态下,仿真的初级感觉传入神经发放是否与实验测量相匹配的方法验证。研究结果表明,在姿态控制中,神经系统可能采取了对肌梭灵敏度的非线性控制策略,使肌梭运动神经承担起中枢向外周系统传输关节角度编码信息的功能。4.研究肌肉间同步活动对上肢帕金森震颤运动的贡献,及其产生的大脑和脊髓神经回路基础。建立实验采集帕金森震颤运动和肌电图的方法,采用相干和互相关等方法分析帕金森震颤中各肌肉的节律性发放活动,提出“paired coherence”和“pool-averaged coherence”两种评价指标定量评估肌群间活动的同步水平,并分析其与关节震颤运动幅度间的相关性。分析显示,不同患者的上肢肌群震颤活动的同步水平存在较大差异,且与上肢关节的震颤幅度成正相关。该结果表明,肌肉非自主发放活动在肌群间的同步,是驱动帕金森震颤运动的重要因素。这一研究为最新的解释帕金森震颤运动产生机制的“调制-开关模型”假说提供了支持,并进一步提出源自小脑和基底节系统的中枢振荡信号,经过脊髓固有神经元网络的传输、整合和处理,支配对抗肌群中的交替发放活动,导致震颤运动产生的神经模型。综上,本论文建立了一个基于模型仿真和实验运动分析的神经运动控制研究平台,并应用于研究生理和病理状态下,中枢脊髓神经网络在运动控制中的作用。研究结果为认识神经运动控制开辟了新思路和新方法,也对运动康复治疗和临床前评估具有理论指导意义。
[Abstract]:The study of physiological and pathological exercise control is an important content of understanding the structure and physiological functions of the sensory motor system, especially the central nervous system of the brain spinal cord. It is also the basis of providing exercise function compensation and rehabilitation for the patients with motor dysfunction. The research method of neural motion control is divided into experimental observation analysis and calculation model simulation The two: the experimental study of the identification of all levels of the loop in the sensorimotor system and the understanding of the behavior function is the basis for the establishment of the computational model, and the simulation results of the computational model can provide the neural information of the state of the system that can not be observed under the experimental conditions. Therefore, the experimental analysis and model simulation methods are aware of the perception of the central nervous system. The scientific problem discussed in this paper is that, under the theoretical framework of dual control of attitude and motion, each nerve loop in the brain and spinal cord is integrated with sensory information to control the modularized function in the process of motion execution. The work of this thesis breaks through the bottleneck of the present research method and the sensory system will be used. A large number of data and known information are integrated into a multi-scale computing model, and the neural control mechanism of physiological and pathological movements is studied from the system level. On the basis of earlier work, this paper improves the simulation platform of multi scale virtual arm model, and verifies the authenticity and effectiveness of the virtual arm model. The virtual arm model is used to analyze and propose the hypothetical central nonlinear coding hypothesis of the muscle spindle gamma nerve innervation, to build an experimental collection and analysis platform for the human upper limb movement experiment, and to analyze the correlation between the synergistic distribution of the muscle group and the amplitude of the tremor in the limb tremor of Parkinson's disease, and to further explore the God of the tremor of Parkinson's disease by using the virtual arm model. The main research contents and results of this paper include the following four aspects: (1.) developing a multi scale integrated virtual arm model on the basis of earlier work. We have developed a biomechanical model of the upper limb skeletal system, virtual muscle, muscle spindle, and Golgi tendon organ, and new new models. The spinal reflex loop and the spinal intrinsic neuron network model module are integrated and transplanted into the SIMULINK modeling and simulation platform. The parameters are optimized and the interface is integrated into a virtual arm system model with real physiological characteristics. We adopt a modular model structure to make it easy to update and replace the module, and have strong expansibility. According to the characteristics of customization and individuation, the model has great potential in simulating normal and pathological neural motion behavior, studying the mechanism of neural motion control, and providing a reliable and efficient simulation platform and environment.2. for the development of rehabilitation application and preclinical evaluation to verify the true neural behavior of the virtual arm model. The contribution of feedforward and feedback control to the stability of the upper limb endpoint attitude is quantitatively evaluated. The hand stiffness reflects the impedance characteristics of the hand, and the motion variability of the hand describes the stability of the hand under the influence of the intrinsic neural noise. These two indicators can quantitatively evaluate the stability of the hand under the attitude control. The neural behavior of the