Olaf SpornspersName, Giulio TononipersName, Rolf KötterpersName (2005date). “The Human Connectome: A Structural Description of the Human Brain.” PLOS Computational Biologytitle·j, BloomingtonplaceName.
Object of study — The complete map of anatomical connectionsterm·key in the human brainterm·key - what it would contain, at what scale it could be built, and by which techniques.
Research question — At what level of descriptionterm·var - single neuronsterm·key, neuronal populations, or brain regionsterm·dom - could a first draft of a complete human connectionterm·key map actually be assembled?
Contribution — Names the brainterm·key's connection matrixterm·mth the connectometerm·dom, argues the macroscaleterm·mth is the only feasible level for a first draft, and proposes combined diffusion and functional imaging as the route to it.
phenomenon · claim-type · cross-domain-transfer · mechanism · claim-relation · replication-status · limitation — grounded in the full text| phenomenon | structural connectivity map of the human brain (the connectome) — “The connection matrix of the human brain (the human “connectome”) represents an indispensable foundation for basic and applied neurobiological research.” | |
| claim-type | method — “We propose a research strategy to achieve this goal, and discuss its potential impact.” | |
| method — evidence | — “The purpose of this article is to discuss research strategies aimed at a comprehensive structural description of the network of elements and connections forming the human brain.” | |
| object of study — evidence | — “The purpose of this article is to discuss research strategies aimed at a comprehensive structural description of the network of elements and connections forming the human brain.” | |
| research question — evidence | — “It is important to determine which level of description is the most appropriate for a first draft of the human connectome.” | |
| contribution — evidence | — “We propose to call this dataset the human “connectome,” and we argue that it is fundamentally important in cognitive neuroscience and neuropsychology.” | |
| transfer | from network science / graph theory: applying complex-network / graph-theoretic analysis (small-world structure, motifs, global measures) to map and characterize the brain connectome — “Both genome and connectome constitute complex networks” | transfers[0] |
| transfer | from genomics: the genome as the model for a complete, standardised structural inventory - invoked here as the comparison against which the connectome's elements are shown to be harder to define — “The compilation of the genome was aided by the fact that base pairs and genes are relatively straightforward choices as basic structural elements.” | transfers[1] |
| mechanism | a brain region can be delineated by its connection profile, because the elements of one region share a long-range connectivity pattern that differs from its neighbours' — “What defines a segregated brain region is that all its structural elements share highly similar long-range connectivity patterns, and that these patterns are dissimilar between brain regions.” | mechanisms[0] |
| mechanism | the single-neuron scale is ruled out not only by size but by plasticity: individual synapses change too fast to be a stable element — “We suggest that the vast number, high variability, and fast dynamics of individual neurons and synapses render them inappropriate as basic elements for an initial draft of the connectome.” | mechanisms[1] |
| mechanism | cognition is treated as a network phenomenon organised by functional segregation and functional integration, which is why the connection map is held to be the missing piece — “A potential theoretical framework for conceptualizing cognition as a network phenomenon is based on two main organizational principles found in the cerebral cortex, functional segregation, and functional integration [ 6 , 7 ].” | mechanisms[2] |
| builds-on | Mountcastle [54, 55], the cortical minicolumn as basic functional unit — “Mountcastle [ 54 , 55 ] originally proposed the cortical minicolumn as a basic functional unit of mammalian cerebral cortex.” | claim_relations[0] |
| builds-on | existing large-scale connection matrices and their matrix conventions [16, 17] — “In keeping with conventions adopted by other authors [ 16 , 17 ], confirmed absence of a connection is denoted by a ij = 0” | claim_relations[1] |
| contradicts | existing connectional models of the human brain [13] built by extrapolating from macaque anatomy — “Connectional models of the human brain are scarce and poorly defined [ 13 ], and they are largely based on extrapolating anatomical information from other primate species such as the macaque monkey, an approach that is problematic [ 14 ], in part, because of our incomplete understanding of evolutionary changes.” | claim_relations[2] |
| builds-on | existing large-scale connection matrices for macaque, cat and rat — “Computational analyses of these datasets have revealed a broad range of network characteristics [ 28 ], including the existence of clusters of brain regions [ 29 ], hierarchical organization [ 30 , 31 ], small-world attributes [ 32 , 33 ]” | claim_relations[3] |
| contradicts | connectional models of the human brain extrapolated from macaque anatomy — “they are largely based on extrapolating anatomical information from other primate species such as the macaque monkey, an approach that is problematic [ 14 ]” | claim_relations[4] |
| sense | network - the brain's anatomical wiring, whose elements are neurons or regions and whose connections are pathways — “the network of anatomical connections linking the neuronal elements of the human brain is still largely unknown” | senses[0] |
| sense | network - cognition itself read as a network phenomenon, organised by functional segregation and integration rather than by anatomy alone — “A potential theoretical framework for conceptualizing cognition as a network phenomenon is based on two main organizational principles found in the cerebral cortex, functional segregation, and functional integration [ 6 , 7 ].” | senses[1] |
| sense | network - natural and technological networks generally, invoked as the class the brain belongs to and then partly excluded from — “Compiling the connectome faces two significant challenges not shared by other natural or technological networks.” | senses[2] |
| sense | neuroinformatics - the discipline arriving with the rapid proliferation of neuroscience datasets — “Recent research in neuroscience has resulted in a rapid proliferation of neuroscience datasets and in the arrival of a new discipline, neuroinformatics [ 1 – 4 ].” | senses[3] |
| sense | functional segregation - one of two cortical principles in a network account of cognition — “A potential theoretical framework for conceptualizing cognition as a network phenomenon is based on two main organizational principles found in the cerebral cortex, functional segregation, and functional integration [ 6 , 7 ].” | senses[4] |
| sense | functional integration - one of two cortical principles in a network account of cognition — “A potential theoretical framework for conceptualizing cognition as a network phenomenon is based on two main organizational principles found in the cerebral cortex, functional segregation, and functional integration [ 6 , 7 ].” | senses[5] |
| sense | mesoscale - neuronal groups or populations between the microscale and macroscale — “Different kinds of structural descriptions could target at least three rather distinct levels of organization. At opposite ends of the scale are the level of single neurons and synapses (microscale) and the level of anatomically distinct brain regions and inter-regional pathways (macroscale). Between these two levels is the level of neuronal groups or populations (mesoscale).” | senses[6] |
| sense | cortical minicolumn - a proposed basic functional unit of mammalian cerebral cortex — “Mountcastle [ 54 , 55 ] originally proposed the cortical minicolumn as a basic functional unit of mammalian cerebral cortex.” | senses[7] |
| sense | elementary processing units - local neuronal populations such as cortical minicolumns — “A further step in this direction involves the characterization of connection patterns among elementary processing units, corresponding to local populations of neurons such as cortical minicolumns.” | senses[8] |
| sense | probabilistic tractography - diffusion-based tracing used with connectivity profiles to parcellate cortex — “Step 1 is to perform diffusion-weighted imaging followed by probabilistic tractography of thalamocortical tracts as well as corticocortical interareal pathways, using correlations in connectivity profiles to assist in parcellating human cortical regions. The end result is a voxel-wise probabilistic all-to-all structural connectivity matrix for the human brain.” | senses[9] |
| sense | structural connectivity matrix - a voxel-wise probabilistic all-to-all matrix from imaging and tractography — “Step 1 is to perform diffusion-weighted imaging followed by probabilistic tractography of thalamocortical tracts as well as corticocortical interareal pathways, using correlations in connectivity profiles to assist in parcellating human cortical regions. The end result is a voxel-wise probabilistic all-to-all structural connectivity matrix for the human brain.” | senses[10] |
| sense | functional connectivity matrix - a voxel-wise all-to-all correlation matrix of registered brain activity — “Step 2 is to perform a correlation analysis of spatially registered, equally resolved resting activity and/or multistimulus/multitask activation data (functional magnetic resonance imaging and/or magnetoencephalography) recorded in the same person [ 73 ], emphasizing strong functional relationships that are consistent across tasks [ 74 ]. The end result is a voxel-wise all-to-all functional connectivity matrix for the human brain.” | senses[11] |
