YNIMG-10548; No.

of pages: 12; 4C: 5, 6, 8

NeuroImage xxx (2013) xxxxxx

Contents lists available at SciVerse ScienceDirect

NeuroImage
journal homepage: www.elsevier.com/locate/ynimg

Review

Connectomic approaches before the connectome

Marco Catani a,, Michel Thiebaut de Schotten a,b, David Slater c, Flavio Dell'Acqua c,d
a

Natbrainlab, King's College London, Institute of Psychiatry, Department of Forensic and Neurodevelopmental Sciences, London SE5 8AF, UK
UMR_S 975, CNRS UMR 7225, Centre de Recherche de l'Institut du Cerveau et de la Moelle pinire, Groupe Hospitalier Piti-Salptrire, 75013 Paris, France
c
Natbrainlab, King's College London, Institute of Psychiatry, Department of Neuroimaging, London SE5 8AF, UK
d
NIHR Biomedical Research Centre for Mental Health at South London and Maudsley NHS Foundation Trust and Institute of Psychiatry, King's College London, UK
b

a r t i c l e

i n f o

Article history:
Accepted 20 May 2013
Available online xxxx
Keywords:
Connectome
Connection
Diffusion
Tractography
Hodology
Networks

a b s t r a c t
Connectome is a term with a short history but a long past. Since the origins of neuroscience the concept of a
map of neural connections has been a constant inspiring idea for those who believed the brain as the organ
of intellect. A myriad of proto-connectome maps have been produced throughout the centuries, each one
reecting the theory and method of investigation that prevailed at the time. Even contemporary denitions
of the connectome rest upon the formulation of a neuronal theory that has been proposed over a hundred
years ago. So, what is new? In this article we attempt to trace the development of certain anatomical and
physiological concepts at the origins of modern denitions of the connectome. We argue that compared to
previous attempts current connectomic approaches benet from a wealth of imaging methods that in part
could justify the enthusiasm for nally succeeding in achieving the goal. One of the unique advantages of contemporary approaches is the possibility of using quantitative methods to dene measures of connectivity
where structure, function and behaviour are integrated and correlated. We also argue that many contemporary maps are inaccurate surrogates of the true anatomy and a comprehensive connectome of the human
brain remains a far distant point in the history to come.
2013 Elsevier Inc. All rights reserved.

Contents
Introduction . . . . . . . . . . . . . . . . .
Early proto-connectome maps (Table 1) . . . .
Modern cartography (Table 2) . . . . . . . . .
Contemporary and future connectomes between
Acknowledgments . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . .

. . . . . .
. . . . . .
. . . . . .
macroscopic
. . . . . .
. . . . . .

. . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .
metaphors and microscopic myths
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .

Introduction
The pattern of scientic investigation cannot be understood in
isolation: it must be set against the background of wider trends in
the sciences, methodological advancements, and general culture of
the time (Clarke and Jacyna, 1987). Social, biological, and technological
network models dominate contemporary approaches to complexity
(Egerstedt, 2011). Telecommunications, social networks, transportation
logistics, molecular interactions, and metabolic pathways are just some
examples in which network analysis is used to described complex
dynamics (Sporns, 2011; Strogatz, 2001).
Corresponding author at: Natbrainlab, PO50 Department of Forensic and Neurodevelopmental Sciences, Institute of Psychiatry, King's College London, London SE5 8AF, UK.
E-mail address: m.catani@iop.kcl.ac.uk (M. Catani).

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

.
.
.
.
.
.

0
0
0
0
0
0

The same approach has been proposed to describe the complexity of

the nervous system. Here, interactions between 86 billion of neurons
could be dened using nodes, hubs, connections and their properties
quantied in terms of efciency, centrality, global and local integration,
etc. (Bullmore and Sporns, 2009). There is also a general perception
that network analysis could bring us closer to a true understanding
of the real working of the human brain and its disorders. Two
recent multicentre research projects testify to the interest and
commitment of the international scientic community to this endeavour. The Human Connectome Project (http://www.humanconnectome.
org) is a $40 million NIH funded study to map the human connectome
in 1200 healthy subjects using large scale functional and structural imaging. The Human Brain Project (http://www.humanbrainproject.eu) is
one of the agship projects of the European Commission that is likely to
receive funding in the region of 1 billion for the simulation of the

1053-8119/$ see front matter 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.neuroimage.2013.05.109

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

entire human brain connectivity at a neuronal level and emulation of its

computational capabilities. With the recent announcement of President
Obama of a budget of possibly $100 million a year for the Brain Research
through Advancing Innovative Neurotechnologies (BRAIN) Initiative,
this eld of research will see an even greater expansion.
This unprecedented support is in large part due to the development
of new methods to image networks in the living human brain and the
computational capability of processing and storing large amounts of
data. The connectome approach, although new in its overarching conception, represents the culmination of converging lines of research,
each of which have developed over the course of many centuries
(Tables 1 and 2). In this paper we attempt to trace timelines of those anatomical (both at the macro- and microscopical scale), physiological
and methodological advances that helped to generate brain maps and
shape their evolution throughout history. Tables 1 and 2 describe
some of the pivotal discoveries in neurosciences from the eld of microscopy, electrophysiology/computational sciences, and anatomy/
neuroimaging. These timelines are not intended to be exhaustive
but rather highlight only a number of key discoveries and developments that have some relevance to contemporary connectomic
approaches. Our aim is to give credit to pioneers and put current
connectome projects into a wider context. The gures reproduced
in the article are examples of historical brain maps that can be
considered as possible forerunners of contemporary connectomes.
In the nal part we try to take advantage of our historical survey

to identify those steps needed to ll the gap between current

connectome maps and the real underlying anatomy of the human
brain.
Early proto-connectome maps (Table 1)
A connectome is dened by its nodes and connections, whose anatomy and function have been varyingly dened throughout history,
according to the predominant theoretical constructs and the latest methodological advancements available at the time. For a long time the most
popular representations of the brain and its functions consisted of
diagrams depicting a variable number of intercommunicating ventricular cells (Fig. 1). The ventricular theory was the direct result of the use
of brain dissections as originally proposed by Herophilus of Alexandria
in the fourth century BC and two centuries later by Galen (Clarke and
O'Malley, 1996). In these maps nodes correspond to ventricular reservoirs or cells with specialised functions. These chambers communicate
through a system of interventricular foramina and to peripheral organs
by means of hollow nerves. In these proto-connectome maps, essential
elements of contemporary connectome descriptions can be found. Implicit to the ventricular theory was the anatomical differentiation and
functional specialisation of ventricular cells communicating through a
network system. Further, brain functions were seen as emerging from
the dynamic and regulated exchange between hierarchically organised
ventricular cells.

