基于兴奋性与抑制性反馈协同调控的神经临界态机制及其信息处理特性研究

The information processing characteristics of brain is optimal at criticality, which implies that neural criticality is an important information processing mechanism in brain. Up to now, most researches of neural criticality are conducted using complex neural network model, which indicated that synaptic plasticity can cause the criticality in neural system. Excitatory and inhibitory feedback is the typical connection in cortical neural network, and plays a key role in modulating the dynamics and function of neural system. On the context, we will study the neural criticality mechanism underlined by excitatory and inhibitory feedback at the scale of neural circuit, and analyze the information processing characteristics at criticality. Firstly, we will build the relationship between excitatory and inhibitory feedback to drive the system to its criticality. Secondly, we will analyze the information capacity and information transmission characteristics at criticality, respectively. Finally, we will provide explain on how the interaction between excitatory and inhibitory feedback drive the system to its criticality. Through the above mentioned researches, we aim to elucidate the criticality mechanism underlined by excitatory and inhibitory feedback, to provide a guideline for understanding the criticality mechanism of complex neural network, to further promote the development of the theory of neural criticality, to provide an insight into how the excitatory and inhibitory feedback exert influence on the information processing characteristics within the criticality-theoretic framework.

大脑在临界态时具有最优的信息处理性能，神经临界态被认为是潜在的神经信息处理机制。已有研究限于神经网络层面，表明突触可塑性是神经临界态产生的重要机制。兴奋性与抑制性反馈连接是脑皮层神经网络的典型连接模式，对神经状态与功能特性具有重要的调控作用。据此在神经回路层面研究兴奋性与抑制性反馈对神经临界态的协同调控机制及临界态时的信息处理特性。包括：①建立临界态时兴奋性与抑制性反馈应满足的定量关系，即临界态产生条件；②兴奋性与抑制性反馈如何协同调控使系统趋向临界态？即临界态产生机制③分析临界态时的信息容量/编码能力及信息传输特性。 意义：①揭示基于兴奋性与抑制性反馈协同调控的神经临界态新机制。②为理解更高层次神经网络临界态机制及信息处理特性提供基础。③通过临界态概念将反馈结构与神经信息处理特性予以有机联系，有利于深入剖析反馈对神经信息处理特性的调控规律。④推动神经临界态理论与应用的进一步发展与完善。

大脑在临界态时具有最优的信息处理性能，神经临界态被认为是潜在的神经信息处理机制。兴奋性与抑制性反馈连接是脑皮层神经网络的典型连接模式，对神经状态与功能特性具有重要的调控作用，据此研究兴奋性与抑制性反馈对神经临界态的协同调控机制及临界态时的信息处理特性。首先，构建神经群尺度的反馈神经回路模型，分别应用分岔分析及描述函数方法确定临界态时兴奋性与抑制性反馈应满足的定量关系，揭示了基于兴奋性与抑制性反馈协同调控的临界态机制。其次，构建神经元尺度的反馈神经回路模型，并在模型中引入抑制性突触可塑性，研究了基于突触可塑性的兴奋性与抑制性反馈平衡机制，表明在兴奋性与抑制性平衡时反馈神经回路居于临界态，揭示了兴奋性与抑制性反馈对神经临界态的协同调控规律；并进一步研究了信息传输特性在兴奋性与抑制性反馈二维平面的分布，据此给出了兴奋性与抑制性反馈对信息传输的调控规律，表明在临界态时神经回路取得最优的信息传输性能。最后，应用临界稳定的概念，分别研究了基于比例积分和比例微分的癫痫电活动的闭环调控方案，在控制器参数域给出临界稳定界的图解确定方法，为癫痫电活动闭环调控参数的选取提供依据，仿真分析表明了方案的可行性和有效性。本研究揭示了基于兴奋性与抑制性反馈协同调控的神经临界态机制及信息处理特性，将推动神经临界态理论与应用的进一步发展与完善。