| sense | cross-level integration - combining macroscale and minicolumn datasets into a mesoscale connectome — “Thus, macroscale and minicolumn descriptions deliver complementary datasets that need cross-level integration to achieve a single mesoscale version of the connectome.” | senses[12] |
| sense | structure–function relationships - not directly readable from connectivity and requiring separate empirical and computational study — “We emphasize that structure–function relationships are not directly evident from the connectional dataset itself. Rather, their elucidation will require further intense empirical and computational study.” | senses[13] |
| sense | structural benchmarks - connectome references for model comparison and integration — “If the connectome is sufficiently comprehensive and accessible, it could also provide a set of structural benchmarks that might facilitate the comparison and integration of the numerous specialized and structurally based models that have already been proposed in computational neuroscience.” | senses[14] |
| edge | brain region —operationalised-as→ connectivity patterns — “What defines a segregated brain region is that all its structural elements share highly similar long-range connectivity patterns, and that these patterns are dissimilar between brain regions.” | concept_edges[5] |
| evidence status note | concept — FLATTENED EXPONENT IN THIS QUOTE - THE NUMBER IS WRONG BY ORDERS OF MAGNITUDE AS IT STANDS. The staged text reads '3 × 10 9', where the paper writes 10 to the power 9. The superscript was pushed onto the line by the PDF extractor. The quote is verbatim to the staged text and its anchor is sound; the STAGED TEXT is what is damaged. NOT repaired - repair means re-staging the body and re-anchoring the record. — “The human genome is composed of approximately 3 × 10 9 base pairs, containing around 20,000–30,000 genes [ 15 ].” | concepts[8] |
| for which use | limitation — the feasibility of the programme: COMPILING THE CONNECTOME FACES TWO SIGNIFICANT CHALLENGES NOT SHARED BY OTHER NATURAL OR TECHNOLOGICAL NETWORKS — “Compiling the connectome faces two significant challenges not shared by other natural or technological networks.” | limitations[0] |
| for which use | limitation — the single-neuron scale, ruled out: ATTEMPTING TO ASSEMBLE THE HUMAN CONNECTOME AT THE LEVEL OF SINGLE NEURONS IS UNREALISTIC AND WILL REMAIN INFEASIBLE AT LEAST IN THE NEAR FUTURE — “Attempting to assemble the human connectome at the level of single neurons is unrealistic and will remain infeasible at least in the near future.” | limitations[1] |
| for which use | limitation — the nodes of the proposed network: NO SINGLE UNIVERSALLY ACCEPTED PARCELLATION SCHEME CURRENTLY EXISTS FOR HUMAN BRAIN REGIONS, POSING A SIGNIFICANT OBSTACLE TO CREATING A UNIFIED RESOURCE SUCH AS THE CONNECTOME — “No single universally accepted parcellation scheme currently exists for human brain regions (e.g., areas of the cerebral cortex), posing a significant obstacle to creating a unified resource such as the connectome.” | limitations[2] |
| for which use | limitation — every connection value in the resource: THE PRESENCE OF SIGNIFICANT INTERINDIVIDUAL VARIABILITY IN STRUCTURAL CONNECTION PATTERNS, EVEN AT THE MACROSCALE LEVEL, AND THE FUNDAMENTALLY PROBABILISTIC NATURE OF CONNECTIVITY DATASETS — “The presence of significant interindividual variability in structural connection patterns, even at the macroscale level, and the fundamentally probabilistic nature of connectivity datasets provided in the connectome may be viewed as fundamental weaknesses of the proposal, undermining its comprehensive goal of a definitive structural description of the human brain.” | limitations[3] |
| for which use | limitation — any inference from connectivity to function: WE EMPHASIZE THAT STRUCTURE-FUNCTION RELATIONSHIPS ARE NOT DIRECTLY EVIDENT FROM THE CONNECTIONAL DATASET ITSELF — “We emphasize that structure–function relationships are not directly evident from the connectional dataset itself.” | limitations[4] |
| for which use | limitation — everything the connectome is later used to answer: THERE ARE MANY STRUCTURAL AND FUNCTIONAL ASPECTS THAT THE HUMAN CONNECTOME, AS ENVISIONED IN THIS ARTICLE, DOES NOT CONTAIN OR ADDRESS — “There are many structural and functional aspects that the human connectome, as envisioned in this article, does not contain or address.” | limitations[5] |
| for which use | limitation — four named processes the first draft omits: ITS FIRST DRAFT DOES NOT CAPTURE SYNAPTIC PLASTICITY AND REMODELING, NONSYNAPTIC COMMUNICATION, HORMONAL REGULATION, OR DEGENERATIVE PROCESSES — “Its first draft does not capture synaptic plasticity and remodeling, nonsynaptic communication, hormonal regulation, or degenerative processes.” | limitations[6] |
| for which use | limitation — the same scope as limitations[2] - no agreed parcellation, so the nodes are not yet definable — “No single universally accepted parcellation scheme currently exists for human brain regions (e.g., areas of the cerebral cortex), posing a significant obstacle to creating a unified resource such as the connectome.” | limitations[7] |
| duplicate of | limitation — limitations[1] — “Attempting to assemble the human connectome at the level of single neurons is unrealistic and will remain infeasible at least in the near future.” | limitations[8] |
| for which use | limitation — the same scope as limitations[1] - the proposal is explicitly a compromise on scale — “Attempting to assemble the human connectome at the level of single neurons is unrealistic and will remain infeasible at least in the near future.” | limitations[8] |
| for which use | limitation — everything the proposal rests on methodologically: THE RELATIONSHIP AMONG TENSOR FIELDS, FIBER TRACTS, AND NEURONAL CONNECTIONS REMAINS CONTROVERSIAL — “the relationship among tensor fields, fiber tracts, and neuronal connections remains controversial” | limitations[9] |
| for which use | limitation — the existing human connectional models: THEY ARE LARGELY BASED ON EXTRAPOLATING ANATOMICAL INFORMATION FROM OTHER PRIMATE SPECIES SUCH AS THE MACAQUE MONKEY, AN APPROACH THAT IS PROBLEMATIC — “they are largely based on extrapolating anatomical information from other primate species such as the macaque monkey, an approach that is problematic [ 14 ]” | limitations[10] |
| evidential status | mechanism — the paper's own words: We suggest that — “We suggest that the vast number, high variability, and fast dynamics of individual neurons and synapses render them inappropriate as basic elements for an initial draft of the connectome.” | mechanisms[1] |
| replication | untested | |
| limitation | compiling the connectome faces two significant challenges not shared by other networks — “Compiling the connectome faces two significant challenges not shared by other natural or technological networks.” | limitations[0] |
| limitation | A single-neuron connectome is out of reach for the foreseeable future. — “Attempting to assemble the human connectome at the level of single neurons is unrealistic and will remain infeasible at least in the near future.” | limitations[1] |
| limitation | There is no agreed parcellation scheme for human brain regions, which obstructs building a unified resource. — “No single universally accepted parcellation scheme currently exists for human brain regions (e.g., areas of the cerebral cortex), posing a significant obstacle to creating a unified resource such as the connectome.” | limitations[2] |
| limitation | Interindividual variability and the probabilistic nature of the data may be seen as fundamental weaknesses of the proposal. — “The presence of significant interindividual variability in structural connection patterns, even at the macroscale level, and the fundamentally probabilistic nature of connectivity datasets provided in the connectome may be viewed as fundamental weaknesses of the proposal, undermining its comprehensive goal of a definitive structural description of the human brain.” | limitations[3] |
| limitation | Structure-function relationships will not fall out of the connection data by themselves. — “We emphasize that structure–function relationships are not directly evident from the connectional dataset itself.” | limitations[4] |
| limitation | The envisaged connectome omits many structural and functional aspects of the brain. — “There are many structural and functional aspects that the human connectome, as envisioned in this article, does not contain or address.” | limitations[5] |
| limitation | A first draft would not capture plasticity, nonsynaptic communication, hormonal regulation or degeneration. — “Its first draft does not capture synaptic plasticity and remodeling, nonsynaptic communication, hormonal regulation, or degenerative processes.” | limitations[6] |
| limitation | There is no agreed parcellation of the human cortex, so the nodes of the proposed network are not yet definable. — “No single universally accepted parcellation scheme currently exists for human brain regions (e.g., areas of the cerebral cortex), posing a significant obstacle to creating a unified resource such as the connectome.” | limitations[7] |
| limitation | The single-neuron connectome is stated to be infeasible, so the proposal is explicitly a compromise on scale. — “Attempting to assemble the human connectome at the level of single neurons is unrealistic and will remain infeasible at least in the near future.” | limitations[8] |