Table 1
Milestones in the history of neuroscience from Renaissance to the end of 19th century.

MICROSCOPY

ELECTROPHYSIOLOGY/COMPUTATIONAL

1590 Hans and Zacharias Janssen invent the

microscope
Marcello Malpighi observes the cortex and
1666 describes globules and fibres
1674

Antonius van Leewenhoek provides the first

detailed description of nerve fibres

1649 Rene Descartes describes the reflex as a sensorymotor mechanism for involuntary muscle contraction

1586

Arcangelo Piccolomini, distinguishes the medulla (i.e.

white matter) and the cerebrum (i.e. grey matter)

1660 Jan Swammerdam discovers that the mechanical

stimulation of nerves produces muscle contractions

1664

Thomas Willis speculates on the sensory and motor

nature of ascending and descending projection
pathways

1684

Viessens separates commissural fibres of the

corpus callosum from projection fibres of the
centrum ovale

1786

Felix Vicq dAzyr, describes commissural and

associative pathways

1810

Joseph Gall identifies different cerebral gyri and

localises cerebral functions in the cerebral cortex

Isaac Newton suggests the electrical nature of

1713 nerve signal propagation
Leonhard Euler gives a mathematical formulation of the
Knegsberg Bridge from which originates the modern graph
theory

1736

1791 Luigi Galvani publishes his work on animal electricity and

describes nerves as pathways that conduct electricity
1803 Giovanni Aldini, applies electrical currents to mammalian
brains to trigger motor responses

Christian Ehrenberg, Gabriel Valentin and

1833-8 Jan Purkinje describe single populations of
neurons
1839 Theodor Schwann proposes the cell theory
for all living organisms
Jules Baillarger describes the cortical layers
1840
and compares them to a Galvanic pile

1850

1830s

Carlo Matteucci measures an electrical voltage across

the cell membrane

1843-45

Augustus Waller describes the Wallerian

degeneration of axons

1887

Santiago Ramn y Cajal proposes the neuron

theory

1891 Heinrich Waldeyer-Hartz uses the term neuron to

indicate the functional unit of the nervous system
1897 Charles Sherrington coins the term synapse

1819-26 Karl Burdach extends Reil's work and gives

latin names to association tracts
1825 Jean-Baptiste Bouillaud demonstrates that speech
is localised in the anterior regions of the brain

1848 Emile du Bois-Reymond discovers the action potential

(negative Schwankung)
1850 Herman von Helmholtz measures the propagation
speed of the nerve impulse

Rudolph Kolliker identifies fibres and cells

in the cortex
1860 Otto Deiters describes axonal and dendritic
processes
1862 Carl Weigert develops the first myelin stain

1873 Camillo Golgi discovers the black reaction

Johann Reil describes several association tracts,

1809-12 including the uncinate, arcuate, inferior longitudinal
fasciculus and cingulum

Wilhelm Griesinger and Thomas Laycock develop the

concept of psychic reflex for higher cognitive functions

Bartolomeo Panizza discovers the visual centre in the

1855 occipital lobe
1857 Franois Lauret and Louis Gratiolet propose the lobar
division of the brain

1852

1867 Theodor Meynert describes interlobar

variations in the cortical layering

NEUROANATOMY/NEUROIMAGING

1861 Paul Broca identifies an area for speech production

Julius Bernstein obtains direct recording of the action
potential and its kinetics
1870 Gustav Fritsch and Julius Hitzig use electricity to localize
motor regions
Richard Caton records electrical activity from exposed
1875
rabbit and mouse brains
1878 James Sylvester introduces for the first time the
mathematical term graph

1868

Theodor Meynert formulates the associationist

theory of brain function
1874 Carl Wernicke puts forward the first nerwork model
of higher cognitive functions

1870-85

1886 Vittorio Marchi develops a technique to trace

degenerating axons over long distances
1896 Paul Flechsig obtains myelogenetic maps of the human
brain and distinguishes primary from association areas

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

Table 2
Milestones in the history of neuroscience in the 20th century.

MICROSCOPY

ELECTROPHYSIOLOGY/COMPUTATIONAL
Julius Bernstein advances the hypothesis that the action
1902 potential results from a change in the permeability of the
axonal membrane to ions
1903

Alfred Campbell publishes the first map of the

1905
brain divided into 17 cortical fields

1909

Korbinian Brodmann produces cytoarchitectonic maps of the brain

1910

Oskar and Cecile Vogt work on myeloarchitectonic maps of the human frontal lobe

Constantin Economo and Georg Koskinas

1925 publish the most comprehensive
cytoarchitectonic maps

1938 Ernst Ruska develops the electron microscope

1942 Albert Coons proposes immunoflurescence
staining.

Walle Nauta and Paul Gygax develop a new

1951
staining method for degenerating axons

1977 Hans Kuypers uses fluorescent axonal tracers

Luis Lapicque publishes a model of integrate-and-fire

1907 neurons suggesting a threshold for firing

1918 Walter Dandy introduces ventriculography

1927 Egas Moniz develops cerebral angiography

1927 Hans Berger records the first human EEG

Jacob Moreno presents the first sociogram as a tool to study
1933 and visualize patterns of interpersonal relationships

1937 Wilder Penfield and Edwin Boldrev describe the motor and
sensory homunculus in man
1943 Warren McCulloch and Walter Pitts propose the first
mathematical model of a neural network
1946 Donald Hebb proposes a theory for synaptic plasticity and
learning process
1952 Alan Hodgkin and Andrew Huxley publish a mathematical
model for nerve excitation

1959 David Hubel and Torsten Wiesel describe oriented

receptive fields in the cat's primary visual contex
1969-71David Marr and James Albus develop a neurobiological
and computational theory of cerebellar function

Eduardo Macagno, uses a serial electron

1979 microscopy to map an isolated neuron in the
water flea
1986 Jay McClelland and David Rumelhart apply parallel distributed
processing theories to cognitive psychology and cognitive neuroscience

Gabriella Ugolini introduces the use of viruses

1987 as transneuronal tracers

Tim Berners-Lee develops a new hypertext system that

1989 runs across the internet, the world wide web

Rabies virus are used to study polysynaptic

neural networks

1935 Joseph Klinger describes a new procedure to

perform blunt dissections of white matter tracts

Felix Block and Edward Purcell independently

1946 describe the NMR phenomenon for liquid and solid
1950s

The first Positron Emission Tomography

scanners are developed

1960s David Cohen develops MEG

1970s Godfrey Hounsfield produces Computerised Tomography
1973 Paul Lauterbur publishes the first NMR image
1980 First Clinical MRI scanner

Electron microscopy is used to describe the first

1986 complete connectomme of the C. elegans (302
neurons)