system model simulation (hand stiffness and variability) is compared with the experimental data of the human body to confirm the authenticity and validity of the virtual arm model. Because of the limitations of the experimental measurement methods, we can not accurately measure the contribution of the spinal reflex loop to the posture stability of the upper extremity at the present stage. In different motion control theories, there are different interpretations of the role of the proprioceptive feedback in motion control. The virtual arm model enables us to quantitatively evaluate the effect of closed loop sensory feedback on the resistance and stability of the system by designing the simulation experiments under open and closed loop conditions. Under the action of closed loop sensory feedback, the area of the hand stiffness ellipse increases by 35.75 + 16.99% (mean mean deviation, the same below), and the area of the hand variable ellipse is reduced by 49.41 + 21.19%. It shows that the noumenon sensory feedback in the closed loop attitude control of the spinal cord can resist the noise of the internal nerve and the disturbance of the external environment and maintain the stability of the multiple joints. .3. is a significant contribution to the study of the possible joint angle coding pattern in the muscle spindle motor control and the influence of the joint angle encoding in the peripheral proprioceptive afferent. Experimental evidence indicates that the downward movement of the muscle spindle motor nerve is related to the trajectory of the joint movement, and the joint angle information is coded as a central peripheral posture in some form. A reference signal of state control. At the same time, the muscle spindle motor plays the role of maintaining the muscle fiber tension of the muscle spindle in the different posture, regulating the sensitivity of the sensory endings to control the linear coding of the joint angle in the frequency of the proprioception. We do not understand the coding mode of the joint angle in the muscle spindle motor nerve. Based on the virtual arm model simulation experiment platform, we have proposed and tested three possible coding mode assumptions of the muscle spindle motion control. The correctness of each hypothesis is verified by the hypothesis of the assumed muscle spindle motor neural coding mode. The results show that in attitude control, the neural system may take a nonlinear control strategy for the sensitivity of the muscle spindle in the attitude control, which makes the muscle spindle motor function.4. to study the intermuscular synchronization activity in the central peripheral peripheral system to transmit the joint angle information. The contribution of the Parkinson tremor movement of the upper limb and the basis of the neural circuit of the brain and spinal cord produced by the upper limb. A method of experimental acquisition of Parkinson's tremor movement and electromyography is established. The rhythmic activities of the muscles in Parkinson's tremor are analyzed by means of coherence and cross correlation, and "paired coherence" and "pool-averaged coherence" are proposed. The two evaluation indexes quantified the synchronous level of inter group activity and analyzed its correlation with the motion amplitude of joint tremor. The analysis showed that the synchronization level of the tremor activity of the upper limbs of the different patients was different, and was closely related to the amplitude of the tremor of the upper limb joints. The synchronization between the muscle groups is an important factor in driving the Parkinson tremor movement. This study provides support for the latest "modulation switching model" hypothesis to explain the mechanism of Parkinson's tremor movement, and further proposes the central oscillation signals derived from the cerebellum and basal ganglia system, through the transmission of the neural network of the spinal cord and the integration of the spinal cord. In this paper, a research platform for neural motion control based on model simulation and experimental motion analysis is established, and the role of the middle spinal cord neural network in motion control is studied under physiological and pathological conditions. The results have opened up new ideas and new methods for understanding neural motor control, and also have theoretical guiding significance for exercise rehabilitation and pre clinical assessment.

【学位授予单位】：上海交通大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：R742.5

【相似文献】