Abstract
A connectome is a comprehensive map of the connections in a nervous system — its wiring diagram. The term was coined in 2005 to name, for the brain, what the genome is for heredity: a complete inventory of elements and the links between them. Connectomes are built at three scales: the microscale of individual neurons and synapses, the mesoscale of neuronal populations, and the macroscale of brain regions and their fibre pathways. Because a connectome is a network, it is analysed with graph theory, which reveals that brains are organised as small-world networks with densely interconnected hubs. Connectomics matters to cognitive psychology because cognition depends not on isolated regions but on the pattern of connections among them, and individual differences in that pattern predict differences in behaviour.
Keywords: connectome, brain networks, graph theory, connectomics
The connectome is the idea that the mind's biological substrate is best understood not as a collection of specialised parts but as a network — a set of neural elements and the connections that join them. The word was introduced in 2005 by deliberate analogy to the genome: just as the genome is the complete set of genes, the connectome is the complete set of neural connections, and the promise was that mapping it would do for systems neuroscience what sequencing did for genetics (Sporns et al., 2005). The analogy is imperfect, but the ambition it names — a complete, principled description of neural wiring — has reorganised how neuroscientists and cognitive scientists think about the brain.
That reorganisation is why connectomics belongs in cognitive psychology and not only in anatomy. A wiring diagram is interesting to the study of mind because cognitive functions are increasingly understood as properties of distributed networks rather than of single regions: attention, memory, and control each depend on the coordinated activity of many areas, and it is the connections among them that make coordination possible (Park & Friston, 2013). The sections below set out what a connectome is, the three scales at which one is built, how connectomes are mapped, the graph-theoretic tools that turn a wiring diagram into quantitative claims about topology, and how the structure of the connectome relates to function and cognition.
- A connectome is a complete map of the connections in a nervous system, named by analogy to the genome (Sporns et al., 2005).
- Connectomes are built at three scales — microscale (neurons), mesoscale (populations), and macroscale (regions) — and different methods serve each (Sporns, 2013).
- The only complete synaptic-resolution connectomes belong to small organisms; the C. elegans connectome, with 302 neurons, was the first (White et al., 1986).
- Graph theory shows brain networks are small-world, with high clustering, short path lengths, and a rich club of interconnected hubs (Bullmore & Sporns, 2009; van den Heuvel & Sporns, 2011).
- Individual patterns of functional connectivity are stable enough to identify a person and to predict behaviour (Finn et al., 2015).
Figure 1
The Three Scales of the Connectome
What a Connectome Is
A connectome is a comprehensive map of the neural connections within a nervous system, often described as its wiring diagram. Formally it is a network: a set of nodes, which are the neural elements, together with a set of edges, which are the connections between them. What counts as a node and an edge depends on the scale of description, but the essential idea is constant — the connectome catalogues both the parts of a nervous system and, crucially, how those parts are linked (Sporns et al., 2005).
The term was coined in 2005 by explicit analogy to the genome, and the analogy carries a specific claim: that the connections are not incidental to understanding the brain but are themselves the object worth mapping completely (Sporns, 2013). Just as knowing an organism's full genetic sequence became a foundation for molecular biology, the hope was that knowing the brain's full connectional layout would become a foundation for understanding how neural activity gives rise to behaviour and cognition. The analogy has limits — a connectome is not a blueprint that generates a brain the way a genome contributes to generating an organism, and it changes across the lifespan — but as a statement of ambition it has been remarkably productive.
Two properties distinguish a connectome from an ordinary anatomical description. The first is completeness: a connectome aspires to map every connection within its chosen scope, not a selected subset. The second is that it is a network object, which means it can be studied with the mathematics of networks. This second property is what links raw anatomy to cognition, because it allows the wiring diagram to be summarised by quantitative measures — how clustered it is, how efficiently signals can travel across it, which elements are its hubs — that can then be related to what the brain does (Bullmore & Sporns, 2009).
Scales of the Connectome
A connectome can be described at any of three scales, and the choice of scale determines both what a node represents and what method can build the map (Betzel & Bassett, 2017). At the microscale, nodes are individual neurons and edges are the synapses between them. This is the finest resolution and the truest to the biology, but it is attainable only for small nervous systems, because reconstructing every neuron and synapse requires serial electron microscopy of the entire tissue — an enormous undertaking even for a few hundred cells.
