Department of Biosciences, University of Helsinki, Finland.
Neuroimage. 2012 Apr 2;60(2):1562-73. doi: 10.1016/j.neuroimage.2011.12.080. Epub 2012 Jan 8.
Development of neuronal connections relies on proper neuronal activity, and it starts during the time when early preterm babies are treated in the neonatal intensive care units. While synchrony has been a key element in visual assessment of neonatal EEG signals, there has been no unambiguous definitions for synchrony, and no objective measures available for neonatal signals. Estimation of phase locking value (PLV) has been an established paradigm in adults, but many unique characteristics of the neonatal EEG have precluded its applicability in them. In the present paper, we developed the existing PLV-based methods further to be applicable for neonatal signals at two different temporal scales, oscillations and events, where the latter refers technically to quantitating phase synchrony (PS) between band-specific amplitude envelopes (bafPS). In addition, we present a measure for quantitation based on assessing cumulative proportion of time with statistically significant synchrony between the given signal pair. The paper uses real EEG examples and the prior neurobiological knowledge in the process of defining optimal parameters in each step of the procedure. Finally, we apply the method to a set of dense array EEG recordings from very early preterm babies, recorded at conceptional age of less than 30 weeks. By comparing PS and bafPS from babies without and with major cerebrovascular lesion, we show that the effects of brain lesions may be selective both in space and in frequency. These findings do by nature escape visual detection in the conventional EEG reading, however they have intriguing correlates in the current concept of how somatosensory networks are thought to develop and/or become disorganized in the early preterm babies.
神经元连接的发育依赖于适当的神经元活动,而这一过程始于早期早产儿在新生儿重症监护病房接受治疗的时期。尽管同步性一直是新生儿脑电图信号视觉评估的关键要素,但目前还没有明确的同步性定义,也没有针对新生儿信号的客观测量方法。相位锁定值(PLV)的估计一直是成人研究中的一种既定范例,但新生儿脑电图的许多独特特征使其无法在其中应用。在本文中,我们进一步开发了现有的基于 PLV 的方法,使其能够适用于两个不同时间尺度的新生儿信号,即振荡和事件,后者在技术上是指量化特定频段幅度包络(bafPS)之间的相位同步(PS)。此外,我们还提出了一种基于评估具有统计学意义的同步性的时间累积比例的定量测量方法。本文在定义该方法的每个步骤的最优参数的过程中使用了真实的脑电图示例和先验的神经生物学知识。最后,我们将该方法应用于一组来自极早期早产儿的密集阵列脑电图记录,记录时间在妊娠 30 周之前。通过比较无和有重大脑血管病变的婴儿的 PS 和 bafPS,我们表明大脑病变的影响可能在空间和频率上都具有选择性。这些发现本质上逃避了传统脑电图阅读中的视觉检测,但它们与当前关于体感网络如何在早期早产儿中发育和/或变得紊乱的概念有有趣的相关性。