| limitation | What diffusion imaging actually measures is still disputed, and it is the method the proposal rests on. — “the relationship among tensor fields, fiber tracts, and neuronal connections remains controversial” | limitations[9] |
| limitation | Existing human connectional models are extrapolated from macaque anatomy, which the paper calls problematic. — “they are largely based on extrapolating anatomical information from other primate species such as the macaque monkey, an approach that is problematic [ 14 ]” | limitations[10] |
finding · quantity · variable · definition · key-term — grounded in the full text| finding | The connection matrix of the human brain, the connectome, is proposed as an indispensable foundation for neurobiological research, but the network of anatomical connections linking its neuronal elements is still largely unknown. — “The connection matrix of the human brain (the human “connectome”) represents an indispensable foundation for basic and applied neurobiological research. However, the network of anatomical connections linking the neuronal elements of the human brain is still largely unknown” | content_tags.findings[0] |
| evidence status note | finding — QUOTE RECOVERED FROM THE PAPER'S OWN CHARACTERS. The recorded quote was LETTERS-ONLY IDENTICAL to this stretch of the staged text but not an exact substring of it - spaces, hyphens, ligature glyphs or mathematical operators differed, so it could never anchor. This is the file's own text between the same two letters: not produced by a rule about the damage, copied out of the file. Round-trip checked (the recovered span is letters-only equal to the original) and length-checked. The text as originally recorded is kept in evidence as extracted. | content_tags.findings[0] |
| finding | A structural connectome would significantly increase understanding of how functional brain states emerge from their structural substrate; network analyses already describe small-world attributes, distinct functional streams and motifs. — “significantly increase our understanding of how functional brain states emerge from their underlying structural substrate... small-world attributes [ 32 , 33 ], distinct functional streams [ 34 ], motifs [ 35 ]” [UNVERIFIED AGAINST THE STAGED TEXT] | content_tags.findings[1] |
| evidence status note | finding — marked quote-not-in-current-staged-text - today's merge gate would REFUSE this quote: Paper.locate() does not find it in the current staged text, and no disclosure on the row explains why. Correction, 15 September 2026: the clause above, 'no disclosure on the row explains why', no longer holds - the composite is now declared in the composite-quote note. The marker stays true: the quote is two passages, not one exact string in the staged text. | content_tags.findings[1] |
| not machine-vouched | finding — not-in-our-copy - The quote is in NEITHER copy we hold - not in the text we tag against and not in the page we fetched ('fulltext.html') - checked with ligatures, hyphens, spacing and punctuation set aside. This body carries no measurable character damage at all. We do not know what it was taken from, and no class is claimed for it. | content_tags.findings[1] |
| finding | The connectome will consist of two main descriptors: neural elements and neural connections. — “The connectome will consist of two main descriptors defining its network architecture: neural elements and neural connections.” | content_tags.findings[2] |
| finding | Brain regions and the pathways between them are perhaps the most feasible level for compiling a first draft of the connectome. — “Nonetheless, anatomically distinct brain regions and inter-regional pathways represent perhaps the most feasible organizational level for compiling a first draft of the human connectome.” | content_tags.findings[3] |
| finding | Computational analyses of existing connection datasets have revealed clusters of brain regions, hierarchical organization and small-world attributes. — “Computational analyses of these datasets have revealed a broad range of network characteristics [ 28 ], including the existence of clusters of brain regions [ 29 ], hierarchical organization [ 30 , 31 ], small-world attributes [ 32 , 33 ]” | content_tags.findings[4] |
| finding | A first draft would be a probabilistic voxel-by-voxel connectivity map of roughly 10^4-10^5 elements and 10^5-10^7 structural connections. — “This first draft of the human connectome would take the form of a probabilistic map of voxel-by-voxel connectivity embedded within standard coordinates containing approximately 10 4 –10 5 elements and approximately 10 5 –10 7 structural connections” | content_tags.findings[5] |
| quantity | 10^11 neurons, 10^15 connections — size of a single-neuron connectome, given as the reason it is infeasible — “the size of the connectome would be several orders of magnitude larger than that of the genome, comprising an estimated 10 11 neurons, with 10 15 connections between them” | content_tags.quantities[0] |
| quantity | 3 × 10^9 base pairs, 20,000-30,000 genes — size of the human genome, the comparison the connectome is scaled against — “The human genome is composed of approximately 3 × 10 9 base pairs, containing around 20,000–30,000 genes [ 15 ].” | content_tags.quantities[1] |
| evidence status note | quantity — FLATTENED EXPONENT IN THIS QUOTE - THE NUMBER IS WRONG BY ORDERS OF MAGNITUDE AS IT STANDS. The staged text reads '3 × 10 9', where the paper writes 10 to the power 9. The superscript was pushed onto the line by the PDF extractor. The quote is verbatim to the staged text and its anchor is sound; the STAGED TEXT is what is damaged. NOT repaired - repair means re-staging the body and re-anchoring the record. | content_tags.quantities[1] |
| quantity | on the order of 100 regions — estimated number of anatomically distinct human cortical areas - and it is unknown — “neurons are arranged in an unknown number of anatomically distinct regions and areas, perhaps on the order of 100 [ 24 ] or more” | content_tags.quantities[2] |
| quantity | 10^4-10^5 elements, 10^5-10^7 connections — projected size of a first-draft macroscale connectome — “This first draft of the human connectome would take the form of a probabilistic map of voxel-by-voxel connectivity embedded within standard coordinates containing approximately 10 4 –10 5 elements and approximately 10 5 –10 7 structural connections” | content_tags.quantities[3] |
| variable | level of description (independent) — “It is important to determine which level of description is the most appropriate for a first draft of the human connectome.” | content_tags.variables[0] |
| definition | connectome — “The connection matrix of the human brain (the human “connectome”) represents an indispensable foundation for basic and applied neurobiological research.” | content_tags.definitions[2] |
| definition | connection matrix A — “The set of all N elements constitutes the columns (targets) and rows (sources) of an N 2 – N connection matrix A, whose a ij entries represent connections between individual elements i and j .” | content_tags.definitions[3] |
| definition | segregated brain region — “What defines a segregated brain region is that all its structural elements share highly similar long-range connectivity patterns, and that these patterns are dissimilar between brain regions.” | content_tags.definitions[4] |
| definition | microscale — “the level of single neurons and synapses (microscale) and the level of anatomically distinct brain regions and inter-regional pathways (macroscale).” | content_tags.definitions[6] |
| definition | macroscale — “the level of single neurons and synapses (microscale) and the level of anatomically distinct brain regions and inter-regional pathways (macroscale).” | content_tags.definitions[7] |
| key-terms | connectome, brain, network, connections, structural, neurons, connectivity, nodes, matrix, connection matrix, neural elements, neural connections, microscale, macroscale, mesoscale, small-world, parcellation, corticothalamic system, functional segregation, functional integration, genome |
| concept | connectome — “The connection matrix of the human brain (the human “connectome”) represents an indispensable foundation for basic and applied neurobiological research.” | concepts[0] |
| concept | connection matrix — not defined in this paper's own words; an N x N array whose entries record the presence or absence of a connection between element i and element j — “The set of all N elements constitutes the columns (targets) and rows (sources) of an N 2 – N connection matrix A, whose a ij entries represent connections between individual elements i and j .” | concepts[4] |
| concept | neural elements — not defined in this paper's own words; the units the connectome is built from - neurons, neuronal populations or brain regions depending on scale; what graph theory calls nodes — “The connectome will consist of two main descriptors defining its network architecture: neural elements and neural connections.” | concepts[5] |
| concept | neural connections — not defined in this paper's own words; the anatomical pathways between neural elements; what graph theory calls edges — “The connectome will consist of two main descriptors defining its network architecture: neural elements and neural connections.” | concepts[6] |
| concept | technological networks — not defined in this paper's own words; engineered networks used as the comparison class for the brain, and from which the paper distances it — “Compiling the connectome faces two significant challenges not shared by other natural or technological networks.” | concepts[7] |