At the mesoscale, nodes are local populations or columns of neurons and edges are the projections between them. This intermediate level is often the natural scale for animal-model connectomics, where tract-tracing can reveal how defined populations connect without resolving every individual synapse. At the macroscale, nodes are whole brain regions — typically the parcels of a cortical atlas — and edges are the large fibre bundles or the statistical couplings between regions. The macroscale is the scale of human connectomics, because it is the only level at which the entire human brain can currently be mapped, and it is mapped non-invasively (Van Essen et al., 2013).
The scales are not merely a matter of zoom; they trade resolution against coverage, and phenomena visible at one may be invisible at another (Betzel & Bassett, 2017). The first complete connectome of any organism was reconstructed at the microscale: the nervous system of the nematode Caenorhabditis elegans, with its 302 neurons and roughly 7,000 synapses, painstakingly traced from electron micrographs over more than a decade (White et al., 1986). No complete microscale connectome of a mammalian brain exists, and the human connectome is, for now, a macroscale object. Table 1 summarises how node, edge, method, and resolution change across the three scales.
| Scale | Node | Edge | Method | Resolution | Example |
|---|---|---|---|---|---|
| Microscale | Individual neuron | Synapse | Volume electron microscopy | ~nanometre | C. elegans (302 neurons) |
| Mesoscale | Neuronal population or column | Projection between populations | Tract-tracing | ~micrometre | Animal-model circuit maps |
| Macroscale | Brain region (atlas parcel) | Fibre pathway or statistical coupling | Diffusion MRI / functional MRI | ~millimetre | Human Connectome Project |
The connectome at three scales
A connectome can be described at three scales, and the scale determines what a node and an edge stand for, which method can build the map, and how much of the brain it can cover. Select a scale to see what changes. Resolution decreases and coverage increases as you move from microscale to macroscale — which is why only small nervous systems can be mapped neuron by neuron while the whole human brain is a macroscale object.
At the Macroscale, a node is a whole brain region (an atlas parcel) and an edge is a fibre bundle or a statistical coupling. The map is built by diffusion MRI and functional MRI at ~millimetre resolution, spanning on the order of 10² – 10³ regions.
The scale of human connectomics — the only level at which the entire human brain can currently be mapped, and mapped non-invasively.
Schematic only; node counts are order-of-magnitude ranges. The scales trade resolution against coverage, so no single method spans them all (Betzel & Bassett, 2017; Sporns, 2013).
How Connectomes Are Mapped
The method used to build a connectome follows from its scale. Microscale connectomes are reconstructed with volume electron microscopy: the tissue is sliced into ultrathin sections, each section is imaged at nanometre resolution, and the images are aligned and traced — increasingly with machine assistance — to follow each neuron's processes and identify its synapses. This yields a connectome of extraordinary detail but is limited to small volumes, which is why complete microscale maps exist for the worm, and now the fruit fly, but not for any mammal's whole brain (Sporns, 2013).
The human connectome is built at the macroscale from magnetic resonance imaging, and it comes in two varieties that must be kept distinct. A structural connectome is derived from diffusion MRI, which measures the directional diffusion of water along white-matter fibres and allows the major fibre pathways between regions to be reconstructed by tractography (Hagmann et al., 2008). A functional connectome is derived from functional MRI, which measures the correlation in activity between regions over time; an edge in a functional connectome represents statistical coupling, not a physical fibre (Finn et al., 2015). The two are related but not identical, and one of the central problems of the field is understanding how the fixed structural scaffold gives rise to the flexible, changing patterns of functional coupling (Suárez et al., 2020).
The scale of this enterprise prompted dedicated infrastructure. The Human Connectome Project, launched in 2010, was designed to map the macroscale connectome of hundreds of healthy adults with harmonised, high-quality imaging, and to release the data openly so that the field could analyse a common reference dataset rather than idiosyncratic single-laboratory scans (Van Essen et al., 2013). Its retrospective assessment records both the technical advances it drove — in acquisition, preprocessing, and parcellation — and the shift it produced toward large, shared, standardised connectome datasets (Elam et al., 2021).
