Brain Complex Network Characteristic Analysis of Fatigue during Simulated Driving Based on Electroencephalogram Signals

Figures

• 

  Figure 1. Simulated driving experiment interface.

• 

  Figure 2. The illustration of the simulated driving experiment procedure.

• 

  Figure 3. The 62 channel electrode positions of the brain.

• 

  Figure 4. The evolution of the complex network characteristics in the delta rhythm for one subject (T=0.74). (a) The functional connectivity graph at several specific times; (b) The heatmap of binary brain network at several specific times (The yellow point represents the element with the adjacency matrix of 1.); (c) The topographic map of the degree centrality at several specific times; (d) and (e) are the changes of the average shortest path length and clustering coefficient, respectively. (The red lines are the results of first-order linear quasi-sum based on the mean point of each window.)

• 

  Figure 5. The relative changes of the average shortest path length and the average clustering coefficient in phases T1, T2 and T3 compared to phase T0 in the brain functional network from all subjects in the delta rhythm. (a) The relative change of the average shortest path length; (b) The relative change of the average clustering coefficient. Vertical lines represent thresholds with significant differences compared to phase T0 (p<0.05).

• 

  Figure 6. The amount of functional connectivity (line) and the relative change (histogram) of phases T1–T3 compared to phase T0 from all subjects in (a) delta, (b) theta, (c) alpha and (d) beta rhythms. Four different colored shaded columns represent different phases. The histogram corresponds to the bottom and left axis, while the line corresponds to the bottom and right axis. Vertical lines represent the amount of functional connectivity with significant differences compared to phase T0 (p<0.05).

• 

  Figure 7. Results of the average shortest path length for (a) delta, (b) theta, (c) alpha and (d) beta rhythms. Vertical lines represent the average shortest path length with significant differences compared to phase T0 (p<0.05).

• 

  Figure 8. Results of the average clustering coefficient for (a) delta, (b) theta, (c) alpha and (d) beta rhythms. Vertical lines represent the average clustering coefficient with significant differences compared to phase T0 (p<0.05).

• 

  Figure 9. Results of the relative degree centrality for (a) delta, (b) theta, (c) alpha and (d) beta rhythms in phases T1–T3 compared to phase T0.

• 

  Figure 10. The brain topographic maps of the degree centrality for all subjects in different rhythms. Red circles represent the degree centrality with significant differences compared to phase T0 (p<0.05).