Dynamic functional MRI markers of drowsiness during sleep onset period

Figures

• 

  Figure 1. Distribution of the percentage of time spent in each drowsiness state (A) Percentage of overall time spent by all participants in each drowsiness state. (B) Percentage of time spent by each participant in the drowsiness states.

• 

  Figure 2. Surface rendering of power spectral density (PSD) of BOLD oscillations within region of interests in the atlas AAL3 Variations in PSD, across frequency bands ranging from 0 to 0.1 Hz (rows) and across drowsiness states (columns), are depicted. The red box highlights the frequency band with the highest PSD. Each drowsiness state column contains two sub-columns. The sub-columns represent left-brain hemisphere with the outer and inner faces at the right and left sides of the figure, respectively.

• 

  Figure 3. Surface rendering of the results from post-hoc comparisons of the power spectral density of BOLD signal oscillations at 0.05 Hz (PSD0.05), between different drowsiness states within each ROI in the AAL3 atlas Effect sizes, calculated using Cohen’s d, are depicted in regions where the difference between two states was statistically significant after Bonferroni correction (p < 0.00005, as 984 comparisons were computed). Outer and inner faces of the left-brain hemisphere are represented.

• 

  Figure 4. Power spectral density of the oscillations of BOLD signal at 0.05 Hz (PSD0.05), for each network in the YEO atlas Bar chart depicting PSD0.05 of each drowsiness state within each network. Data are presented as mean ± SEM. Networks are identified by colors illustrated in the legend at the top right, as well as in the surface rendering below it, and drowsiness states by colors depicted in the legend at the top left corner within the bar chart. ∗p < 0.05; ∗∗p < 0.01 and ∗∗∗p < 0.001 for Bonferroni-corrected pairwise comparisons after a linear mixed model.