Changes in Whole Brain Dynamics and Connectivity Patterns during Sevoflurane- and Propofol-induced Unconsciousness Identified by Functional Magnetic Resonance Imaging
Daniel Golkowski,Stephen Karl Larroque,Audrey Vanhaudenhuyse,Alain Plenevaux,Melanie Boly,Carol Di Perri,Andreas Ranft,Gerhard Schneider,Steven Laureys,Denis Jordan,Vincent Bonhomme,Rüdiger Ilg
DOI: https://doi.org/10.1097/ALN.0000000000002704
IF: 8.986
Anesthesiology
Abstract:Background: 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. Methods: 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. Results: 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). Conclusions: 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.