| concept | genome — not defined in this paper's own words; the complete DNA sequence, invoked as the model for what a complete structural inventory looks like and as a contrast in how easily its elements are defined — “The human genome is composed of approximately 3 × 10 9 base pairs, containing around 20,000–30,000 genes [ 15 ].” | concepts[8] |
| concept | parcellation — not defined in this paper's own words; the division of the cortex into anatomically distinct regions, which must be settled before the connectome has nodes — “No single universally accepted parcellation scheme currently exists for human brain regions (e.g., areas of the cerebral cortex), posing a significant obstacle to creating a unified resource such as the connectome.” | concepts[9] |
| concept | macroscale — not defined in this paper's own words; the level of anatomically distinct brain regions and the pathways between them — “At opposite ends of the scale are the level of single neurons and synapses (microscale) and the level of anatomically distinct brain regions and inter-regional pathways (macroscale).” | concepts[10] |
| concept | microscale — the level of single neurons and synapses — “At opposite ends of the scale are the level of single neurons and synapses (microscale) and the level of anatomically distinct brain regions and inter-regional pathways (macroscale).” | concepts[11] |
| concept | small-world — not defined in this paper's own words; a network property already found in the mammalian connection matrices this proposal builds on — “Computational analyses of these datasets have revealed a broad range of network characteristics [ 28 ], including the existence of clusters of brain regions [ 29 ], hierarchical organization [ 30 , 31 ], small-world attributes [ 32 , 33 ]” | concepts[12] |
| concept | connectivity patterns — not defined in this paper's own words; the profile of long-range connections shared by the elements of one region, used to delineate that region — “What defines a segregated brain region is that all its structural elements share highly similar long-range connectivity patterns, and that these patterns are dissimilar between brain regions.” | concepts[13] |
| concept | brain region — not defined in this paper's own words; an anatomically segregated part of the cortex, proposed as the node of the first-draft connectome — “What defines a segregated brain region is that all its structural elements share highly similar long-range connectivity patterns, and that these patterns are dissimilar between brain regions.” | concepts[14] |
| concept | neuroinformatics — not defined in this paper's own words; the discipline associated with the proliferation, organization, and use of neuroscience datasets — “Recent research in neuroscience has resulted in a rapid proliferation of neuroscience datasets and in the arrival of a new discipline, neuroinformatics [ 1 – 4 ].” | concepts[15] |
| concept | functional segregation — not defined in this paper's own words; one of the two cortical organizational principles used to conceptualize cognition as a network phenomenon — “A potential theoretical framework for conceptualizing cognition as a network phenomenon is based on two main organizational principles found in the cerebral cortex, functional segregation, and functional integration [ 6 , 7 ].” | concepts[16] |
| concept | functional integration — not defined in this paper's own words; one of the two cortical organizational principles used to conceptualize cognition as a network phenomenon — “A potential theoretical framework for conceptualizing cognition as a network phenomenon is based on two main organizational principles found in the cerebral cortex, functional segregation, and functional integration [ 6 , 7 ].” | concepts[17] |
| concept | mesoscale — not defined in this paper's own words; the intermediate organizational level of neuronal groups or populations, between single neurons and brain regions — “Different kinds of structural descriptions could target at least three rather distinct levels of organization. At opposite ends of the scale are the level of single neurons and synapses (microscale) and the level of anatomically distinct brain regions and inter-regional pathways (macroscale). Between these two levels is the level of neuronal groups or populations (mesoscale).” | concepts[18] |
| concept | cortical minicolumn — not defined in this paper's own words; a proposed basic functional unit of mammalian cerebral cortex and an example of a local neuronal processing population — “Mountcastle [ 54 , 55 ] originally proposed the cortical minicolumn as a basic functional unit of mammalian cerebral cortex.” | concepts[19] |
| concept | elementary processing units — not defined in this paper's own words; local neuronal populations, exemplified by cortical minicolumns, whose connection patterns form a finer description than brain regions — “A further step in this direction involves the characterization of connection patterns among elementary processing units, corresponding to local populations of neurons such as cortical minicolumns.” | concepts[20] |
| concept | probabilistic tractography — not defined in this paper's own words; tractography following diffusion-weighted imaging, used with connectivity-profile correlations to parcellate cortical regions — “Step 1 is to perform diffusion-weighted imaging followed by probabilistic tractography of thalamocortical tracts as well as corticocortical interareal pathways, using correlations in connectivity profiles to assist in parcellating human cortical regions. The end result is a voxel-wise probabilistic all-to-all structural connectivity matrix for the human brain.” | concepts[21] |
| concept | structural connectivity matrix — not defined in this paper's own words; a voxel-wise probabilistic all-to-all matrix produced from diffusion imaging and tractography — “Step 1 is to perform diffusion-weighted imaging followed by probabilistic tractography of thalamocortical tracts as well as corticocortical interareal pathways, using correlations in connectivity profiles to assist in parcellating human cortical regions. The end result is a voxel-wise probabilistic all-to-all structural connectivity matrix for the human brain.” | concepts[22] |
| concept | functional connectivity matrix — not defined in this paper's own words; a voxel-wise all-to-all matrix derived from correlations in registered resting or task-related activity — “Step 2 is to perform a correlation analysis of spatially registered, equally resolved resting activity and/or multistimulus/multitask activation data (functional magnetic resonance imaging and/or magnetoencephalography) recorded in the same person [ 73 ], emphasizing strong functional relationships that are consistent across tasks [ 74 ]. The end result is a voxel-wise all-to-all functional connectivity matrix for the human brain.” | concepts[23] |
| concept | cross-level integration — not defined in this paper's own words; integration of complementary macroscale and minicolumn datasets into a single mesoscale connectome — “Thus, macroscale and minicolumn descriptions deliver complementary datasets that need cross-level integration to achieve a single mesoscale version of the connectome.” | concepts[24] |
| concept | structure–function relationships — not defined in this paper's own words; relationships not directly visible in the connectional dataset and requiring further empirical and computational study — “We emphasize that structure–function relationships are not directly evident from the connectional dataset itself. Rather, their elucidation will require further intense empirical and computational study.” | concepts[25] |
| concept | structural benchmarks — not defined in this paper's own words; connectome-derived reference structures for comparing and integrating specialized computational neuroscience models — “If the connectome is sufficiently comprehensive and accessible, it could also provide a set of structural benchmarks that might facilitate the comparison and integration of the numerous specialized and structurally based models that have already been proposed in computational neuroscience.” | concepts[26] |
| correlation cross-correlation matrix | — “Binarized connection patterns were then used to calculate a cross-correlation matrix, which was examined for the existence of distinct clusters of voxels with shared connection patterns.” | content_tags.statistics[0] |
| probability P | — “First, connection probabilities from voxels within the medial frontal cortex to all other voxels in the rest of the brain were obtained.” | content_tags.statistics[1] |
| claim type | method — a second reading reaching the same value; recorded: method — “We suggest that the correlated use of noninvasive structural and functional imaging methods offers the most promising experimental route toward the human connectome.” | alternatives.claim_type |
| method | review — a second reading reaching the same value; recorded: review — “We propose a research strategy to achieve this goal, and discuss its potential impact.” | alternatives.method |
| phenomenon | the absence of a structural connection map of the human brain, and what building one would require — a second reading DIFFERING from the recorded value; recorded: structural connectivity map of the human brain (the connectome) — “the network of anatomical connections linking the neuronal elements of the human brain is still largely unknown” | alternatives.phenomenon |
Full text — tagged
The reference list is not part of the text we hold for this paper. Our copy was extracted from the publisher's page and the bibliography did not survive that extraction, so the article below ends before its references. Nothing has been removed by us.
