From the Department of Neurology (D.G., R.I.) Department of Anesthesiology (A.R., G.S., D.J.), Klinikum rechts der Isar, Technical University Munich, München, Germany GIGA-Consciousness, Coma Science Group (S.K.L., C.D.P., S.L.) GIGA-Consciousness, Sensation and Perception Research Group (A.V., V.B.) GIGA Research, University, and Department of Algology and Palliative Care, Department of Neurology (S.L.) Department of Anesthesia and Intensive Care Medicine (V.B.) CHU University Hospital of Liège (C.D.P.), Liège, Belgium GIGA-Cyclotron Research Center: In Vivo Imaging, University of Liège, Liège, Belgium (A.P.) University Department of Anesthesia and Intensive Care Medicine, CHR Citadelle, Liège, Belgium (V.B.) Department of Neurology, University of Wisconsin, Madison, Wisconsin (M.B.) Asklepios Clinic, Department of Neurology, Bad Tölz, Germany (R.I.).
Anesthesiology. 2019 Jun;130(6):898-911. doi: 10.1097/ALN.0000000000002704.
A key feature of the human brain is its capability to adapt flexibly to changing external stimuli. This capability can be eliminated by general anesthesia, a state characterized by unresponsiveness, amnesia, and (most likely) unconsciousness. Previous studies demonstrated decreased connectivity within the thalamus, frontoparietal, and default mode networks during general anesthesia. We hypothesized that these alterations within specific brain networks lead to a change of communication between networks and their temporal dynamics.
We conducted a pooled spatial independent component analysis of resting-state functional magnetic resonance imaging data obtained from 16 volunteers during propofol and 14 volunteers during sevoflurane general anesthesia that have been previously published. Similar to previous studies, mean z-scores of the resulting spatial maps served as a measure of the activity within a network. Additionally, correlations of associated time courses served as a measure of the connectivity between networks. To analyze the temporal dynamics of between-network connectivity, we computed the correlation matrices during sliding windows of 1 min and applied k-means clustering to the matrices during both general anesthesia and wakefulness.
Within-network activity was decreased in the default mode, attentional, and salience networks during general anesthesia (P < 0.001, range of median changes: -0.34, -0.13). Average between-network connectivity was reduced during general anesthesia (P < 0.001, median change: -0.031). Distinct between-network connectivity patterns for both wakefulness and general anesthesia were observed irrespective of the anesthetic agent (P < 0.001), and there were fewer transitions in between-network connectivity patterns during general anesthesia (P < 0.001, median number of transitions during wakefulness: 4 and during general anesthesia: 0).
These results suggest that (1) higher-order brain regions play a crucial role in the generation of specific between-network connectivity patterns and their dynamics, and (2) the capability to interact with external stimuli is represented by complex between-network connectivity patterns.
人类大脑的一个关键特征是能够灵活地适应不断变化的外部刺激。这种能力可以通过全身麻醉来消除，全身麻醉的状态表现为无反应、失忆和（很可能）无意识。先前的研究表明，在全身麻醉期间，丘脑、额顶叶和默认模式网络内的连接性降低。我们假设这些特定脑网络内的变化导致网络之间的通信和它们的时间动态发生变化。
我们对以前发表的 16 名志愿者在异丙酚全身麻醉和 14 名志愿者在七氟醚全身麻醉期间的静息态功能磁共振成像数据进行了合并空间独立成分分析。与以前的研究相似，得到的空间图谱的平均 z 分数作为网络内活动的测量指标。此外，相关的时间过程的相关性作为网络之间连接性的测量指标。为了分析网络间连接的时间动态，我们在 1 分钟的滑动窗口中计算相关矩阵，并在全身麻醉和清醒状态下对矩阵应用 k-均值聚类。
在全身麻醉期间，默认模式、注意力和突显网络内的活动减少（P < 0.001，中位数变化范围：-0.34，-0.13）。全身麻醉期间平均网络间连接性降低（P < 0.001，中位数变化：-0.031）。在全身麻醉和清醒状态下观察到了不同的网络间连接模式，而与麻醉药物无关（P < 0.001），并且在全身麻醉期间网络间连接模式的转换较少（P < 0.001，清醒状态时的中位数转换次数为 4 次，全身麻醉时为 0 次）。
这些结果表明，（1）高级脑区在产生特定的网络间连接模式及其动力学方面起着关键作用，（2）与外部刺激相互作用的能力由复杂的网络间连接模式来表示。
