Functional Brain Networks and EEG Microstate Dynamics During Light Sleep: An EEG-fMRI Study
Poster No:
2342
Submission Type:
Abstract Submission
Authors:
Yujie Long1, Jing Xu1, Shuqin Zhou2, Guangyuan Zou3,4, Jiayi Liu3,4, Qihong Zou3, Jia-hong Gao3,4,5,1
Institutions:
1Shanghai International Studies University, Shanghai, China, 2McLean Imaging center, McLean Hospital, Harvard Medical School, Belmont, MA 02478, USA, Boston, MA, 3Center for MRI Research, Peking University, Beijing, China, 4Beijing City Key Lab for Medical Physics and Engineering, Institution of Heavy Ion Physics, School of Physics, Peking University, Beijing, China, 5McGovern Institute for Brain Research, Peking University, Beijing, China
First Author:
Co-Author(s):
Shuqin Zhou
McLean Imaging center, McLean Hospital, Harvard Medical School, Belmont, MA 02478, USA
Boston, MA
McLean Imaging center, McLean Hospital, Harvard Medical School, Belmont, MA 02478, USA
Boston, MA
Guangyuan Zou
Center for MRI Research, Peking University|Beijing City Key Lab for Medical Physics and Engineering, Institution of Heavy Ion Physics, School of Physics, Peking University
Beijing, China|Beijing, China
Center for MRI Research, Peking University|Beijing City Key Lab for Medical Physics and Engineering, Institution of Heavy Ion Physics, School of Physics, Peking University
Beijing, China|Beijing, China
Jiayi Liu
Center for MRI Research, Peking University|Beijing City Key Lab for Medical Physics and Engineering, Institution of Heavy Ion Physics, School of Physics, Peking University
Beijing, China|Beijing, China
Center for MRI Research, Peking University|Beijing City Key Lab for Medical Physics and Engineering, Institution of Heavy Ion Physics, School of Physics, Peking University
Beijing, China|Beijing, China
Jia-hong Gao
Center for MRI Research, Peking University|Beijing City Key Lab for Medical Physics and Engineering, Institution of Heavy Ion Physics, School of Physics, Peking University|McGovern Institute for Brain Research, Peking University|Shanghai International Studies University
Beijing, China|Beijing, China|Beijing, China|Shanghai, China
Center for MRI Research, Peking University|Beijing City Key Lab for Medical Physics and Engineering, Institution of Heavy Ion Physics, School of Physics, Peking University|McGovern Institute for Brain Research, Peking University|Shanghai International Studies University
Beijing, China|Beijing, China|Beijing, China|Shanghai, China
Introduction:
Light sleep, encompassing non-rapid eye movement stages N1 and N2, is vital for cognitive function and physical health, acting as a bridge to deeper sleep phases. Characterized by distinct EEG patterns, these stages are crucial for understanding brain dynamics during sleep. Light sleep is composed of non-rapid eye movement stages including N1 and N2, characterized by distinct EEG patterns, with the latter stages serving as crucial transitions from wakefulness to deeper sleep phases. EEG microstates, brief quasi-stable periods of electric field topography, offer unique perspectives into these dynamic brain functions. In previous research, four consistent microstate maps (A-D), have been identified across wake and sleep stages. While recent EEG-fMRI advancements have identified their correlations with resting-state networks (RSNs), relationships during N1 and N2 remain unknown.
Methods:
In this study, simultaneous EEG-fMRI data were acquired during the N1 and N2 stages to investigate the relationships between EEG microstates and fMRI networks.
Stage-specific N1 and N2 template maps were applied to each subject's EEG data, enabling the computation of key microstate parameters including Mean Correlation, Global Explained Variance (GEV), Mean Duration, and Time Coverage. ANOVA with factors "stage" and "microstate" was performed on microstate parameters.
Custom regressors based on microstate maps were used in a General Linear Model (GLM) to model BOLD responses, involving three critical steps: 1. Aligning group-level microstate template maps with individual EEG data; 2. Convolving spatial correlation time series with the hemodynamic response function for GLM analysis; 3. The beta maps resulting from the GLM analysis were analyzed using AFNI's 3dLMEr (Chen et al., 2013) for group-level linear mixed-effect analysis, considering sleep stages, sessions, and subjects. Multiple comparison corrections were applied using the autocorrelation function (ACF) modeling approach, involving 3dFWHMx for ACF estimation and 3dClustSim for determining the cluster-size threshold (Cox, 1996; Eklund et al., 2016). This methodology ensured a corrected significance level of p < .05 with a minimum cluster size of 150 voxels. The analysis was replicated for four different EEG microstate maps.