1995

1901 Joseph Dejerine describes the topographic

distribution of fibres within the internal capsule

Ivan Pavlov discovers the conditioned reflex as an

automatic form of learning

1959 Mountcastle and Powell identify the columnar organization

of the cortex
Intra-axonal tract tracing compounds are
1960-70
developed (e.g. horseradish peroxidase
and radiolabeled amino acids )

NEUROANATOMY/NEUROIMAGING

Duncan Watts and Steven Strogatz present a

1998 mathematical model to describe small world networks

The ventricular model had a long lasting inuence despite the fact that
obvious experimental evidence of its fallacy emerged during the Renaissance. Leonardo da Vinci, for example, obtained a wax cast of the ventricles that was clearly against the classical representation of the
ventricular system (Pevsner, 2002). Similarly, Vesalius showed that the
ventricular anatomy described by Galen was questionable and its doctrine did not t with the evidence from dissections (Vesalius, 1543).
From the sixteenth century the ventricular theory co-existed with a
new spirit in science based on the experimental method (Galilei, 1638)
and the renewed belief of an intimate relationship between anatomy
and function (Catani, 2007; Catani and Thiebaut de Schotten, 2012).
Post-mortem dissection became the primary method of investigation of
the nervous system and this led to important anatomical discoveries,
among them the distinction between the cerebrum (i.e. grey matter
or cortex) and the medulla (i.e. white matter) by Arcangelo Piccolomini
in 1586:
I call the cerebrum [grey matter] that whole ashen-colored body,
darkening from white, which very closely encompasses the medulla.
The medulla is the whole of the white and more solid body, which is
concealed within the ashen-colored one. Thus the cerebrum differs and
is distinguished from the medulla by color, because the cerebrum is
ashen-colored but the medulla is white; in consistency, because the cerebrum is softer and the medulla a little harder and more compact; in location, because the medulla is in the middle of the cerebrum which wholly
covers it over; also the ashen-colored body is distinguished from the

1985 Denis Le Bihan applies Diffusion MRI to the

living human brain
1988 Jean Talairach and Pierre Tournoux publish the
first atlas in a common space of reference

1990 Seiji Ogawa describes the BOLD effect

1994 Peter Basser develops diffusion tensor imaging
1999 First in-vivo human diffusion tractography
reconstructions

white by certain lines. The cerebrum commences everywhere by convolutions and extends as far as the corpus callosum...
This distinction has direct relevance to modern connectomics
where hubs are located in the grey matter and connections in the
white matter. This is particularly true for those approaches that use
neuroimaging and electrophysiological methods to map whole brain
networks at the macroscopic level.
The study of the white matter anatomy expanded in the seventeenth
century when many scientists recognised that white matter contains bres whose trajectories could be followed and described if specimens
were carefully prepared using all the necessary precautions (Steno,
1669):
bres must be disposed in the most artful manner, since all the
diversity of our sensations and movements depend upon them. We admire the contrivance of the bres in each muscle, and ought still more
to admire their disposition in the brain, where conned in a very small
space, each execute their particular ofces without confusion or disorder.
From these anatomical studies a new view of the white matter
emerged; no more a homogenous support structure around the ventricles, but rather a complex medium composed of tubular laments
for the passage of uid between central cells and peripheral nerves
(Descartes, 1662). These connecting laments were found to originate from the cortex (Malpighi, 1666), specialise in motor and

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

anatomy (Fig. 2, left) (Descartes, 1664). Others produced images that

had the primary objective of displaying true anatomical ndings
(Fig. 2, right) (Vieussens, 1685). By now the concept of connectivity
was implicit in the idea of bres and nerves.
Originally prompted by the ventricular theory, the study of the
white matter pathways remained linked for a long time to a hydraulic
physiology of the brain. After all if, according to the ventricular theory,
cogitation emerges from the ow and passage of uid through hollow
bres, an exact description of these pathways could reveal the mechanisms of brain function. With the progressive accumulation of new anatomical ndings the ventricular theory evolved, with a shift from
central ventricular specialisation of function to a compartmentalised
system of individual gyri containing animal spirits (Willis, 1664):
But a no less important reason and necessity for the twistings [gyri] in
the brain arises from the distribution of the animal spirits. Since for
the various act of imagination and memory the animal spirits must
be moved back and forth repeatedly within certain distinct limits
and through the same tracts or pathways, therefore numerous folds
and convolutions of the brain are required for these various arrangements of the animal spirits; that is, the appearance of perceptible
things are stored in them, just as in various storerooms and warehouses, and at given times can be called forth from them. Hence these
folds or convolutions are far more numerous and larger in man than
in any other animal because of the variety and number of acts of the
higher faculties
Fig. 1. The ventricular system represented in a drawing dating from about 1310
(reproduced from Clarke and Dewhurst, 1972). The ve cells are named according to
their specialisation of function: the most anterior cells are the sensus communis (e.g.
common sense) and the ymaginatio (e.g. visual sense) connected to the eyes through
the optic nerves. Behind are the cell estimativa and the cell cogitativa, the latter
connected to a fth cell, the vis memorativa located below the vermis of the cerebellum. Cognitive processes result from the passage of spirits from one cell to the other.
Also note the hierarchical arrangement of the cells with the visual sense acting as a
major hub connected to three other cells and the eye, while the other cells are only
connected to either one or two cells.

sensory functions (Willis, 1664), and coordinate a wide range of behaviours, from simple reex responses to cognition (Descartes,
1664). Some of the maps of this period had little anatomical accuracy.
The gure produced by Descartes to illustrate the complexity of white
matter connections, for example, was pictorial in its representation of
a philosophical concept rather than a faithful portrait of the real

At the beginning of the nineteenth century brain maps of white matter connections became gradually more rened with differentiation of
tracts into callosal, projections and association pathways. Individual association tracts were also identied and their anatomy described in detail (Burdach, 1822; Reil, 1809, 1812). The convergence of anatomical
advancements with the progressive corticalisation of brain functions
culminated in Gall and Spurzheim's organology theory (Fig. 3). They believed that the brain is the organ of the mind and itself is made up of multiple organs, to be identied with the convolutions (Gall and Spurzheim,
1810):
The convolutions, as far as they constitute an organ, receive their bers from different regions These bres or bre bundles have a constant and uniform direction, different however in each region; they
form their own expansions and their own convolutions; they develop

Fig. 2. The beginning of the modern study of white matter connections based on methods for bre dissection. Left) Descartes' (1664) representation of the intricate system of white
matter passages in the human brain had little anatomical correspondence. Right) Vieussens (1684) used post-mortem dissections to identify white matter tracts and was the rst to
separate the centrum ovale composed of projection bres from the commissural bres of the corpus callosum, in this gure partially removed in the midline.