The Human Connectometerm·dom: A Structuralterm·key Description of the Human Brainterm·key Olaf SpornspersName Olaf SpornspersName *, a , Giulio TononipersName Giulio TononipersName a , Rolf KötterpersName Rolf KötterpersName a Olaf SpornspersName is in the Department of Psychology and Programs in Neuroscience and Cognitive Science, Indiana University, Bloomington, Indiana, United States of America. Giulio TononipersName is in the Department of Psychiatry, University of Wisconsin, Madison, Wisconsin, United States of America. Rolf KötterpersName is at C. and O. Vogt Brainterm·key Research Institute and Institute of Anatomy II, Heinrich Heine University, Düsseldorf, Germany.
✉ *To whom correspondence should be addressed. E-mail: osporns@indiana.edu Issue date 2005 Sep.
Copyright: © 2005 SpornspersName et al.
This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
PMCID: PMC1239902 PMID: 16201007 ABSTRACT The connection matrixterm·mth of the human brainterm·key (the human “connectometerm·dom”) represents an indispensable foundation for basic and applied neurobiological research. However, the networkterm·key of anatomical connectionsterm·key linking the neuronal elements of the human brainterm·key is still largely unknown. While some databases or collations of large-scale anatomical connectionterm·key patterns exist for other mammalian species, there is currently no connection matrixterm·mth of the human brainterm·key, nor is there a coordinated research effort to collect, archive, and disseminate this important information. We propose a research strategy to achieve this goal, and discuss its potential impact.
Introduction To understand the functioning of a networkterm·key, one must know its elements and their interconnections. The purpose of this article is to discuss research strategies aimed at a comprehensive structuralterm·key description of the networkterm·key of elements and connectionsterm·key forming the human brainterm·key. We propose to call this dataset the human “connectometerm·dom,” and we argue that it is fundamentally important in cognitive neuroscience and neuropsychology. The connectometerm·dom will significantly increase our understanding of how functional brainterm·key states emerge from their underlying structuralterm·key substrate, and will provide new mechanistic insights into how brainterm·key function is affected if this structuralterm·key substrate is disrupted. It will provide a unified, time-invariant, and readily available neuroinformaticsterm resource that could be used in virtually all areas of experimental and theoretical neuroscience.
Recent research in neuroscience has resulted in a rapid proliferation of neuroscience datasets and in the arrival of a new discipline, neuroinformaticsterm [ 1 – 4 ]. Despite considerable advances in experimental techniques and computational paradigms, we still have an incomplete understanding of how human cognitive function emerges from neuronal structure and dynamics. Here, we focus on the relative lack of information about the different types of neural elementsterm·dom and their neural connectionsterm·dom in the human brainterm·key. While a larger number of anatomical studies of the human brainterm·key have been carried out at the macroscopic (cerebral lobes, surface landmarks, and white matter tracts) or microscopic (cytoarchitectonics, myeloarchitectonics, chemoarchitectonics, etc.) anatomical level, there is virtually no information on the finer connectivity patternsterm·mth, including neuronal connectionterm·key densities or laminar projection patterns in relation to anatomically segregated cortical areas or intraregional differentiation. Furthermore, none of the available information is deposited in a single standardized data format, nor can it be accessed through a public database.
Experimental approaches to human cognition have been significantly enhanced by the arrival of functional neuroimaging [ 5 ], a set of techniques that can be applied to study a broad range of cognitive functions, with ever-increasing spatial and temporal resolution. But the mechanistic interpretation of neuroimaging data is limited, in part due to a severe lack of information on the structure and dynamics of the networksterm·key that generate the observed activation patterns. A potential theoretical framework for conceptualizing cognition as a networkterm·key phenomenon is based on two main organizational principles found in the cerebral cortex, functional segregationterm, and functional integrationterm [ 6 , 7 ]. Emerging networkterm·key theories of cognition emphasize the contextual [ 8 ], distributed [ 9 ], dynamic [ 10 ], and degenerate [ 11 , 12 ] nature of structure–function mappings in the brainterm·key. To successfully map structure to function in the human brainterm·key, we urgently need a comprehensive, detailed structuralterm·key model of neuronal units and their connectionsterm·key. Connectional models of the human brainterm·key are scarce and poorly defined [ 13 ], and they are largely based on extrapolating anatomical information from other primate species such as the macaque monkey, an approach that is problematic [ 14 ], in part, because of our incomplete understanding of evolutionary changes.
Our proposal to assemble the human connectometerm·dom has several components. First, we attempt to define a level of scale at which a first draft of the human connectometerm·dom might be assembled. We consider several potential experimental and neuroinformaticsterm approaches for creating this first draft. We then discuss potential problems and limitations of the connectometerm·dom, including the central issues of individual variability and development. Finally, we sketch out the potential impact of the connectometerm·dom in computational and cognitive neuroscience.
Scales and Levels of Structuralterm·key Description The human genometerm·dom is composed of approximately 3 × 10 9 base pairs, containing around 20,000–30,000 genes [ 15 ]. The compilation of the genometerm·dom was aided by the fact that base pairs and genes are relatively straightforward choices as basic structuralterm·key elements. The connectometerm·dom will consist of two main descriptors defining its networkterm·key architecture: neural elementsterm·dom and neural connectionsterm·dom. Data fields for these elements specify superordinate or subordinate structures, a normalized position within a standard coordinate system, and additional parameters such as physiological/biophysical metadata that are crucial for specifying neural dynamics. The set of all N elements constitutes the columns (targets) and rows (sources) of an N 2 – N connection matrix Aterm·def, whose a ij entries represent connectionsterm·key between individual elements i and j . In keeping with conventions adopted by other authors [ 16 , 17 ], confirmed absence of a connectionterm·key is denoted by a ij = 0, while confirmed presence of a connectionterm·key (irrespective of its strength or physiological characteristics) results in a ij = 1. Once a connectionterm·key is confirmed to be present, its nonzero matrixterm·key element receives additional data entries cataloguing a range of structuralterm·key and physiological parameters, such as fiber direction, connectionterm·key density, strength, sign (excitatory/inhibitory), conduction delay, potential modulatory effects (voltage dependence), etc. The union of the binary connection matrixterm·mth and connection-specific physiological data then results in a structuralterm·key description that combines connectionterm·key topology and biophysics.
Compiling the connectometerm·dom faces two significant challenges not shared by other natural or technological networksterm·dom. First, the human brainterm·key is a highly complex organ with a great number of structurally distinct, heterogeneous, yet interconnected components. Because a primary application of the connectometerm·dom will likely be a structuralterm·key substrate for understanding human cognitive function and interpreting neuroimaging studies, a first draft of the connectometerm·dom may focus on the structuralterm·key description of the corticothalamic systemterm·key, including all regions of the cortex and their associated thalamic nuclei. Extensions of this first draft might include additional connected regions (striatum, cerebellum, etc.), with the ultimate aim of compiling the connectometerm·dom of the whole brainterm·key.
