脑电非周期活动与睡眠及警觉性变化的研究进展

A Review of Aperiodic EEG Activity in Relation to Changes in Sleep and Vigilance

摘要: 脑电功率谱中的非周期成分描述了跨频率的背景变化。本文采用叙述性综述方法，梳理自然睡眠、睡眠剥夺及清醒期警觉表现的相关研究，并讨论非周期参数的生理解释与方法学限制。已有研究发现，非周期特征随睡眠阶段、REM亚状态和恢复睡眠进程变化；清醒期睡眠剥夺后的偏移量变化与主观警觉性有关，但其与客观行为表现的关系尚不明确。这些差异受到拟合频段、模型与记录条件的影响。兴奋–抑制平衡、膜与树突滤波以及神经活动的特征时间尺度，均可能参与形成背景谱，单一参数尚不能区分这些机制。单斜率模型适于描述近似线性的局部谱形，膝点模型则容许特征频率的存在，但拟合形式本身不构成生理机制的证明。今后需结合干预、多尺度记录和独立样本验证，检验非周期指标能否在传统脑电特征之外提供可靠的警觉性信息。

Abstract: The aperiodic component of the EEG power spectrum describes background variation across frequencies. This narrative review examines evidence from natural sleep, sleep deprivation, and waking vigilance, and discusses the physiological interpretation and methodological limitations of aperiodic parameters. Aperiodic features vary with sleep stages, REM substates, and the progression of recovery sleep. During wakefulness after sleep deprivation, offset changes have been associated with subjective alertness, whereas their relationship with objective performance remains unclear. These findings depend on frequency range, model specification, and recording conditions. Excitation-inhibition balance, membrane and dendritic filtering, and characteristic neural timescales may all contribute to background spectra; individual parameters cannot yet distinguish these mechanisms. Fixed models describe approximately linear spectral regions, whereas knee models allow characteristic frequencies, but neither model alone establishes a physiological mechanism. Intervention studies, multiscale recordings, and independent validation are needed to determine whether aperiodic measures provide reliable vigilance information beyond conventional EEG features.