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

Fig. 3. Organology and phrenology according to Gall and Spurzheim (1810). Left) Lateral view of a human brain where different groups of gyri are indicated with progressive numbering according to their functional specialisation (Organology). Right) Protuberances on the skull that according to the phrenological theory resulted from the progressive expansion of the underlying gyri. Please note the correspondence of the numbers between the two gures.

at different stages of life; their number varies greatly in different

kinds of animal each organ is independent and acts by itself by
the virtue of its own powers and it contains directly within itself the
proximate cause of the phenomena which it offers.
But their view, although very modern in some respects, fell into disrepute mainly for two reasons. First, organology failed to embrace new
ideas on the physiology of the nervous system based on experimental
work of Galvani and Aldini (see below). Second, organology lost
scientic credibility when Gall and Spurzheim later suggested that
mental faculties are localised in well-developed cortical organs and
concluded that larger organs leave greater impressions on the skull.
With their new Phrenological theory they hoped to ascertaining the
several intellectual and moral dispositions of man and animal, by the
conguration of their heads (Gall and Spurzheim, 1810). Despite the
fast spreading of phrenological ideas and their large diffusion across
the continents, others continue to work on alternative models.
In particular the rst half of the nineteenth century saw the
emergence of a new paradigm: animal electricity (Galvani, 1791). This
sparked debates in Italy and Europe and propelled a new series of experiments that signposted the origin of modern electrophysiology. The new
paradigm attracted illustrious gures from other elds and promoted
cross-fertilisation between anatomy and physics. The layering of the
cortex, for example, was seen as an electrical generator of the human

brain (Baillarger, 1840) and current was applied to the brain to

elicit movement (Aldini, 1803). New recording methods were also
introduced to study the characteristics of the action potential in the
muscles (Bois-Reymond, 1848; Matteucci, 1830) and peripheral
nerve (Bernstein, 1868). By the mid-nineteenth century, the concept
of spinal reex was already established. This was extended from the
spine to the brain (Griesinger, 1843; Laycock, 1845) and, with the
experimental work of Ivan Pavlov (1903), conditioned reex became
a basic aspect of physiological psychology.
The second half of the nineteenth century witnessed also the explosion of advanced methods for microscopy. In 1839 Schwann proposed
the cell theory for all living organisms and many microscopists begun
to describe new cells in the brain (Schwann, 1839). Klliker divided histological features of the cortex into myelo- (i.e. the pattern of distribution of cortical bres) and cytoarchitectonic (i.e. the pattern of cellular
distribution in the cortex) (Klliker, 1859) and Meynert observed
interregional variations of the cortical layering (Meynert, 1868).
Methods for bre staining advanced even at a faster pace. The introduction, for example, of methods for myelin staining by Carl Weigert and
Vittorio Marchi, and the development of the precision microtome for
the study of serial sections improved the visualisation of small bres
in the normal and pathological brains (Bentivoglio and Mazzarello,
2010). Bernhard von Gudden perfected an experimental method in
animals whereby he could produce secondary degeneration and atrophy

Fig. 4. Cortical centers and connections in the late nineteenth century dened using clinico-anatomical correlation and post-mortem dissections. Left) Cerebral centres in the human
brain dedicated to motor, somatosensory and language functions as displayed in one of the most popular neuroanatomy textbooks of the time (Testut, 1897): I) writing centre of
Exner; II) Broca's centre for speech; III) motor centre, lower limb; IV) motor centre, upper limb; V) motor centre, face and tongue; VIVII) Dejerine's centre for reading; and VIII)
Wernicke's acoustic centre for verbal comprehension. Blue and purple areas are zones of the association centres according to Flechsig (1896). Right) Dejerine's representation of the
white matter tracts of the human brain responsible for language and reading (1895). Note that at that time the exact correspondence between centres and cortical projections was
not well established.

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

Fig. 5. Microscopy applied to the study of microcircuits of the hippocampus. Left) Hippocampal histology according to Camillo Golgi, inventor of the reazione nera (i.e. black reaction) method in 1873. Right) Hippocampal histology according to Santiago Ramon y Cajal (1911). Cajal also used the black reaction and introduced new concepts derived from
his anatomical observations, including the directionality of impulse propagation (here indicated by the arrows).

of the nerve nuclei and their connections by removing the peripheral

sense organs, such as the eyes, ears, or various cranial nerves (von
Gudden, 1870). Constantin von Monakow an assistant of Gudden, continued the work of his mentor on the thalamic and motor projection bres
(von Monakow, 1897). Following extirpation of circumscribed areas of
the cortex he was able to follow retrogradely the degenerating tract to
the thalamic nuclei and other subcortical nuclei. Von Monakow was
also the rst to draw attention to the effects of a lesion on other distant
regions of the brain, a mechanism that he called diaschisis (von
Monakow, 1914). Paul Emile Flechsig studied the developing pathways
with his myelogenetic method consisting of staining myelin in brains
of foetuses or newborns and mapping a chronological maturational

sequence of the white matter pathways in the human brain. Considering

that the projection tracts are among the rst to myelinate, he was able
to trace the origin and course of the corticospinal tract and describe for
the rst time asymmetry in its crossing (i.e. decussation) at the level of
the medulla (Flechsig, 1876). For many a true understanding of the nervous system was possible only through a precise depiction of its connections (Dejerine and Dejerine-Klumpke, 1895, 1901; von Bechterew, 1900).
At the same time clinicians begun to apply their neuroanatomical insights to patients (Catani et al., 2012a). The clinicalanatomical correlation method replaced phrenology and became a popular investigation
to infer possible functional correlates of cortical areas (Bouillaud, 1825;
Broca, 1861). Disorders of the nervous system were explained in terms

Fig. 6. Cortical myelogenetic maps and white matter projections according to Flechsig (1896). Left) Flechsig identied different areas according to their degree of myelination at
birth. Primordial areas are already myelinated at birth and have some correspondence with primary sensory and motor areas (densely dotted red areas). Tertiary association
areas myelinate after birth and correspond to large regions of the frontal, parietal, temporal and occipital lobes. Sparsely red-dotted areas show intermediate degrees of myelination
at birth. Right) Flechsig was able to reconstruct the trajectories of the main projection bres from his myelogenetic maps.

