Mapping the structural core of human cerebral cortex

Abstract

Structurally segregated and functionally specialized regions of the human cerebral cortex are interconnected by a dense network of cortico-cortical axonal pathways. By using diffusion spectrum imaging, we noninvasively mapped these pathways within and across cortical hemispheres in individual human participants. An analysis of the resulting large-scale structural brain networks reveals a structural core within posterior medial and parietal cerebral cortex, as well as several distinct temporal and frontal modules. Brain regions within the structural core share high degree, strength, and betweenness centrality, and they constitute connector hubs that link all major structural modules. The structural core contains brain regions that form the posterior components of the human default network. Looking both within and outside of core regions, we observed a substantial correspondence between structural connectivity and resting-state functional connectivity measured in the same participants. The spatial and topological centrality of the core within cortex suggests an important role in functional integration.

Figure 1. Extraction of a Whole Brain Structural Connectivity Network

(1) High-resolution T1 weighted and diffusion spectrum MRI (DSI) is acquired. DSI is represented with a zoom on the axial slice of the reconstructed diffusion map, showing an orientation distribution function at each position represented by a deformed sphere whose radius codes for diffusion intensity. Blue codes for the head-feet, red for left-right, and green for anterior-posterior orientations. (2) White and gray matter segmentation is performed from the T1-weighted image. (3a) 66 cortical regions with clear anatomical landmarks are created and then (3b) individually subdivided into small regions of interest (ROIs) resulting in 998 ROIs. (4) Whole brain tractography is performed providing an estimate of axonal trajectories across the entire white matter. (5) ROIs identified in step (3b) are combined with result of step (4) in order to compute the connection weight between each pair of ROIs. The result is a weighted network of structural connectivity across the entire brain. In the paper, the 66 cortical regions are labeled as follows: each label consists of two parts, a prefix for the cortical hemisphere (r = right hemisphere, l = left hemisphere) and one of 33 designators: BSTS = bank of the superior temporal sulcus, CAC = caudal anterior cingulate cortex, CMF = caudal middle frontal cortex, CUN = cuneus, ENT = entorhinal cortex, FP = frontal pole, FUS = fusiform gyrus, IP = inferior parietal cortex, IT = inferior temporal cortex, ISTC = isthmus of the cingulate cortex, LOCC = lateral occipital cortex, LOF = lateral orbitofrontal cortex, LING = lingual gyrus, MOF = medial orbitofrontal cortex, MT = middle temporal cortex, PARC = paracentral lobule, PARH = parahippocampal cortex, POPE = pars opercularis, PORB = pars orbitalis, PTRI = pars triangularis, PCAL = pericalcarine cortex, PSTS = postcentral gyrus, PC = posterior cingulate cortex, PREC = precentral gyrus, PCUN = precuneus, RAC = rostral anterior cingulate cortex, RMF = rostral middle frontal cortex, SF = superior frontal cortex, SP = superior parietal cortex, ST = superior temporal cortex, SMAR = supramarginal gyrus, TP = temporal pole, and TT = transverse temporal cortex.

Figure 2. Node Degree and Node Strength Distributions

(A) Ranked distribution of node degree for left and right cerebral hemispheres. Shaded bars represent means across five participants and symbols indicate data for individual participants. (B) Ranked distribution of node strength for left and right cerebral hemispheres. (C) ROI strength obtained from high-resolution connection matrices. The plot shows how consistently ROI strength ranked in the top 20% across participants.

Figure 3. High-Resolution Connection Matrix, Network Layout and Connectivity Backbone (Participant A, scan 2)

(A) Matrix of fiber densities (connection weights) between all pairs of n = 998 ROIs. ROIs are plotted by cerebral hemispheres, with right-hemispheric ROIs in the upper left quadrant, left-hemispheric ROIs in the lower right quadrant, and interhemispheric connections in the upper right and lower left quadrants. The color bars at the left and bottom of the matrix correspond to the colors of the 66 anatomical subregions shown in Figure 1. All connections are symmetric and displayed with a logarithmic color map. (B) Kamada-Kawai force-spring layout of the connectivity backbone. Labels indicating anatomical subregions are placed at their respective centers of mass. Nodes (individual ROIs) are coded according to strength and edges are coded according to connection weight (see legend). (C) Dorsal and lateral views of the connectivity backbone. Node and edge coding as in (B).

Figure 4. Average Regional Connection Matrix, Network Layout, and Connectivity Backbone

(A) Matrix of inter-regional fiber densities between pairs of anatomical subregions, obtained by averaging over fiber densities for all pairs of ROIs within the regions, and averaging across all five participants. Connection weights are symmetric and are plotted on a logarithmic scale. For corresponding plots for all individual participants, see Figure S2. (B) Network layout. (C) Dorsal and medial views of the connectivity backbone in anatomical coordinates.

Figure 5. Structural Network Cores

(A) Network cores for each individual participant derived by k-core decomposition of a binary connection matrix obtained by thresholding the high-resolution fiber densities such that a total of 10,000 connections remain in each participant. Nodes are plotted according to their core number, counted backwards from the last remaining core. (B) Average network core across all five participants. (C) Ranked distribution of core numbers for left and right cerebral hemispheres. Shaded bars represent means across five participants and symbols indicate data for individual participants

Figure 6. Modularity and Hub Classification

The modularity was derived from the average regional connection matrix. Modules are listed in Table S1. (A) The plot shows a dorsal view, with nodes representing anatomical subregions. The spatial position of each region corresponds to the center of mass coordinates calculated from participant A, scan 2 (as seen in Figure 4C). Six modules are shown as gray circles centered on their center of mass and sized according to their number of members. Edges correspond to the average connection densities of each region with the member regions of each of the six modules, plotted between that region's spatial coordinates and the center of mass of each module Connector hubs are defined as regions with above average strength and a participation index p ≥ 0.3, indicating a high proportion of cross-module connectivity. These regions are marked as filled yellow circles. Provincial hubs have above-average strength and P < 0.3; they are marked as unfilled yellow circles. (B) Connector hubs obtained from analyses of high-resolution connection matrices. ROIs are displayed according to how consistently a given ROI was identified as a connector hub across participants.

Figure 7. Centrality and Efficiency

(A) ROI centrality obtained from analyses of high-resolution connection matrices. The plot shows how consistently ROI centrality ranked in the top 20% across participants. (B) Lateral views of the right and left cerebral hemispheres showing ROI centrality, averaged across all five participants and projected onto the cortical surface of participant A. (C) Ranked distribution of betweenness centrality for left and right cerebral hemispheres. Shaded bars represent means across five participants and symbols indicate data for individual participants. (D) Ranked distribution of efficiency for left and right cerebral hemispheres.

Figure 8. Comparison of Structural and Functional Connectivity

(A) Map of functional correlations from resting state fMRI for a cluster of five seed ROIs located within 10 mm of the Talairach coordinate [–5 −49 40] (marked by a white circle). Correlations are averaged over the five ROIs and over scanning sessions for all five participants. The plot shows a lateral and medial view of the left cerebral hemisphere. (B) Scatter plot of structural and functional connections of the precuneus and posterior cingulate cortex (PCUN and PC, left and right hemisphere), for all five participants. (C) Scatter plots for structural and functional connections averaged over all five participants, for all anatomical subregions in both hemispheres.