A second challenge is that basic structuralterm·key elements of the human brainterm·key, in terms of networkterm·key nodesterm·key and connectionsterm·key, are difficult to define. Different kinds of structuralterm·key descriptions could target at least three rather distinct levels of organization. At opposite ends of the scale are the level of single neuronsterm·key and synapses (microscaleterm·mth) and the level of anatomically distinct brain regionsterm·dom and inter-regional pathways (macroscaleterm·mth). Between these two levels is the level of neuronal groups or populations (mesoscaleterm). It is important to determine which level of descriptionterm·var is the most appropriate for a first draft of the human connectometerm·dom.
Microscaleterm·mth: Single Neuronsterm·key and Synapses Attempting to assemble the human connectometerm·dom at the level of single neuronsterm·key is unrealistic and will remain infeasible at least in the near future. With single neuronsterm·key as the basic element, the size of the connectometerm·dom would be several orders of magnitude larger than that of the genometerm·dom, comprising an estimated 10 11 neuronsterm·key, with 10 15 connectionsterm·key between them (approximately 10 10 neuronsterm·key and 10 13 connectionsterm·key in the cortex alone) [ 18 , 19 ]. Recording or tracing 10 15 connectionsterm·key is not only technically impossible, it may also be unnecessary. While a genomic mutation in a single base pair can have dramatic consequences, alterations of single synapses or cells have not been shown to have similar macroscopic effects. Instead, there is overwhelming evidence that human cognitive functions depend on the activity and coactivity of large populations of neuronsterm·key in distributed networksterm·key, including the corticothalamic systemterm·key [ 20 ]. Furthermore, individual neuronsterm·key and connectionsterm·key are subject to rapid plastic changes. These changes include synaptic weights as well as structuralterm·key remodeling of dendritic spines and presynaptic boutons [ 21 ], possibly switching synaptic connectionsterm·key between large numbers of potential synaptic sites [ 22 , 23 ]. We suggest that the vast number, high variability, and fast dynamics of individual neuronsterm·key and synapses render them inappropriate as basic elements for an initial draft of the connectometerm·dom.
Macroscaleterm·mth: Brain Regionsterm·dom and Pathways An advantage of single neuronsterm·key is that the elements themselves are relatively easily demarcated and well defined. In contrast, brainterm·key areas and neuronal populations are more difficult to delineate. No single universally accepted parcellationterm·mth scheme currently exists for human brain regionsterm·dom (e.g., areas of the cerebral cortex), posing a significant obstacle to creating a unified resource such as the connectometerm·dom. In the human cerebral cortex, neuronsterm·key are arranged in an unknown number of anatomically distinct regions and areas, perhaps on the order of 100 [ 24 ] or more. Different subdivisions of the human brainterm·key (e.g., brainterm·key stem, thalamus, cerebellum, or cortex) may require different criteria for parcellationterm·mth.
Nonetheless, anatomically distinct brain regionsterm·dom and inter-regional pathways represent perhaps the most feasible organizational level for compiling a first draft of the human connectometerm·dom. Several neuroinformaticsterm resources recording large-scale connectionterm·key patterns in the cerebral cortex of various mammalian species already exist, for example, for most cortical regions of the macaque monkey [ 16 , 17 , 25 ], cat [ 26 ], and rat [ 27 ]. Computational analyses of these datasets have revealed a broad range of networkterm·key characteristics [ 28 ], including the existence of clusters of brain regionsterm·dom [ 29 ], hierarchical organization [ 30 , 31 ], small-worldterm·dom attributes [ 32 , 33 ], distinct functional streams [ 34 ], motifs [ 35 ], and areal contributions to global networkterm·key measures [ 36 ].
A broad range of experimental approaches exist at the macroscaleterm·mth. Cerebral white matter has traditionally been taken as a marker of the amount of connectivityterm·key within a cortical system. The relative contribution of cerebral white matter has increased throughout phylogeny to such an extent that its volume and metabolic requirements may present a limitation to further increases in connectional complexity [ 37 ]. The structuralterm·key organization of white matter has been investigated by dissection, histological staining, degeneration methods, and axonal tracing [ 38 ]. Axonal tracing methods are the main basis for existing mammalian large-scale connectionterm·key matrices [ 16 , 25 – 27 ], and their systematic compilation is currently being refined [ 17 , 39 ] to extract more sophisticated data, perhaps with the help of automated text analysis. In the human brainterm·key, postmortem tracing with carbocyanine dyes has provided details of connectivityterm·key within and between adjacent areas [ 40 , 41 ]. However, because of the slow transport and the length of fibers in the human brainterm·key, this method fails to reveal more remote connectionsterm·key. Another tracing method employs the in vivo detection by magnetic resonance imaging of high-contrast rare earth ions (e.g., manganese [ 42 ]) that have been injected into fiber tracts or inhaled, and taken up by neuronsterm·key. The invasive nature and potential toxicity of the procedure makes it an unlikely candidate for human connectivityterm·key analyses.
Recent noninvasive imaging methods (diffusion tensor imaging [DTI] in its several variants, commonly followed by computational tractography) have been shown to produce results that are consistent with known pathways formed by major fiber tracts in the human brainterm·key [ 43 – 45 ], although there continues to be some limitations in data acquisition and processing algorithms [ 46 ]. To disambiguate signals produced by crossing or intersecting fibers, advances in diffusion imaging technology may allow the resolution of multiple axonal pathways within a single image voxel [ 47 , 48 ]. Despite the promise of diffusion imaging, a systematic atlas of DTI-based neuroanatomy, including probabilistic data gathered from individual brainsterm·key, has not yet been produced, and the relationship among tensor fields, fiber tracts, and neuronal connectionsterm·key remains controversial. This controversy is likely to be resolved only through a combination of anatomical tracing techniques with noninvasive diffusion and functional imaging.
Perhaps the most promising avenue for compiling the human connectometerm·dom originates from the notion that individual brain regionsterm·dom maintain individual connectionterm·key profiles. What defines a segregated brain regionterm·def is that all its structuralterm·key elements share highly similar long-range connectivity patternsterm·mth, and that these patterns are dissimilar between brain regionsterm·dom. These connectivity patternsterm·mth determine the region's functional properties [ 49 ], and also allow their anatomical delineation and mapping. Diffusion imaging can be used to identify borders between cortical areas [ 50 ], most clearly on the basis of differences in long-range connectionsterm·key to the thalamus [ 51 , 52 ]. The idea that patterns of connectivityterm·key can be used to identify areal boundaries has also been tested in a combination of functional imaging and DTI/tractography in the human medial frontal cortex [ 53 ]. First, connectionterm·key probabilities from voxels within the medial frontal cortex to all other voxels in the rest of the brainterm·key were obtained. Binarized connectionterm·key patterns were then used to calculate a cross-correlation matrixterm·key, which was examined for the existence of distinct clusters of voxels with shared connectionterm·key patterns. Such clusters were taken to represent anatomically segregated areas corresponding to human supplementary motor area and pre- supplementary motor area, respectively. Functional mapping revealed close correspondence between DTI and functional activation patterns. While this combined structuralterm·key–functional approach is computationally intensive, nothing prevents its application to the entire corticothalamic systemterm·key. We suggest that the correlated use of noninvasive structuralterm·key and functional imaging methods offers the most promising experimental route toward the human connectometerm·dom.