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

Fig. 7. Early cytoarchitectonic maps of the human brain. Left) Campbell's division of the cortex in 17 elds according to the interregional differences in cortical cyto- and
myeloarchitecture (1905). Right) Brodmann's maps according to purely cytoarchitectonic criteria (1909). Note that while Campbell believed that its areas had some functional
correspondence (hence terms like visuo-sensory and visuo-psychic), Brodmann rejected any correlation between individual areas and functional localisation.

of either cortical damage or disconnection syndromes (Fig. 4) (Dejerine

and Dejerine-Klumpke, 1895, 1901). Behind these maps was the idea
that brain should be understood as a whole, a system of integrated and
interconnected areas. This approach required a good degree of approximation, often resulting in a trivialised transposition of complex anatomical details into simplied diagrams (Catani and ffytche, 2005; Catani and
Mesulam, 2008).
Modern cartography (Table 2)
By the end of the nineteenth century two lines of investigation
emerged from the eld of microscopy. The rst important development
was the detailed description of microcircuits (Fig. 5) (Golgi, 1873; Cajal,
1893). Ramon y Cajal championed this approach and introduced important concepts in neurosciences, including a set of fundamental biological laws of neuronal organisation. He described, for example, how the
shape of axons and dendrites are constrained by biophysical parameters
such as cytoplasmic volume, space and conduction time. He also introduced the law of dynamic polarisation that identies dendrites and
cell body as the main receiving structures of the action potential and axonal terminations as the main output. The existence of these laws enabled him to formulate the neuron doctrine and to infer directionality
in signal ow between neurons (Cajal, 1911).
The second line of investigation attempted to produce whole-brain
maps based on the study of interregional variations of the cyto- and
myeloarchitecture of the cortex. The development of new staining
methods combined with painstaking observations of large numbers of
brain slices led to important landmark discoveries, including the identication of primary and associative areas of the brain (Fig. 6) (Flechsig,

1896), delineation of maturational trajectories of the long white matter

projection tracts (Flechsig, 1896) and improved localisation of functions
and associated symptoms (Campbell, 1905; von Economo and Koskinas,
1925). However, cartographers differed among them in the ultimate
objective of their work and disputed on whether it could be possible
to understand brain function from their cortical maps. Two schools of
thought emerged.
For many, cortical maps were primarily anatomical divisions without
any inference on function. Surprisingly, among them was Korbinian
Brodmann who was clearly disdainful of any functional localisation in
specic cortical areas (Brodmann, 1909):
In reality there is only one psychic centre: the brain as a whole with
all its organs activated for every complex psychic event, either all together or most at the same time, and so widespread over the different
parts of the cortical surface that one can never justify any separate
specially differentiated psychic centres within this whole.
Paradoxically, in contemporary cognitive neurosciences, Brodmann's
maps have become the lingua franca of cortical localisation numbers
used universally as shorthand for a precise cortical locus and, in some
cortical elds, a precise functional role (Fig. 7). Brodmann's approach
to clinico-anatomical correlation was shared by many other inuential
gures of his time and had long lasting consequences, especially in the
eld of psychology. Those who insisted on approaching the brain function from an anatomical point of view were disparagingly referred to as
the diagram makers and localisation theory were given very little importance (Head, 1926; Lashley, 1950).
The second approach was based on the assumption that cortical
divisions derived from cytoarchitectonic or myeloarchitectonic could

Fig. 8. Cortical divisions of the human brain and corresponding clinical syndromes according to Economo and Koskinas (1925). Left) Distribution of the ve principal types of neocortex in lateral surface of the human brain: 1. agranular, 2. frontal, 3. parietal, 4. granular, 5. polar. Right) Corresponding clinical syndromes (in German). Please note the attempt to
localise not only neurological syndromes but also psychiatric conditions (e.g. depressive vs expansive mental disorders in the frontal lobe).

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

reveal important functional divisions. This approach, pioneered by

Baillarger, Meynert, Wernicke and many others, had as ultimate goal
the establishment of solid scientic bases for an anatomically-based
classication of psychiatric disorders (Baillarger, 1840; Meynert, 1868;
Wernicke, 1906). The work of these psychiatrists had a great inuence on future generations. In Germany Paul Flechsig developed a
method for staining myelinated bres and was able to distinguish
areas according to the order of myelination during the perinatal
development (Fig. 6). In England, Alfred Campbell combined cytoand myeloarchitectonic observations with histological studies of
patients with various disorders to produce a map of distinctive cortical elds in the brain of Homo sapiens and other primate species
(Fig. 7) (Campbell, 1905). Campbell's monograph was a monumental
achievement for several reasons, the most important being the emphasis given to function. Campbell's project went beyond cytoarchitectonic
cartography, attempting to integrate clinical, anatomical, and physiological evidence to provide a guide to function (ffytche and Catani,
2005). Indeed, Campbell's 17 cortical elds are labelled not by numbers
but by function (e.g. visuo-sensory, audito-psychic, olfactory, precentral motor, and post-central somatosensory).
The era of cortical mapping culminated with the work of Cecile and
Oskar Vogt and Economo and Koskinas. By studying the variation of cortical myeloarchitectonic, the Vogt's tandem identied more than 200
areas, many of which represent subdivisions of the cytoarchitectonic
areas of Brodmann (Vogt and Vogt, 1926). Despite working fourhandedly on the project and relying on the help of other assistants,
they never nished their monumental project and their anatomical
endeavour has remained incomplete for most of the temporal and occipital
cortex (Nieuwenhuys, 2013). In 1925 Economo and Koskinas succeeded in
publishing a prodigious atlas containing the cytoarchitectonic analysis of
107 cortical areas, for each of which quantitative measurements were
recorded for variations in cortical thickness and volume, form, size,

number of cells, their density, grouping in stripes and layers (Fig. 8).
The atlas is a monumental work, which, despite being considered by
many the denitive text on cortical cartography, never met the favour
of the scientic community. This is probably in part due to its encyclopaedic proportions, the lack of clear boundaries between some of the
smallest areas and possibly the general feeling against the crazy paving
school of cortical research, which peaked shortly after (Le Gros Clark,
1952).
By the mid-twentieth century the combination of histological
methods with neurophysiological techniques applied to the animal
brain provided cortical architecture with a precise functional meaning.
This approach, as exemplied in the work of Mountcastle and Powell
(1959) on the somatosensory cortex of the monkey and of Hubel and
Wiesel (1962) on the visual cortex of the cat, was based on the use of
single- or multi-unit recording, axonal tracing by means of microlesion
or dye injection and combined with the cytoarchitectural description of
sections later cut from the same brain. Key principles of cellular organisation and neuronal physiology were discovered, such as columnar organisation of the cortex and oriented receptor elds. At the same time
the use of disconnection procedures in the monkey, combined with behavioural studies, intracortical recording, and axonal tracing revitalised
a network approach to brain functions (Fig. 9) (Mishkin, 1966). New
methods became available for the identication of single axons and
their exact cortical projection and termination (Fink and Heimer,
1967; Nauta and Gygax, 1951). By the end of 1960s novel powerful
tracers were developed based on the active transport of proteins and
other elements along the axonal bres. These methods require injection
of tract tracers into a predetermined cortical or subcortical region of the
nervous system. Once injected, the tracers enter the neuron and are
transported from the body of the neuron to its terminations (i.e. anterograde direction) or in the opposite direction (i.e. retrograde direction)
(Morecraft et al., 2009). Tracer compounds would differ for their ability