The connectome as a matrix
A connectome is stored as a connectivity matrix: cell (i, j) records the connection between region i and region j. This network has twelve regions in three modules, with three hub regions (0, 4, 8) that link the modules together. Switch between the structural view — a binary map of which fibres exist, after thresholding — and the functional view, where the shade of each cell is the strength of statistical coupling. Click a region to highlight its row and column.
Region 0 (a hub) is in module 1 and has degree 11 in the functional view. high degree — connects across modules
Illustrative deterministic network (★ marks hubs). The block structure on the diagonal is the modular organisation of the connectome; the off-block cells the hubs add are the long-range links that make the network small-world (Bullmore & Sporns, 2009; van den Heuvel & Sporns, 2011).
Graph Theory and Network Topology
Once a connectome is represented as a network, it can be characterised with graph theory, the branch of mathematics that studies nodes and edges (Bullmore & Sporns, 2009). This is the step that converts a picture of wiring into quantitative, testable claims. A handful of measures do most of the work. The degree of a node is the number of connections it has. The clustering coefficient measures how often a node's neighbours are also connected to each other — how cliquey the network is locally. The characteristic path length is the average number of steps along edges needed to get from one node to any other — how efficiently the network can be traversed globally (Bullmore & Bassett, 2011).
Applied to brain networks, these measures reveal a consistent and non-trivial organisation. The small-world idea comes from network science: Watts and Strogatz showed that rewiring only a few connections of a regular lattice toward random targets sharply shortens path lengths while local clustering stays high, producing a network that is both clustered and efficiently traversable (Watts & Strogatz, 1998). Brains are small-world networks in exactly this sense: they combine high local clustering, like a regular lattice, with short global path lengths, like a random graph (Bullmore & Sporns, 2009). This combination is efficient, supporting both specialised local processing and rapid global integration, and it is not what either a purely regular or a purely random wiring would produce. Brains are also not homogeneous: a small number of high-degree nodes act as hubs, and these hubs are more densely interconnected with one another than their degree alone would predict — a rich-club organisation that forms a central backbone for communication across the whole network (van den Heuvel & Sporns, 2011).
The small-world index
A network is small-world when it keeps the high local clustering of a regular lattice while achieving the short path length of a random graph. The small-world index σ makes that precise by comparing the measured clustering C and path length L to a random graph of the same size and degree: σ = (C/Crand) / (L/Lrand). Values well above 1 mean small-world. The reference graph is the article’s N = 1,000, k = 10 example (Crand = 0.010, Lrand = 3.0). The defaults reproduce the Worked Example.
With C = 0.53 and L = 2.5, the clustering ratio is C/Crand = 53.0 and the path-length ratio is L/Lrand = 0.83, so σ = 63.6. strongly small-world
Reference values from the Worked Example (N = 1,000, k = 10): Crand = 0.010, Lrand = 3.0. The default C = 0.53, L = 2.5 give σ = 63.6, matching the article. σ > 1 is the standard threshold for small-worldness (Bullmore & Sporns, 2009).
The rich club is costly and consequential. Its long-range connections are metabolically and materially expensive, and because so much traffic is routed through a few hubs, damage to them is disproportionately disruptive — a structural fact that helps explain why hub regions figure prominently in a wide range of brain disorders (Fornito et al., 2015). Graph theory thus does more than describe: it identifies which elements of the connectome matter most, and why their disruption should have outsized effects on cognition.
Structure, Function, and Cognition
The reason connectomics reaches into cognitive psychology is that the connectome's structure constrains and shapes what the brain can compute. Cognitive functions are not localised to single regions so much as they emerge from the coordinated activity of distributed networks, and the connectome is the scaffold on which that coordination runs (Park & Friston, 2013). Understanding a cognitive function increasingly means understanding the network that supports it and how that network is embedded in the whole-brain connectome.
The link between the fixed structural connectome and the changing functional one is a central research problem (Suárez et al., 2020). Structure constrains function — two regions joined by a strong fibre pathway tend to show correlated activity — but the mapping is not one-to-one, because functional coupling also arises indirectly, through multi-step paths, and shifts with task and state. Relating the two is the work of network neuroscience, the field that has grown up around the connectome to treat the brain explicitly as a network across scales (Bassett & Sporns, 2017).
