Resting-State EEG Functional Connectivity and Independent Correlates of Cognitive Impairment in Patients with Atrial Fibrillation

Figures

• 

  Figure 1. Schematic overview of participant inclusion, grouping, and study design. Eligible participants were consecutively enrolled and allocated to one of the three study groups. Boxes represent the main study steps, arrows indicate the study flow, and colored boxes denote the three study groups.

• 

  Figure 2. Brain topographic maps of group differences in EEG power across frequency bands. (A) Absolute power spectral density (PSD); (B) relative band power. Rows, from top to bottom, show the δ, θ, α, β, and γ frequency bands. Colors indicate the one-way ANOVA F values at each electrode, with warmer colors indicating larger group differences. P values were adjusted using the Benjamini–Hochberg false discovery rate (FDR) method. Electrode locations follow the International 10–20 System.

• 

  Figure 3. Violin plots showing the distribution of weighted phase-lag index (wPLI) values across five frequency bands for the HC, PT-WoCI, and PT-CI groups. (A) δ-wPLI; (B) θ-wPLI; (C) α-wPLI; (D) β-wPLI; (E) γ-wPLI. Each violin plot represents the data distribution, with the embedded box plot indicating the median and interquartile range, and individual data points shown as scatter plots. Brackets indicate significant post hoc pairwise comparisons using Dunn’s test with Benjamini–Hochberg correction; only corrected significant comparisons are shown (*P < 0.05, **P < 0.01, and ***P < 0.001).

• 

  Figure 4. Channel pairs showing statistically significant differences in weighted phase-lag index (wPLI) across frequency bands among the three study groups. (A) δ-wPLI; (B) θ-wPLI; (C) α-wPLI; (D) β-wPLI; (E) γ-wPLI. Scalp maps are displayed using the International 10–20 System electrode layout. Yellow lines indicate channel pairs with significant overall group effects based on one-way ANOVA after Benjamini–Hochberg false discovery rate (FDR) correction.

• 

  Figure 5. Forest plot of the multivariable logistic regression analysis. Points represent odds ratios, and horizontal lines denote 95% confidence intervals estimated using the profile likelihood method. The model incorporates age, α-wPLI_z, cardiac function classification, sex, educational attainment, and occupational status. The x-axis is displayed on a logarithmic scale, with the vertical reference lines indicating an odds ratio of 1. Categorical variables are presented as the comparisons shown in parentheses, with the latter category serving as the reference.

• 

  Figure 6. Receiver operating characteristic curve of the primary multivariable logistic regression analysis. The model incorporates age, sex, educational attainment, occupational status, cardiac function classification, and α-wPLI_z. The apparent AUC calculated in the study sample was 0.814 (95% CI: 0.729–0.900). The dashed diagonal line represents random classification.