Fig. 9. Connectivity maps based on animal studies. One of the advantages of animal studies is the ability to use data obtained with different methods and directly test network-based
models of brain functions using experimental approaches (e.g. disconnection lesions). Lower left) Leslie Ungerleider and Mortimer Mishkin used animal electrophysiology, axonal tracing
and lesion studies to formulate a dual stream model of visual processing (lower left) (Mishkin et al., 1983). Right) Felleman and Van Essen's (1991) projectome of the visual system showing the hierarchical connectivity of the visual areas. One of the limitations of these approaches is the inability to quantify the strength of connectivity between different areas. This means
that the arrows and lines in the diagrams could be representative of single axons or large bundles. Also these maps have been transposed to humans without a direct anatomical verication. Bailey and von Bonin (1951), for example (upper left) applied their ndings from the monkey directly to the human brain without experimental verication. The use of
tractography could help to identify equivalence and between species differences in the anatomy of these tracts (Thiebaut de Schotten et al., 2011, 2012).

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

to follow a predominantly anterograde or retrograde direction although

most classical tracers are bidirectional (Glover et al., 1986; Kuypers et
al., 1977; Mesulam, 1982). The eld evolved further with methods
that combined different tracers (e.g. several conventional compounds
or conventional with viral transneuronal tracers) and multiple sites of
injection (Morecraft et al., 2009). These methods led to a greater
understanding of the main features of many cortico-cortical and
cortico-subcortical pathways, such as feedback and forward organisation, hierarchical arrangement, and parallel organisation (Felleman
and Van Essen, 1991; Mesulam, 1998). They also provided solid foundations for computational approaches to brain function. One limitation of
these approaches to connectivity is the general assumption that ndings from animals can be directly translated to humans. This view
may be valid for sensory and motor functions but there is some doubt
that it may hold true for other aspects of cognition such as language.
Post-mortem methods for cortical mapping of the human brain have
also evolved signicantly in the last two decades. Karl Zilles and collaborators, for example, have developed a number of methods for automatic cytoarchitectonic and receptor mapping of post-mortem human
brains (Zilles and Amunts, 2009). The advantage of these maps is that
they are provided within a standard template of reference and can be
used to guide, complement and integrate other in vivo imaging
methods (Caspers et al., this issue).
In contemporary neuroscience neuroimaging methods have inaugurated a new era in the study of functional and anatomical connectivity
in the living human brain. Although many methods are still in development and their use is limited, especially in the clinical settings, some
principles of brain function are emerging from their application to
large population of subjects. PET and fMRI studies for example, showed
the existence of a default mode network consisting a group of medial
and lateral regions that is active during the resting state, a condition
in which the majority of the subjects engage in an introspective,
self-directed stream of thought (i.e. similar to daydreaming) (Raichle
et al., 2001; Raichle and Snyder, 2007). A synchronous deactivation of
the default network areas is observed in the transition between the resting state and the execution of goal directed tasks, including working
memory, focusing attention to sensorially driven activities, understanding other people's intention (mentalising or theory of mind), prospective
thinking (envisioning the future) and memory for personal events
(autobiographic memory) (Raichle and Snyder, 2007). Alteration of the
default network activation has been reported in functional imaging
studies of patients with neuropsychiatric disorders, such as autism and
schizophrenia (Broyd et al., 2009).
Another signicant contribution from neuroimaging is the possibility of measuring cortical thickness in healthy subjects and patients
with neurological and psychiatric disorders. The use of this method on
patients with progressive loss of language due to a primary neurodegenerative disorder (i.e. Primary Progressive Aphasia), for example,
has signicantly contributed to expand the classical model of language
networks based on stroke studies to regions of the anterior temporal
lobe and medial frontal cortex (Rogalski et al., 2011). Similarly
tractography studies based on diffusion tensor imaging and spherical
deconvolution are revealing the existence of novel tracts underlying
frontal lobe functions (Catani et al., 2012b) and language (Catani
et al., 2005; Catani et al., in press). A fundamental contribution of contemporary neuroimaging is related to the description of interindividual
differences in brain connectivity and the possibility of quantifying
parameters that give indirect measurements of the functional and anatomical strength of connections between regions (Catani et al., 2007;
Thiebaut de Schotten et al., 2011). The above examples are indicative
of the far reaching potential of neuroimaging methods and their ability
to give answers to questions that were not possible to address before.
With the introduction of the concept of a brain connectome the eld
moved a step farther (Sporns et al., 2005; Hagmann et al., 2007). Here
the ambition is to dene the overall structural and functional brain architecture to understand how anatomical networks inuence neuronal