The most direct evidence that the connectome carries cognitively relevant information comes from individual differences. The pattern of functional connections is stable and distinctive enough to act as a fingerprint: a person can be identified from their functional connectome across separate scanning sessions, and the same individual patterns predict behavioural traits such as fluid intelligence (Finn et al., 2015). This finding has been generalised into a method — connectome-based predictive modelling — that uses whole-brain connectivity to predict behavioural and cognitive measures from data alone, turning the connectome from a description into a predictor (Shen et al., 2017).
Worked Example
Consider how graph theory decides whether a connectome is small-world. Take a simplified macroscale connectome of N = 1,000 regions in which each region connects to its k = 10 nearest neighbours, and compare it to two reference networks: a regular ring lattice and a random graph with the same size and degree.
For a regular lattice, the clustering coefficient is given by Clattice = 3(k − 2) / [4(k − 1)] = (3 × 8) / (4 × 9) = 24 / 36 = 0.667, and its characteristic path length is long, roughly Llattice ≈ N / (2k) = 1000 / 20 = 50 steps. For an equivalent random graph, clustering is low, Crand ≈ k / N = 10 / 1000 = 0.010, but the path length is short, Lrand ≈ ln(N) / ln(k) = ln(1000) / ln(10) = 6.908 / 2.303 = 3.0 steps.
A real connectome sits between these extremes. Suppose measurement gives it clustering C = 0.53 and path length L = 2.5. The small-world index normalises the network against its random equivalent on both measures (Humphries & Gurney, 2008): σ = (C / Crand) / (L / Lrand) = (0.53 / 0.010) / (2.5 / 3.0) = 53 / 0.833 = 63.6. Because σ is far greater than 1 — the network keeps nearly lattice-like clustering while achieving nearly random-like path length — the connectome is strongly small-world. That single number captures the defining topological signature of brain networks: dense local processing combined with efficient global integration (Bullmore & Sporns, 2009).
Discussion
The connectome has reframed a long-standing question in cognitive science — how the brain's parts give rise to the mind — by insisting that the connections are as much the object of study as the parts. This is a genuine shift. For much of the twentieth century the dominant programme was localisation: assigning functions to regions. Connectomics does not discard localisation but embeds it in a network view, in which what a region does depends on its connectional context, and in which many cognitive functions are properties of distributed circuits rather than of any single area (Park & Friston, 2013).
The framework also has clear limits that temper the genome analogy that named it. A connectome is not static: synaptic connections form and prune across development and learning, so any map is a snapshot of a changing system. Different scales can disagree, and the macroscale human connectome, inferred from MRI, is a coarse and indirect estimate rather than a ground-truth wiring diagram — tractography has known biases, and functional edges are statistical, not anatomical. And a wiring diagram, however complete, does not by itself explain dynamics: knowing the connections of C. elegans did not immediately yield an understanding of its behaviour (White et al., 1986). The connectome constrains function; it does not fully determine it.
What the connectome contributes to cognitive psychology, then, is not a theory of cognition but a substrate for one: a principled, quantitative account of the network on which cognitive processes run, and a set of tools for relating individual differences in that network to individual differences in behaviour (Bassett & Sporns, 2017).
Current Directions
Connectomics is advancing on several fronts at once. The most visible is scale at the microscale: after decades in which the 302-neuron worm was the only complete synaptic map, volume electron microscopy and machine-assisted reconstruction have now delivered a whole-brain connectome of the adult fruit fly, with on the order of 105 neurons and 107 synapses — a step change in the size of nervous system that can be mapped neuron by neuron (Sporns, 2013).
A second front is the maturation of network neuroscience as a discipline in its own right, with a shared vocabulary and a growing methodological rigour around how brain graphs are built and compared (Bassett & Sporns, 2017). Central to this is the structure–function problem: recent work uses communication models and generative models to explain how the fixed structural connectome gives rise to the flexible functional connectome, moving beyond simple correlations between the two (Suárez et al., 2020). A third front is clinical. The connectomics of brain disorders reframes conditions from schizophrenia to Alzheimer's disease as disorders of connectivity — disconnection syndromes or hub failures — and connectome-based predictive modelling is being developed toward individualised prediction of cognitive and clinical outcomes from a person's own connectivity (Fornito et al., 2015; Shen et al., 2017). The reference datasets that make much of this possible continue to grow and standardise in the wake of the Human Connectome Project (Elam et al., 2021).