Stage-specific N1 and N2 template maps were applied to each subject's EEG data, enabling the computation of key microstate parameters including Mean Correlation, Global Explained Variance (GEV), Mean Duration, and Time Coverage. ANOVA with factors "stage" and "microstate" was performed on microstate parameters.
Custom regressors based on microstate maps were used in a General Linear Model (GLM) to model BOLD responses, involving three critical steps: 1. Aligning group-level microstate template maps with individual EEG data; 2. Convolving spatial correlation time series with the hemodynamic response function for GLM analysis; 3. The beta maps resulting from the GLM analysis were analyzed using AFNI's 3dLMEr (Chen et al., 2013) for group-level linear mixed-effect analysis, considering sleep stages, sessions, and subjects. Multiple comparison corrections were applied using the autocorrelation function (ACF) modeling approach, involving 3dFWHMx for ACF estimation and 3dClustSim for determining the cluster-size threshold (Cox, 1996; Eklund et al., 2016). This methodology ensured a corrected significance level of p < .05 with a minimum cluster size of 150 voxels. The analysis was replicated for four different EEG microstate maps.
Results:
ANOVA analysis showed that significant main effects were observed for sleep stages on mean correlation (p < .05), and for microstates on all four dependent microstate parameters (p < .01). Additionally, significant interaction effects between sleep stages and microstates were noted for each dependent variable (p < .01) (see Fig. 1). Notably, microstate B demonstrated larger values of GEV, mean duration, and time coverage in N1, a trend that reversed in N2 (see Fig. 1b-d). In contrast, microstate C showed smaller values in N1, which increased in N2 (see Fig. 1b-d).
The main effects of sleep stage of beta values of microstate B and microstate D were found to be significant. Specifically, microstate B demonstrated a correlation with the right cerebellum, exhibiting activation during N1 and deactivation in N2, as detailed in post hoc analysis (see Fig. 2a). In the case of microstate D, correlations with the bilateral thalamus and the bilateral postcentral gyrus were observed. Interestingly, a transition from N1 to N2 in microstate D was characterized by a reversal in activation patterns: the thalamus shifted from deactivation to activation, while the motor network showed an opposite trend, moving from activation to deactivation (see Fig. 2b).
The main effects of sleep stage of beta values of microstate B and microstate D were found to be significant. Specifically, microstate B demonstrated a correlation with the right cerebellum, exhibiting activation during N1 and deactivation in N2, as detailed in post hoc analysis (see Fig. 2a). In the case of microstate D, correlations with the bilateral thalamus and the bilateral postcentral gyrus were observed. Interestingly, a transition from N1 to N2 in microstate D was characterized by a reversal in activation patterns: the thalamus shifted from deactivation to activation, while the motor network showed an opposite trend, moving from activation to deactivation (see Fig. 2b).
Conclusions:
In conclusion, our study sheds light on the relationships between EEG microstates and functional brain networks during the N1 and N2 stages of light sleep, establishing EEG microstates as crucial electrophysiological markers of these networks. Building upon prior research on thalamocortical network activity transitions from wakefulness to sleep, we postulate that microstates B and D represent EEG manifestations during early sleep stages.
Brain Stimulation:
Non-invasive Magnetic/TMS
Modeling and Analysis Methods:
EEG/MEG Modeling and Analysis
Novel Imaging Acquisition Methods:
BOLD fMRI 1
EEG
Perception, Attention and Motor Behavior:
Sleep and Wakefulness 2
Keywords:
Cortex
Data analysis
FUNCTIONAL MRI
Thalamus
1|2Indicates the priority used for review
Provide references using author date format
Cox, R. W. (1996). AFNI: Software for Analysis and Visualization of Functional Magnetic Resonance Neuroimages. Computers and Biomedical Research, 29(3), 162–173. https://doi.org/10.1006/cbmr.1996.0014
Eklund, A., Nichols, T. E., & Knutsson, H. (2016). Cluster failure: Why fMRI inferences for spatial extent have inflated false-positive rates. Proceedings of the National Academy of Sciences, 113(28), 7900–7905. https://doi.org/10.1073/pnas.1602413113