Associating EEG functional networks and the effect of sleep deprivation as measured using psychomotor vigilance tests

Fig. 1. CR Protocol. Raster plot depicting the laboratory protocol across the 2 baseline days (B1, B2), the 2 days of CR (CR1, CR2) and 2 recovery days (R1, R2)—nominal 08:00 waketime and 23:00 DLMO as denoted by the red triangle. AD, admissions; D/C, discharge; circle EEG acquisition (120 s); circle and cross, first EEG acquisition after awakening (120 s); circle and plus, EEG acquisition at \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\text {DLMO}-11\text { h}$$\end{document} (120 s)—here DLMO was set at 23:00, as denoted by the red triangle, to provide an example of the first acquisition after awakening differing from the EEG acquisition at \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\text {DLMO}-11\text { h}$$\end{document}.

Fig. 2. Epoch selection. (a) A single 30 s epoch selected using the Resting State Epoch Selector Algorithm, which was used for calculating a FN and the graph metrics mean node strength, characteristic path length and clustering coefficient. (b) The sliding window approach used to select 30 s epochs with a 15 s overlap. For each epoch a FN is calculated and the median 2-D correlation between each pair of FNs was taken as representative of stability.

Fig. 3. Psychomotor vigilance test performance split into groups based on high and low impairment. Timeseries showing (a) mean RT, (b) median RT, (c) standard deviation of RT and (d) the percentage of lapses when the participants are split into two groups based on the level of impairment in that measure. The solid blue line () and dashed red line () are the mean across the 6 lowest and 6 highest impaired participants, respectively. The corresponding blue and red shaded area is the standard error of the mean. The dotted vertical lines indicate the times considered to be in the WMZ (3 h before DLMO to 5 min after). The significance of permutation tests (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$n=10,000$$\end{document}) comparing the mean value of the low and high impairment groups over adjacent 12-h periods depicted by different shades of grey are represented by, *\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy}…

Fig. 4. Significant relationships between PVT impairment and baseline graph metrics. Scatterplots relating the PVT impairment as measured using std of RT to (a) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\text {B}_\text {FA}$$\end{document} mean node strength, (b) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\text {B}_\text {-11}$$\end{document} mean node strength, (c) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\text {B}_\text {FA}$$\end{document} characteristic path length, (d) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs}…

Fig. 5. Pearson’s correlation between B\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\mathbf {_{FA}}$$\end{document} graph metrics and standard deviation of RT. How the Pearson’s correlation coefficient between baseline (a) mean node strength, (b) clustering coefficient, (c) characteristic path length and (d) stability for the first acquisition after awakening and the PVT performance as measured using std of RT. In all plots the ☆ indicates where the p-value associated with the correlation is significant \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$p<0.05$$\end{document} (uncorrected) and the dotted vertical lines indicate the times considered to be in the WMZ (3 h before DLMO to 5 min after) and the relative time awake provides an indication of the time awake for the majority of the participants.

Fig. 6. Stability of participant’s FNs. The stability across all participants for a given frequency band (a) Delta, (b) Theta, (c) Alpha and (d) Beta. The solid blue line is the mean across the participants stability and the shaded area is the standard error of the mean. In all plots the dotted vertical lines indicate the times considered to be in the WMZ (3 h before DLMO to 5 min after) and the relative time awake provides an indication of the time awake for the majority of the participants.

Fig. 7. Stability of participant’s FNs with low and high impairment. The stability across the participants with the lowest PVT impairment and the highest PVT impairment as measured using the percentage increase in standard deviation of RT. The solid blue line () is the mean stability of the 6 participants with the lowest impairment and the dashed red line () is the mean stability of the 6 participants with the highest impairment. The corresponding blue and red shaded area is the standard error of the mean. Results are given for (a) Delta, (b) Theta, (c) Alpha and (d) Beta frequency bands. The significance of permutation tests (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$n=10,000$$\end{document}) comparing the mean value of the low and high impairment groups over adjacent 12 h periods, depicted by different shades of grey, are represented by *\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs}…