Common Misconceptions
- “The human connectome has been completely mapped.”
- It has not. The complete, synaptic-resolution human connectome does not exist and is far beyond current technology. What exists for humans is the macroscale connectome — a map of regions and pathways inferred from MRI — which is coarse and indirect. Complete synaptic maps exist only for small organisms (Sporns, 2013).
- “The connectome is fixed, like a genome.”
- The genome analogy names an ambition, not an identity. Unlike a genome, a connectome changes: connections are formed, strengthened, and pruned throughout development and in response to learning and experience, so any connectome is a snapshot of a plastic, changing system (Bassett & Sporns, 2017).
- “A structural and a functional connectome are the same thing.”
- They are distinct. A structural connectome maps physical fibre pathways from diffusion MRI; a functional connectome maps statistical correlations in activity from functional MRI. An edge in a functional connectome need not correspond to a direct fibre, and relating the two is an open problem (Suárez et al., 2020).
- “Knowing the wiring diagram explains behaviour.”
- A connectome constrains function but does not fully determine it. The complete wiring of C. elegans has been known since 1986, yet a full account of how that circuit produces the worm's behaviour is still not settled. Structure is necessary for understanding dynamics, not sufficient (White et al., 1986).
Glossary
- Characteristic path length.
- The average number of edges on the shortest path between pairs of nodes; a measure of how efficiently signals can traverse the whole network. Short path lengths indicate high global integration.
- Clustering coefficient.
- The fraction of a node's neighbours that are also connected to each other; a measure of local interconnectedness, or how cliquey the network is around a node.
- Connectome.
- A comprehensive map of the connections in a nervous system; the set of neural elements (nodes) and the connections between them (edges), described at a chosen scale.
- Connectomics.
- The scientific enterprise of producing and analysing connectomes, spanning data acquisition, reconstruction, and network analysis.
- Degree.
- The number of connections attached to a node; nodes of unusually high degree are candidate hubs.
- Diffusion MRI.
- A magnetic-resonance method that measures the directional diffusion of water to infer the orientation of white-matter fibres, enabling reconstruction of the structural connectome by tractography.
- Edge.
- A connection between two nodes in a network; in a connectome an edge may be a synapse, a fibre pathway, or a statistical coupling, depending on the scale.
- Functional connectome.
- A network whose edges represent statistical correlations in activity between regions over time, typically derived from functional MRI; distinct from the physical wiring.
- Graph theory.
- The branch of mathematics dealing with networks of nodes and edges; the toolkit that turns a connectome into quantitative measures of topology.
- Hub.
- A node with a large number of connections that occupies a central position in the network; hubs are disproportionately important for integration and disproportionately costly to lose.
- Node.
- A neural element in a connectome — a neuron, a population, or a brain region — depending on the scale of description.
- Rich club.
- A set of high-degree hubs that are more densely interconnected with one another than their degrees alone would predict, forming a central backbone for whole-brain communication.
- Small-world network.
- A network that combines high local clustering with short global path lengths; brain networks are small-world, balancing specialised local processing against efficient global integration.
- Structural connectome.
- A network whose edges represent physical connections — synapses or fibre pathways — between neural elements; the anatomical wiring diagram.
- Tractography.
- The computational reconstruction of white-matter fibre pathways from diffusion-MRI data; the method by which the macroscale structural connectome is estimated.