dynamics (Sporns, 2013). A popular approach uses network analysis

frameworks based on graph theory (Fornito et al., 2013; Hagmann et
al., 2010). This special issue is entirely dedicated to neuroimaging
methods for mapping the connectome and we refer to other papers
for an in depth illustration of the advantages and limitations of this
approach.
However, it is important to bear in mind that neuroimaging is only
one of many methods for mapping the connectome and an important
distinction should be made between those approaches that adopt
neuroimaging for mapping whole brain networks and other methods
that characterise the most detailed features of microconnections using
advanced microscopy technology (Lichtman and Denk, 2011; Seung,
2012; Sporns, 2011). Clearly for the two approaches the sufx -ome
refers to different concepts, one related to the totality of the brain (i.e.
the connectome as a map of the entire brain connections), the other
to the totality of all its constituents (the connectome as the most detailed description of the elements that form neuronal connections).
Hence, the rst approach provides a global overview (whole-brain) of
the principal brain networks at a macroscopic level (i.e. large-scale connections or bundles), the second aspires to the most detailed description
of local networks at a micro- and nano-scale (e.g. single axons and dendrites). While the rst approach can only be achieved through data reduction and oversimplication (i.e. connectome as a metaphor), the
second may never realise for the entire brain (i.e. connectome as a
myth). In the last paragraph we argue that the two approaches could
converge at the mesoscale level, at least for post-mortem studies in the
near future and perhaps in vivo in the long term.
Contemporary and future connectomes between macroscopic
metaphors and microscopic myths
Current approaches to brain mapping result from the coalescence
of fast paced advancements in computing (data processing and storage, software development, etc.), quantitative and statistical neuropsychological testing, MRI capability (higher resolutions, plethora of
sequences for structural and functional imaging) and computational
theories (Lichtman and Denk, 2011; Seung, 2012; Sporns, 2011;
Dell'Acqua and Catani 2012). In post-mortem brains the use of automated histological analysis combined with transmitter receptor
distribution and microarray proling is beginning to delineate a
new landscape of human cartography where multiple information is
available for each area (e.g. cytoarchitectonic, receptor and gene
expression) (Hawrylycz et al., 2012; Zilles and Amunts, 2009). This
information can be applied to in vivo imaging using atlas-based
approaches and correlated with differences in functional activity, diffusion connectivity and behaviour. This could lead for the rst time to
multimodal brain maps that dene interindividual variability among
the general population and help in understanding not only neural
mechanisms of normal cognition, but also identify vulnerable connectivity patterns in those at risk for mental illness, and predict treatment response and recovery after injury (Bullmore and Sporns,
2009; Jbabdi et al., 2007; Stephan et al., 2009).
Nevertheless, contemporary approaches based on brain imaging are
not without limitations. One risk of contemporary connectome maps is
to divert from the real functional anatomy of the brain and to follow instead a path of their own as it happened before in other disciplines. The
eld of articial neural networks, for example, just after McCulloch and
Pitts proposed the rst mathematical model for a neural network in
1943, quickly evolved into statistical and pattern recognition tools far
removed from the action of complex neurophysiological mechanisms
subserving their biological counterparts (Rosenblatt, 1958). In the
case of the connectome there are several factors that may contribute
to a similar outcome. Mapping methods, for example, are typically
based on data acquired at low resolution (a few millimeters) with several
distortions due to high background noise and eld dishomogeneity.
Signal derived from the MRI scans reects an average information,

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

10

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

which derives from the combination of complex biological features of the

underlying tissue, thus making any interpretation of the results
unspecic and often speculative. Further the number of passages related
to the data processing introduces artifacts and distortions that are
difcult to distinguish from the real structural and functional anatomy.
This is particularly true for connectomes based on diffusion imaging
where the limitations of both the tensor model and more advanced
methods (e.g. Diffusion Spectrum Imaging and High Angular Resolution
Diffusion Imaging) (Dell'Acqua et al., 2010; Dell'Acqua et al., in press-a;
Descoteaux et al., 2009; Tournier et al., 2007; Wedeen et al., 2012) may
produce connectivity maps that do not reect the real underlying anatomy (Catani et al., 2012c; Hagmann et al., 2010; Jones et al., 2013; Wedeen
et al., 2012).
With diffusion datasets acquired at higher spatial resolution (currently ~ 1mm) some of these limitations may reduce. In post-mortem
samples even higher resolutions can be obtained (Kleinnijenhuis et
al., 2012; Leuze et al., 2013; McNab et al., 2009) with the advantage
of direct validation with histology of tractography reconstructions
(Takahashi et al., 2013; Dell'Acqua et al., in press-b).
Another risk is that the language used today in contemporary mapping and analysis techniques can be ambiguous at times as it uses a terminology that alludes to anatomical properties when in fact it reects
only features of the derived maps or graphs. Thus, a distance between
two nodes in a network described using graph theory is not equivalent
to the axonal length between the neurons that constitute those nodes
but is the minimum number of steps required to connect the two
nodes, regardless of their physical distances (Bullmore and Sporns,
2009; Fornito et al., 2013). Similarly, the average length of streamlines
in tractography is not equivalent to the average length of the connecting
tracts. Moreover, many apparently functional properties of the neuroimaging networks (e.g. connectivity, synchrony, etc.) are mathematical
and statistical concepts rather than physiological. Therefore, concluding
that the brains of certain groups of patients have reduced connectivity
because of the reduced fMRI co-activation or shorter length of their
tractography streamlines calculated from DTI scans could be incorrect.
Today we are certainly in a better position to effectively integrate
different neuroimaging modalities and complement structural and
functional ndings. There is, however, an inevitable tension in contemporary mapmaking between the desire to cram in more and more information and the need to keep things clear. This happened before in the
history of neuroscience and history tends to repeat itself. Our current
connectomes too often reect a distorted image of the real architecture
of the brain and we tend to see in our connectomes patterns of neuronal
organisation that may just represent a biased view of the real anatomy.
Here again validation is paramount (Mesulam, 2012) perhaps not only
using post-mortem approaches but also in vivo studies in animal
models (Lee et al., 2012).
At the microscopic level the use of advanced imaging methods
based on uorescent protein staining, electron microscopy and
super-resolution light microscopy provides more condence for the
interpretation of the connectome results (Lichtman and Sanes,
2008; Denk et al., 2012). Here the problem is of a different nature
and mainly related to its gargantuan scale. The realisation of the complete connectome of the 302 neurons and 5000 synapses that form
the nervous system of the Caenorhabditis elegans (White et al., 1986)
was certainly an encouraging result that gave hope for similar achievements in larger brains. Completing the connectome of the Drosophila
could also represent a signicant leap forward in the phylogeny scale.
When we deal with the human brain the problem is related not only
to the 86 billion neurons and 100 trillion synapses that form our networks but to the ever changing anatomy of its constituents. Whenever
we engage in an action or thought we reshape our connectome by
forming new synapses and pruning neuronal dendritic trees. In other
words we wake up with one connectome and go to bed with a different
one. Even reading this paper could have an impact on the anatomy of
the reader's connectome. It is this dynamic nature of our connectomes