Structuralterm·key connectionterm·key data obtained using this combined methodology can in principle be validated by histological or tracing methods. Likely, no single method will turn out to be sufficiently powerful or comprehensive. A systematic application of sophisticated diffusion-weighted imaging combined with spatially registered high-resolution anatomical or spectroscopic imaging, regional activation, and coactivation data (e.g., electroencephalography, magnetoencephalography) obtained within the same individual subject, may offer the most feasible strategy for mapping the human connectometerm·dom at the macroscaleterm·mth. This first draft of the human connectometerm·dom would take the form of a probabilistic map of voxel-by-voxel connectivityterm·key embedded within standard coordinates containing approximately 10 4 –10 5 elements and approximately 10 5 –10 7 structuralterm·key connectionsterm·key. It would not only provide information on the large-scale connectionterm·key patterns within the corticothalamic systemterm·key, but also on parcellationterm·mth of human cortex into distinct areas based on a combination of structuralterm·key and functional data in the same individual. Since this connectivityterm·key matrixterm·key is voxel-based, it can be cross-referenced with existing reference templates and with population-based brainterm·key atlases.
Mesoscaleterm: Minicolumns and Their Connectionterm·key Patterns A first draft of the corticothalamic connectometerm·dom at the macroscaleterm·mth might provide a comprehensive dataset comprising several hundred brain regionsterm·dom and thousands of pathways, but it does not incorporate information on functional subdivisions or segregated subcircuits within each brain regionterm·dom. While such a macroscaleterm·mth description will provide many fundamental insights into the large-scale organization of human cortex, it is an insufficient basis for a complete understanding of the human brainterm·key's functional dynamics and information processing capacities. A further step in this direction involves the characterization of connectionterm·key patterns among elementary processing unitsterm, corresponding to local populations of neuronsterm·key such as cortical minicolumnsterm. Mountcastle [ 54 , 55 ] originally proposed the cortical minicolumnterm as a basic functional unit of mammalian cerebral cortex. While details of minicolumn architecture are likely to vary across different brain regionsterm·dom [ 56 – 58 ], minicolumns generally contain approximately 80–100 neuronsterm·key, spanning all cortical layers, with a diameter of approximately 30–50 μm [ 55 ]. Minicolumns may possess relatively stereotypic internal processing, and maintain generic patterns of inputs and outputs with minicolumns in other regions [ 56 – 59 ].
Recent studies have provided evidence for functionally specialized and precisely wired subnetworks of neuronsterm·key coexisting within single cortical columns [ 60 , 61 ]. These studies have shown that cortical columns may contain segregated subnetworks, corresponding to minicolumns, which promote greater intracolumnar functional independence and informational heterogeneity. The members of such subnetworks are selectively interconnected with each other, indicating that connectionsterm·key within and between minicolumns follow more complex rules than simple random patterns or Gaussian (distance-dependent) connectionterm·key profiles.
Minicolumns may be a sensible choice for neural elementsterm·dom at the mesoscaleterm of the connectometerm·dom because they may represent basic functional elements that are crucial for cortical information processing. While tracing or recording all minicolumns in an individual brainterm·key is still impossible, it may be feasible to collect data on minicolumn anatomy for selected brain regionsterm·dom, which can then be “fit” into the appropriate positions within the macroscaleterm·mth connection matrixterm·mth. Functional imaging at columnar resolution has been carried out in animal experiments using high field strength [ 62 ], and may be facilitated in the future through selective imaging of fast and spatially precise capillary cerebral blood flow response components [ 63 ]. Especially important for determining functional responses of brain regionsterm·dom are connectionterm·key patterns between each region's constituent elements. Axonal tracing methods have revealed specific patterns of horizontal connectionsterm·key between individual cells and cell populations within a brain regionterm·dom, which are often found to preferentially link cells with similar response characteristics [ 64 , 65 ], resulting in clustered intra-areal connectivityterm·key. Such patterns might also be accessible in the human brainterm·key [ 40 ], and regional variations in such patterns may provide important clues regarding the way in which information is processed within each region.
The axonal tracing approach delivers minicolumn maps for each distinct brain regionterm·dom, including information on their functional segregationterm and local (intra-areal) interconnectivity. While parcellated brain regionsterm·dom at the macroscaleterm·mth can be identified across individuals, we have no means to resolve the locations of corresponding minicolumns across different brainsterm·key. The mapping of smaller-than-macroscopic units, therefore, requires coordinate-independent mapping approaches [ 66 ], which preserve the anatomical relationships and basic physiological properties of these units. Thus, macroscaleterm·mth and minicolumn descriptions deliver complementary datasets that need cross-level integrationterm to achieve a single mesoscaleterm version of the connectometerm·dom. Effectively, minicolumn maps need to be mapped onto brain regionterm·dom–specific voxel sets rather than individual voxels, with voxel sets providing spatial embedding and probabilistic long-range connectionsterm·key, and minicolumn maps providing local connectivityterm·key, processing, and coding information. The minicolumn description provides a functionally heterogeneous architecture that is unique to each parcellated brain regionterm·dom, with specific (probabilistic) patterns of intra-areal and interareal minicolumn connectionsterm·key. A crucial task will be to convert long-range, nondirected voxel-by-voxel connectivityterm·key into directed, functionally heterogeneous long-range minicolumn interconnectivity. This requires intermediate descriptive levels such as stripes, bands, and blobs in the early visual system, for which specific connectionterm·key patterns have been demonstrated across areal boundaries. When accomplished, this cross-level integrationterm will result in a mesoscaleterm connection matrixterm·mth of the human brainterm·key that might comprise as many as 10 7 –10 8 structuralterm·key elements (comparable in size to the 2005 World Wide Web), with minicolumn elements directly cross-referenced to voxels in the macroscaleterm·mth map.
Individual Variability and Development The large-scale connectivityterm·key structure of the brainterm·key above the synaptic level represents a relatively invariant characteristic of our species. Once the elements and connectionsterm·key in the human brainterm·key are recorded, this dataset will remain stable, essentially forever. However, as in the case of the genometerm·dom, the precise combinations and patterns of elements and their connectionsterm·key exhibit significant variations between individuals, at all levels of scale. Some individual variations may be due to genetic differences, others may be the result of developmental and experiential history, gender differences, pathologies, or responses to injury. To complicate matters further, the human connectometerm·dom undergoes development through time, from early stages of the embryo to adolescence to adult age. Incorporating individual variations and developmental stages is absolutely crucial in rendering the connectometerm·dom an effective resource.
Anatomical and imaging studies have revealed significant interindividual variability in the size and location of brainterm·key areas, as well as in the relationship between macrostructures (e.g., the cortical gyrification pattern) and microstructures (e.g., cytoarchitectonics and cytochemistry). Statistical analysis of variations in macroscopic surface features of the human brainterm·key, for example its sulcal geometry [ 67 ], demonstrates the extent to which even large-scale features of cortical morphology vary between individuals, possibly as a result of genetic differences [ 68 ]. An anatomical study of Broca's area in ten postmortem human brainsterm·key revealed significant variations in size as well as in the area's relation to sulcal landmarks [ 69 ]. Applying structuralterm·key magnetic resonance imaging to map the boundaries of the planum temporale has demonstrated significant variations in its size and position across 50 individuals [ 70 ]. Functional neuroimaging studies have also revealed significant interindividual differences, for example, in the location and extent of area MT/V5 [ 71 ] and other visual cortical areas [ 72 ]. These functional differences are presumably due to variations in underlying structuralterm·key (cytoarchitectonic and connectional) substrates.
The presence of significant interindividual variability in structuralterm·key connectionterm·key patterns, even at the macroscaleterm·mth level, and the fundamentally probabilistic nature of connectivityterm·key datasets provided in the connectometerm·dom may be viewed as fundamental weaknesses of the proposal, undermining its comprehensive goal of a definitive structuralterm·key description of the human brainterm·key. However, we should consider the fact that there is also clear interindividual variability in the human genometerm·dom. Nevertheless, the first draft with a DNA sequence obtained from cells from only a few individuals [ 15 ] has proven immensely useful for gaining insights into general organizational features of the human genometerm·dom. Mapping of interindividual variability in the connectometerm·dom is a necessary further step, but does not detract from the potential insights gained from a first draft that does not yet systematically incorporate these differences.
