基于闭环控制的神经元及集群癫痫状态的研究

发布时间：2018-04-19 12:00

本文选题：癫痫 + 闭环控制　；参考：《天津大学》2014年硕士论文

【摘要】：癫痫是大脑神经元突发性异常放电导致短暂的大脑功能障碍的一种慢性神经系统疾病,针对药物无法控制的难治性癫痫,电刺激正逐渐成为其有效的治疗手段之一,如深度脑刺激(DBS)、经颅直流电刺激(tDCS)等。目前临床使用的开环电刺激方法并不具有普遍适用性,为解决这个问题,通过闭环电刺激来治疗癫痫等精神疾病已成为目前研究的热点。由于癫痫疾病的致病机理尚不清楚加之电生理实验方面的各种限制,神经计算模型分析已成为研究神经系统动力学特性和生理现象的有效方法之一。因此,基于单神经元和神经元集群模型,本文提出了闭环迭代学习控制(ILC)策略与闭环混合控制策略,以实现癫痫态放电的闭环控制。本论文的研究内容主要包括以下三个方面:首先,本文基于海马CA3区Pinsky-Rinzel(PR)模型和大脑皮层神经元集群模型(NMM),分析其动力学特性,考虑单神经元关键参数—钾离子通道反电势V_K、胞体与树突之间的耦合电导cg、化学耦合强度g_(NMDA)、g_(AMPA)对神经元放电模式的影响;神经元集群参数—兴奋性平均突触增益A、耦合强度K_(ij)对神经元集群放电的影响,探索癫痫发病机制。其次,基于关键参数对单神经元放电模式的影响,利用迭代学习控制、延时反馈与PI控制相结合的混合控制分别实现了单神经元癫痫样放电的闭环控制;采用无迹卡尔曼滤波器(UKF)对关键参数进行估计,进而形成反馈信号。通过对关键参数的跟踪反馈调节实现PR神经元放电模式的转换;基于化学耦合强度的影响,利用迭代学习控制实现不同耦合强度的神经元间放电去同步控制。最后,基于多耦合神经元集总参数模型提出多模式闭环深度脑刺激系统。依据兴奋性平均突触增益与耦合强度的影响,产生癫痫态与正常态下的局部场电位(LFP)波形;根据对C_0复杂度、信号频带能量比的分析,提取局部场电位的特征;采用BP神经网络对局部场电位信号进行分类并实现刺激强度的选择;针对不同的放电模式利用迭代学习控制器输出控制信号;通过对刺激电流幅值和周期的调制输出精准的深度脑刺激电流,实现癫痫状态的闭环控制,阻止癫痫样放电的进一步传播。以上的研究结果均证明本文提出的闭环控制策略在控制癫痫发作状态上的有效性,为单神经元及神经元集群闭环电生理实验,以及癫痫电刺激治疗器的研究提供了理论依据。
[Abstract]:Epilepsy is a kind of chronic nervous system disease, which is caused by sudden abnormal discharge of brain neurons. Electrical stimulation is becoming one of the effective treatment methods for intractable epilepsy, which can not be controlled by drugs.For example, deep brain stimulation, transcranial direct current stimulation (TDC), and so on.In order to solve this problem, closed loop electrical stimulation has become a hot research topic in the treatment of mental disorders such as epilepsy.As the pathogenesis of epilepsy is not clear and various limitations in electrophysiological experiments, neural computing model analysis has become one of the effective methods to study the dynamic characteristics and physiological phenomena of nervous system.Therefore, based on the single neuron and neuron cluster model, a closed loop iterative learning control (ILC) strategy and a closed loop hybrid control strategy are proposed to realize the closed loop control of epileptic state discharge.The main contents of this thesis include the following three aspects: firstly, based on the Pinsky-Rinzelpra model of hippocampal CA3 region and the cerebral cortex neuron cluster model, the dynamic characteristics of the model were analyzed.Consider the effect of the key parameter of single neuron VK, the coupling conductance between the cell body and the dendrite, the chemical coupling intensity, the chemical coupling intensity, on the firing pattern of the neurons.The effect of neuronal cluster parameters-excitatory average synaptic gain A, coupling intensity K _ (+ +) on the discharge of neuronal cluster and the pathogenesis of epilepsy were explored.Secondly, based on the influence of key parameters on the single neuron discharge mode, the closed-loop control of single neuron epileptoid discharge is realized by the hybrid control of iterative learning control, delay feedback and Pi control.Unscented Kalman filter (UKF) is used to estimate the key parameters to form feedback signals.Based on the influence of chemical coupling intensity, iterative learning control is used to realize the de-synchronous control of firing between neurons with different coupling intensity.Finally, a closed loop deep brain stimulation system is proposed based on the multi-coupled neuron lumped parameter model.According to the effect of excitatory average synaptic gain and coupling intensity, the local field potential (LFP) waveform was generated in epileptic state and normal state, and the characteristic of local field potential was extracted according to the complexity of CSP 0 and the analysis of signal frequency band energy ratio.BP neural network is used to classify the local field potential signals and to select the stimulus intensity, and the iterative learning controller is used to output the control signals for different discharge modes.By modulating the amplitude of the stimulus current and the period of the stimulus, the amplitude of the stimulation current and the modulation of the cycle output the accurate deep brain stimulation current to realize the closed-loop control of the epileptic state and to prevent the further spread of the epileptoid discharge.The above results all prove the effectiveness of the closed-loop control strategy proposed in this paper in the control of epileptic seizures. It provides a theoretical basis for the research of single neuron and neuron cluster closed loop electrophysiological experiment and the therapeutic device of epileptic electrical stimulation.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：R742.1

【相似文献】