标题： 平衡兴奋性抑制尖峰网络中由涨落引起的不同节律之间的瞬时转换 显示英文标题

标题： Fluctuation induced intermittent transitions between distinct rhythms in balanced excitatory-inhibitory spiking networks

摘要： 在睡眠过程中，与临界动力学相关并表现出幂律分布的间歇性转变是常见的。 这些临界行为在微观层面通过神经元级联显现，在宏观层面通过睡眠阶段之间的转变显现。 为了阐明这些经验观察结果，已提出了基于统计物理的模型。 在皮层活动的介观层面，临界行为表现为各种皮层节律之间的间歇性转变。 例如，利用大鼠脑电图数据的经验研究已经识别出$\delta$和$\theta$节律之间的间歇性转变，其中$\theta$节律的持续时间表现出幂律分布。 然而，目前尚缺乏解释这一现象的动力学模型。 在本研究中，我们引入了一个稀疏耦合的兴奋性和抑制性二次积分-发放（QIF）神经元群体网络，以表明间歇性转变可以从有限尺寸系统的内在波动中产生，特别是在系统处于霍普夫分岔点附近时，这是一个临界点。 所得出的幂律分布和指数与经验观察结果一致。 此外，我们展示了网络连接性的修改如何通过影响稳定极限环的吸引力和振荡频率来影响幂律指数。 我们的研究结果通过神经元网络的基本动力学进行解释，为皮层节律之间间歇性转变的产生提供了一个合理的机制，与经验研究中记录的幂律分布相一致。 显示英文摘要

摘要： Intermittent transitions, associated with critical dynamics and characterized by power-law distributions, are commonly observed during sleep. These critical behaviors are evident at the microscopic level through neuronal avalanches and at the macroscopic level through transitions between sleep stages. To clarify these empirical observations, models grounded in statistical physics have been proposed. At the mesoscopic level of cortical activity, critical behavior is indicated by the intermittent transitions between various cortical rhythms. For instance, empirical investigations utilizing EEG data from rats have identified intermittent transitions between $\delta$ and $\theta$ rhythms, with the duration of $\theta$ rhythm exhibiting a power-law distribution. However, a dynamic model to account for this phenomenon is currently absent. In this study, we introduce a network of sparsely coupled excitatory and inhibitory populations of quadratic integrate-and-fire (QIF) neurons to demonstrate that intermittent transitions can emerge from the intrinsic fluctuations of a finite-sized system, particularly when the system is positioned near a Hopf bifurcation point, which is a critical point. The resulting power-law distributions and exponents are consistent with empirical observations. Additionally, we illustrate how modifications in network connectivity can affect the power-law exponent by influencing the attractivity and oscillation frequency of the stable limit cycle. Our findings, interpreted through the fundamental dynamics of neuronal networks, provide a plausible mechanism for the generation of intermittent transitions between cortical rhythms, in alignment with the power-law distributions documented in empirical researches.