Steps Toward the Human Connectometerm·dom: Its Compilation, Assembly, and Integration Based on a combination of functional and diffusion-weighted imaging, we outline the following five-step strategy for compiling a first draft of the human connectometerm·dom at the macroscaleterm·mth.
Step 1 is to perform diffusion-weighted imaging followed by probabilistic tractographyterm of thalamocortical tracts as well as corticocortical interareal pathways, using correlations in connectivityterm·key profiles to assist in parcellating human cortical regions. The end result is a voxel-wise probabilistic all-to-all structural connectivity matrixterm for the human brainterm·key. Step 2 is to perform a correlation analysis of spatially registered, equally resolved resting activity and/or multistimulus/multitask activation data (functional magnetic resonance imaging and/or magnetoencephalography) recorded in the same person [ 73 ], emphasizing strong functional relationships that are consistent across tasks [ 74 ]. The end result is a voxel-wise all-to-all functional connectivity matrixterm for the human brainterm·key. Step 3 is to perform a cluster analysis of correspondences between the structuralterm·key and functional connectivity matrixterm obtained under steps 1 and 2, with the goal of identifying regions of consistent structure–function relationshipsterm in the human brainterm·key, possibly involving indirect projections [ 75 ]. Step 4 is to compare the results obtained by cluster analysis (step 3) with macaque data in order to identify correspondences (e.g., in visuomotor pathways) and deviations (e.g., in structures such as the fasciculus arcuatus). Step 5 is to validate the strongest predictions generated by assembling the final combined structuralterm·key–functional connectivity matrixterm using custom-designed stimuli and perturbational techniques such as transcranial magnetic stimulation.
The following three steps represent additional stages designed to further test and verify the human connectometerm·dom, including population analyses of individual variability and pathology. Step 6 is to perform a population analysis of healthy subjects and spatial registration to standard brainterm·key coordinates for probabilistic statements about data from steps 1–5. Step 7 is to compare population data on clustered brain regionsterm·dom to histologically identified regions in a probabilistic human brainterm·key atlas to assess correspondence. Step 8 is to compare population data between healthy individuals and patient groups with specific pathologies, to identify connectional differences.
Ultimately, the connectometerm·dom will likely describe connectivity patternsterm·mth at multiple levels of scale, for example, by incorporating linkages between the macroscaleterm·mth of brain regionsterm·dom and pathways in more elementary mesoscaleterm functional units such as minicolumns and their patterns of interconnectivity. As experimental techniques mature, the connectometerm·dom will gradually evolve through different stages of assembly as it is refined, updated, cross-validated, and “filled in.” Standardization of parcellationterm·mth schemes, elimination of unreliable data, and incorporation of additional structuralterm·key levels above and below the one chosen for the initial draft will drive this effort. An additional driving force is the continued innovation in data acquisition and analysis techniques, particularly in diffusion-weighted imaging and tractography, which will result in progressive revision, refinement, and extension of the connectometerm·dom. To become a useful research tool, the connectometerm·dom must be linked to other databases (compiled in parallel efforts) that record additional information mapped across the human brainterm·key, such as receptor distributions or gene expression patterns. We expect that assembling even the first draft of the connectometerm·dom will require significant experimental and computational resources over an extended period of time. Compilation, assembly, and integration efforts are likely to be extensive tasks, requiring large-scale collaboration, coordinated data collection and dissemination, and the establishment of new computational methods, data standards, and mechanisms for controlled validation and quality assurance.
Conclusions: The Potential Impact of the Connectometerm·dom How can the connectometerm·dom be used to map brainterm·key structure to function? The step from structure to function is essential for understanding how cognitive processes emerge from their morphological substrates. Our central motivating hypothesis is that the pattern of elements and connectionsterm·key as captured in the connectometerm·dom places specific constraints on brainterm·key dynamics, and thus shapes the operations and processes of human cognition. In turn, data recording the activity of the human brainterm·key in combination with the structuralterm·key model provided by the connectometerm·dom will help to discern causal interactions in large-scale brain networksterm·dom (e.g., [ 76 – 78 ]). We emphasize that structure–function relationshipsterm are not directly evident from the connectional dataset itself. Rather, their elucidation will require further intense empirical and computational study. Depending on sensory input, global brainterm·key state, or learning, the same structuralterm·key networkterm·key can support a wide range of dynamic and cognitive states. This should not, however, discourage the effort to assemble the connectometerm·dom. Similar difficulties in mapping structure to function exist in the case of the genometerm·dom, although they generally were not foreseen when the project was initiated. Highly complex transcriptional regulatory networksterm·key, signaling pathways, mechanical forces, and elaborate mechanisms of gene expression all play essential roles in translating base sequences into functioning cells, tissues, and organisms. Both genometerm·dom and connectometerm·dom constitute complex networksterm·key [ 79 ], whose functions are not fully understood even if their structuralterm·key substrates are fully catalogued.
An obvious and related use of the human connectometerm·dom would be providing structuralterm·key information that can be implemented as part of large-scale computational models [ 80 , 81 ]. If the connectometerm·dom is sufficiently comprehensive and accessible, it could also provide a set of structural benchmarksterm that might facilitate the comparison and integration of the numerous specialized and structurally based models that have already been proposed in computational neuroscience. Drawing upon human connectional datasets would help ground modeling efforts aimed specifically at brainterm·key mechanisms of human cognitive function (e.g., language). Other computational applications of the connectometerm·dom are topological analyses of networkterm·key structure [ 28 ], perturbational studies aiming at mapping structuralterm·key disruption to functional defects, and synthetic brainterm·key imaging [ 82 – 84 ].
The human connectometerm·dom could potentially have a major impact on our understanding of brainterm·key damage and subsequent recovery. The effects of developmental variations or abnormalities, traumatic brainterm·key injury, or neurodegenerative disease can all be captured as specific structuralterm·key variants of the human connectometerm·dom. The functional consequences of networkterm·key perturbations will allow a better understanding of structuralterm·key causes of dysfunction, and may permit the design of strategies for recovery based on networkterm·key analysis. Understanding the basic networkterm·key causes of brainterm·key diseases may also open new avenues for therapy and prevention by harnessing inherent networkterm·key mechanisms that ensure robustness and compensation.
There are many structuralterm·key and functional aspects that the human connectometerm·dom, as envisioned in this article, does not contain or address. For example, it does not explicitly capture or catalogue the rich variety of neuronal morphologies, the diversity of physiological and biochemical neural subtypes, glial cells, or brainterm·key vascularization. Its first draft does not capture synaptic plasticity and remodeling, nonsynaptic communication, hormonal regulation, or degenerative processes. We do not view these issues as shortcomings that fundamentally undermine the usefulness or potential impact of the connectometerm·dom. Rather, they illustrate the open-ended, integrative nature of the proposed research effort. The connectometerm·dom will represent a foundational resource and a central reference point for a broad range of specialized databases [ 85 ], thus making federating these databases more effective. Most importantly, the connectometerm·dom will provide an important tool for mechanistic modeling and interpretation of human functional brainterm·key data.
Acknowledgments OS was supported by National Institutes of Health National Institute on Drug Abuse grant 1R21DA15647–01. RK was supported by the Deutsche Forschungsgemeinschaft (Graduate School 320 and KO 1560/6–2). Core ideas were developed at a workshop sponsored by the James S. McDonnell Foundation. The authors thank M. Arbib, K. Friston, R. McIntosh, B. Horwitz, and T. Behrens for comments and suggestions.
Abbreviation DTI diffusion tensor imaging