QuantSLEEP: A numerical toolbox for a graph signal processing of the rhythmic and arrhythmic components of sleep EEG recordings.

人的生命有三分之一是在睡觉上度过的。睡眠远非休息状态，而是一个特别重要的状态，在这个状态中，大脑产生和处理信息，并经历突触连接的可塑性重组，以便在即将到来的清醒时间更有效。在信号处理方面，睡眠是一个原始的“实验室”，在这个实验室里，大脑自发活动的电测量给出了大脑在清醒的前一段时间内应对和应对不利威胁的能力的清晰信号。今天，我们对这一活动的理解主要是各种过程之间的相互作用序列。除了可能的振荡爆发耦合之外，大脑生物电活动还显示了短期功能网络之间的相互作用，这些功能网络组织了参与神经元过程的共同激活的远程大脑区域。我们的研究计划将侧重于发展数值方法，用于集成先进的信号处理，具体到睡眠大脑的脑电图（EEG）测量。更具体地说，这项研究计划旨在开发一种创新的软件，以提供睡眠期间涉及的网络之间的时间分辨相互作用。网络将从脑电图信号的不同组成部分获得，从在心律失常组成部分的背景下观察到的特征瞬态节律振荡，到最近发生的“微观状态”事件，这些事件恰好是远程大脑区域稳定共同激活的样本。该提案的主要成果是一个数值工具，该工具集成了解码脑电图信号的创新方法，以便提供与“睡眠大脑”基本生物标志物相关的功能网络的清单和动态相互作用。这种对睡眠脑电图记录的创新描述将在各种情况下进行评估，从儿童时期大脑的早期成熟到癫痫和其他神经退行性疾病的临床应用。

A third of human life is spent sleeping. Far from being a resting state, sleep is a particularly important state during which the brain generates and processes information, and undergoes a plastic reorganization of synaptic connections to be more efficient during the forthcoming awake time. In terms of signal processing, sleep is a pristine `laboratory' in which electric measurements of spontaneous brain activity give a clear signature of the ability of the brain to cope with, and to respond to adverse threats during the previous periods of wakefulness. Today, our understanding of this activity is mostly a matter of sequences of interactions between various processes. In addition to the possible coupling of bursts of oscillations, the brain bioelectrical activity also shows interactions between short-lived functional networks that organize the co-activated remote brain areas involved in the neuronal processes. Our research program will focus on the development of numerical methods for the integration of advanced signal processing specific to electro-encephalography (EEG) measurements of the sleeping brain. More specifically, this research program aims at developing an innovative software to provide the time-resolved interactions between the networks involved during sleep. The networks will be obtained from different components of the EEG signals, from the characteristic transient rhythmic oscillations observable in the background of an arrhythmic component, to the more recent `microstates' events that happen to sample the stable co-activation of remote brain regions. The main outcome of this proposal is a numerical tool that integrates innovative approaches to decode EEG signals in order to provide the inventory and the dynamical interactions of the functional networks associated with the fundamental biomarkers of the `sleeping brain'. This innovative description of the sleep EEG recordings will be evaluated in various contexts, from the early maturation of the brain during childhood to clinical applications in epilepsy and other neuro-degenerative diseases.