Fig. 4. Spatial distribution of recurrent connection strength $w$, external input current $I$, and noise amplitude $\mathit{\sigma }$, and their relationships with resting-state networks.

A Seven resting-state networks³⁹. B Strength of recurrent connection $w$ in 68 Desikan–Killiany cortical ROIs (left) and seven resting-state networks (right). C Strength of external input $I$ in 68 Desikan–Killiany cortical ROIs (left) and seven resting-state networks (right). D Strength of noise amplitude $\sigma$ in 68 Desikan–Killiany cortical ROIs (left) and seven resting-state networks (right). The boxplots comprised values obtained by “transferring” the parameter estimates from the 68 Desikan–Killiany parcels to all vertices (from the underlying cortical meshes) comprising each anatomical parcel. The vertex wise parameter values were then segregated based on the seven resting-state networks. Therefore, there were 3203, 2478, 1523, 1520, 1067, 1438, and 2886 values comprising the boxplots for somatomotor, visual, dorsal attention, ventral attention, limbic, control, and default networks, respectively. The boxes show the inter-quartile range (IQR) and the median. Whiskers indicate 1.5 IQR. Black crosses represent outliers. Recurrent connection strength and noise amplitude increased from sensory-motor to association (limbic, control, and default) networks. On the other hand, external input current was the highest in sensory-motor networks and decreased towards the default network. Source data are provided as a Source Data file.