Key Researchers
Danielle S. Bassett
(living). Physicist and network neuroscientist at the University of Pennsylvania; co-defined the field of network neuroscience and authored foundational treatments of brain graphs and of multi-scale network organisation (Bullmore & Bassett, 2011; Bassett & Sporns, 2017; Betzel & Bassett, 2017). ORCID · Wikipedia · Google Scholar
Sydney Brenner
(1927–2019). Biologist and 2002 Nobel laureate; led the reconstruction of the C. elegans nervous system, the first complete connectome of any organism, establishing that a full wiring diagram was achievable in principle (White et al., 1986). Wikipedia · Wikidata
Edward T. Bullmore
(living). Neuropsychiatrist at the University of Cambridge; with Sporns he brought graph-theoretical analysis to the study of structural and functional brain networks, providing the analytic framework of connectomics (Bullmore & Sporns, 2009; Bullmore & Bassett, 2011). ORCID · Wikipedia · Google Scholar
David C. Van Essen
(living). Neuroscientist at Washington University in St. Louis; principal investigator of the WU-Minn Human Connectome Project, which produced the standardised, openly shared reference dataset of the human macroscale connectome (Van Essen et al., 2013; Elam et al., 2021). ORCID · Wikipedia · Google Scholar
Emily S. Finn
(living). Cognitive neuroscientist at Dartmouth College; showed that an individual's functional connectome is a stable, identifying fingerprint that predicts behavioural traits, opening the study of individual differences in connectivity (Finn et al., 2015). ORCID · Google Scholar
Patric Hagmann
(living). Neuroradiologist at Lausanne University Hospital and the University of Lausanne; mapped the structural core of the human cerebral cortex with diffusion MRI, providing an early empirical macroscale human connectome (Hagmann et al., 2008). ORCID · Google Scholar
Martijn P. van den Heuvel
(living). Neuroscientist at Vrije Universiteit Amsterdam; discovered the rich-club organisation of the human connectome, showing that its hubs form a densely interconnected central backbone (van den Heuvel & Sporns, 2011). ORCID · Google Scholar
Olaf Sporns
(living). Computational neuroscientist at Indiana University Bloomington; coined the term connectome and has been central to establishing network neuroscience, from the rich club to the structure–function problem (Sporns et al., 2005; Sporns, 2013; Bassett & Sporns, 2017). ORCID · Wikipedia · Google Scholar
Frequently Asked Questions
What is a connectome?
A connectome is a comprehensive map of the connections in a nervous system — its wiring diagram. It is a network of nodes, the neural elements, and edges, the connections between them. The term was coined in 2005 by analogy to the genome, to name the ambition of mapping neural connectivity completely.
Has the human connectome been fully mapped?
No. A complete, synaptic-resolution map of the human brain does not exist and is well beyond current technology. What exists for humans is the macroscale connectome, a map of brain regions and the pathways between them inferred non-invasively from MRI. Complete synaptic-level connectomes exist only for small organisms.
What is the difference between a structural and a functional connectome?
A structural connectome maps physical connections — synapses or white-matter fibre pathways — and for humans is derived from diffusion MRI. A functional connectome maps statistical correlations in activity between regions over time, derived from functional MRI. A functional edge reflects coupling, not necessarily a direct anatomical link.
What are the three scales of the connectome?
The microscale, where nodes are individual neurons and edges are synapses; the mesoscale, where nodes are populations of neurons; and the macroscale, where nodes are whole brain regions and edges are fibre pathways or statistical couplings. Resolution decreases and coverage increases from micro to macro.
Why is graph theory used to study connectomes?
Because a connectome is a network, and graph theory is the mathematics of networks. It converts a wiring diagram into quantitative measures — degree, clustering, path length, hubs — that can be compared across brains and related to behaviour, turning anatomy into testable claims about organisation.
What does it mean that the brain is a small-world network?
It means the brain combines high local clustering, like a regular lattice, with short global path lengths, like a random graph. This organisation is efficient: it supports specialised local processing and rapid communication across the whole brain at the same time, which neither purely regular nor purely random wiring achieves.
What is the rich club?
The rich club is a set of high-degree hub regions that are more densely interconnected with each other than their individual degrees would predict. It forms a central backbone for communication across the brain, is metabolically costly, and its disruption is implicated in many brain disorders.
Why does the connectome matter for cognitive psychology?
Because cognitive functions arise from distributed networks rather than isolated regions, and the connectome is the scaffold those networks run on. Individual differences in connectivity are stable enough to identify a person and to predict traits such as fluid intelligence, linking the wiring diagram directly to behaviour.
References
Bassett, D. S., & Sporns, O. (2017). Network neuroscience. Nature Neuroscience, 20(3), 353–364. https://doi.org/10.1038/nn.4502
Betzel, R. F., & Bassett, D. S. (2017). Multi-scale brain networks. NeuroImage, 160, 73–83. https://doi.org/10.1016/j.neuroimage.2016.11.006
Bullmore, E., & Sporns, O. (2009). Complex brain networks: Graph theoretical analysis of structural and functional systems. Nature Reviews Neuroscience, 10(3), 186–198. https://doi.org/10.1038/nrn2575
Bullmore, E. T., & Bassett, D. S. (2011). Brain graphs: Graphical models of the human brain connectome. Annual Review of Clinical Psychology, 7, 113–140. https://doi.org/10.1146/annurev-clinpsy-040510-143934
Elam, J. S., Glasser, M. F., Harms, M. P., Sotiropoulos, S. N., Andersson, J. L. R., Burgess, G. C., . . . Van Essen, D. C. (2021). The Human Connectome Project: A retrospective. NeuroImage, 244, 118543. https://doi.org/10.1016/j.neuroimage.2021.118543
Finn, E. S., Shen, X., Scheinost, D., Rosenberg, M. D., Huang, J., Chun, M. M., Papademetris, X., & Constable, R. T. (2015). Functional connectome fingerprinting: Identifying individuals using patterns of brain connectivity. Nature Neuroscience, 18(11), 1664–1671. https://doi.org/10.1038/nn.4135
Fornito, A., Zalesky, A., & Breakspear, M. (2015). The connectomics of brain disorders. Nature Reviews Neuroscience, 16(3), 159–172. https://doi.org/10.1038/nrn3901
Hagmann, P., Cammoun, L., Gigandet, X., Meuli, R., Honey, C. J., Wedeen, V. J., & Sporns, O. (2008). Mapping the structural core of human cerebral cortex. PLoS Biology, 6(7), e159. https://doi.org/10.1371/journal.pbio.0060159
Humphries, M. D., & Gurney, K. (2008). Network ‘small-world-ness’: A quantitative method for determining canonical network equivalence. PLoS ONE, 3(4), e2051. https://doi.org/10.1371/journal.pone.0002051
Park, H.-J., & Friston, K. (2013). Structural and functional brain networks: From connections to cognition. Science, 342(6158), 1238411. https://doi.org/10.1126/science.1238411
Shen, X., Finn, E. S., Scheinost, D., Rosenberg, M. D., Chun, M. M., Papademetris, X., & Constable, R. T. (2017). Using connectome-based predictive modeling to predict individual behavior from brain connectivity. Nature Protocols, 12(3), 506–518. https://doi.org/10.1038/nprot.2016.178
Sporns, O., Tononi, G., & Kötter, R. (2005). The human connectome: A structural description of the human brain. PLoS Computational Biology, 1(4), e42. https://doi.org/10.1371/journal.pcbi.0010042
Sporns, O. (2013). The human connectome: Origins and challenges. NeuroImage, 80, 53–61. https://doi.org/10.1016/j.neuroimage.2013.03.023
Suárez, L. E., Markello, R. D., Betzel, R. F., & Mišić, B. (2020). Linking structure and function in macroscale brain networks. Trends in Cognitive Sciences, 24(4), 302–315. https://doi.org/10.1016/j.tics.2020.01.008
van den Heuvel, M. P., & Sporns, O. (2011). Rich-club organization of the human connectome. Journal of Neuroscience, 31(44), 15775–15786. https://doi.org/10.1523/JNEUROSCI.3539-11.2011
Van Essen, D. C., Smith, S. M., Barch, D. M., Behrens, T. E. J., Yacoub, E., & Ugurbil, K. (2013). The WU-Minn Human Connectome Project: An overview. NeuroImage, 80, 62–79. https://doi.org/10.1016/j.neuroimage.2013.05.041
Watts, D. J., & Strogatz, S. H. (1998). Collective dynamics of ‘small-world’ networks. Nature, 393(6684), 440–442. https://doi.org/10.1038/30918
White, J. G., Southgate, E., Thomson, J. N., & Brenner, S. (1986). The structure of the nervous system of the nematode Caenorhabditis elegans. Philosophical Transactions of the Royal Society of London. B, Biological Sciences, 314(1165), 1–340. https://doi.org/10.1098/rstb.1986.0056