that holds the key to the real working of the brain and only a
connectome map of the brain in action will capture the anatomical,
electrophysiological and computational elements of those networks
that characterise human cognition and behaviour. Although this may
seem a distant point, at the time of the submission of this paper a report
published in Nature Methods reawakened our optimism. Researchers at
the Howard Hughes Medical Institute's Janelia Farm Research Campus
in Ashburn, Virginia have been able to record activity across a whole larval sh brain, detecting 80% of its 100,000 neurons (Ahrens and Keller,
2013). The imaging system relies on a genetically engineered zebrash
whose neurons make a protein that uoresces in response to uctuations in the concentration of calcium ions, which occur when nerve
cells re. A system composed of a microscope and detectors records activity from the full brain. The neurons are visible thanks to the transparency of almost the entire non-neuronal tissue of the zebrash.
Potentially this system could show the dynamics throughout the
nervous system while the zebrash engages in different behaviours
and during learning paradigms. Certainly the journey from here to application of similar methods in the human brain (perhaps with high resolution functional diffusion imaging) is a long one in the history to
come. In the meantime setting up combined imaging and postmortem histology studies of the white matter, perhaps using cuttingedge methods for network analysis at the axonal level (e.g. Polarised
Light Imaging or Clarity) (Axer et al., 2011; Chung et al., 2013) could
help us to validate our current structural methods and move more secure steps on the steep ascent of the connectome science.
Acknowledgments
We would like to thank Richard Joules, Stefano Sandrone and the
other members of the NatBrainLab (http://www.natbrainlab.com)
for their helpful advice on the manuscript. This work was supported
by Guy's and St Thomas Charity and the NIHR Biomedical Research
Centre for Mental Health at the South London and Maudsley NHS
Foundation Trust.
References
Ahrens, M.B., Keller, P.J., 2013. Whole-brain functional imaging at cellular resolution
using light-sheet microscopy. Nat. Methods 10, 413420.
Aldini, G., 1803. An Account of the Late Improvements in Galvanism : To which is
Added, an Appendix, Containing Experiments on the Body of a Malefactor.
Cuthell & Martin.
Axer, M., Amunts, K., Grssel, D., Palm, C., Dammers, J., Axer, H., Pietrzyk, U., Zilles, K.,
2011. A novel approach to the human connectome: ultra-high resolution mapping
of ber tracts in the brain. Neuroimage 54, 10911101.
Bailey, P., von Bonin, G., 1951. The Isocortex of Man. University of Illinois Press, Urbana.
Baillarger, J.G.F., 1840. Recherches sur la structure de la couche corticale des
circonvolutions du cerveau. J.-B. Baillire.
von Bechterew, W., 1900. Les voies de conduction du cerveau et de la moelle. Storck
and Doin.
Bentivoglio, M., Mazzarello, P., 2010. The anatomical foundations of clinical neurology.
In: Finger, S., Boller, F., Tyler, K.L. (Eds.), Handbook of Clinical Neurology, vol. 95,
pp. 149168 (3rd series).
Bernstein, J., 1868. Ueber den zeitlichen Verlauf der negativen Schwankung des
Nervenstroms. Pugers Arch. 1, 173207.
Bois-Reymond, E.H.D., 1848. Untersuchungen ber thierische elektricitt. G. Reimer.
Bouillaud, J., 1825. Trait clinique et physiologique de l'encphalite. J.B. Baillire, Paris.
Broca, P., 1861. Nouvelle observation d'aphmie produite par une lesion de la partie
postrieure des deuxime et troisime circonvolutions frontales. Bull. Soc. Anat.
Paris 36, 398407.
Brodmann, K., 1909. Vergleichende Localisationslehre der Grosshirnrinde in ihren
Prinzipien dargestellt auf Grund des Zellenbaues. Barth, Leipzig.
Broyd, S.J., Demanuele, C., Debener, S., Helps, S.K., James, C.J., Sonuga-Barke, E.J.S., 2009.
Default-mode brain dysfunction in mental disorders: a systematic review. Neurosci.
Biobehav. Rev. 33, 279296.
Bullmore, E., Sporns, O., 2009. Complex brain networks: graph theoretical analysis of
structural and functional systems. Nat. Rev. Neurosci. 10, 186198.
Burdach, K., 1822. Vom Baue und Leben des Gehirns. In der Dyk'schen Buchhandlung,
Leipzig.
Cajal, S.R. y, 1893. Manual de histologa normal y tcnica microgrca. Libreria de
Pascual Aguilar.
Cajal, S., 1911. Histologie du Systeme nerveux. A. Maloine, Paris.
Campbell, A.W., 1905. Histological Studies on the Localisation of Cerebral Function.
University Press.

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

Caspers, S., Eickhoff, S.B., Zilles, K., Amunts, K., 2013. Microstructural grey matter
parcellation and its relevance for connectome analyses. Neuroimage (S10538119(13)00319-4) (this issue).
Catani, M., 2007. From hodology to function. Brain 130, 602605.
Catani, M., ffytche, D.H., 2005. The rises and falls of disconnection syndromes. Brain
128, 22242239.
Catani, M., Mesulam, M.M., 2008. The arcuate fasciculus and the disconnection theme
in language and aphasia: history and current state. Cortex 44, 953961.
Catani, M., Thiebaut de Schotten, M., 2012. Atlas of Human Brain Connections. Oxford
University Press, Oxford.
Catani, M., Allin, P.G.A., Husain, M., Pugliese, L., Mesulam, M.M., Murray, R.M., Jones,
D.K., 2007. Symmetries in human brain pathways predict verbal recall. Proc. Natl.
Acad. Sci. U. S. A. 104, 1716317168.
Catani, M., Jones, D.K., ffytche, D.H., 2005. Perisylvian language networks of the human
brain. Ann. Neurol. 57, 816.
Catani, M., Dell'acqua, F., Bizzi, A., Forkel, S.J., Williams, S.C., Simmons, A., Murphy, D.G.,
Thiebaut de Schotten, M., 2012a. Beyond cortical localization in clinico-anatomical
correlation. Cortex 48, 12621287.
Catani, M., Dell'acqua, F., Vergani, F., Malik, F., Hodge, H., Roy, P., Valabregue, R.,
Thiebaut de Schotten, M., 2012b. Short frontal lobe connections of the human
brain. Cortex 48, 273291.
Catani, M., Bodi, I., Dell'Acqua, F., 2012c. Comment on The geometric structure of the
brain ber pathways. Science 337, 1605.
Catani, M., Mesulam, M., Jakobsen, E., et al., 2013. A novel frontal pathway underlies
verbal uency in primary progressive aphasia. Brain (in press).
Chung, K., Wallace, J., Kim, S.Y., Kalyanasundaram, S., Andalman, A.S., Davidson, T.J., et
al., 2013. Structural and molecular interrogation of intact biological systems. Nature 497 (7449), 332337.
Clarke, E., Dewhurst, K., 1972. An Illustrated History of Brain Function. Sandford
Publications.
Clarke, E., Jacyna, L.S., 1987. Nineteenth-century Origins of Neuroscientic Concepts.
Diane Pub Co.
Clarke, E., O'Malley, C.D., 1996. The Human Brain and Spinal Cord: A Historical Study
Illustrated by Writings from Antiquity to the 20th Century. Norman Publishing.
Djerine, J.J., Dejerine-Klumpke, A., 1895. Anatomie des centres nerveux, vol. 1. Rueff et
Cie, Paris.
Djerine, J.J., Dejerine-Klumpke, A., 1901. Anatomie des centres nerveux, vol. 2. Rueff et
Cie, Paris.
Dell'Acqua, F., Catani, M., 2012. Structural human brain networks: hot topics in diffusion tractography. Curr. Opin. Neurol. 25, 375383.
Dell'Acqua, F., Scifo, P., Rizzo, G., Catani, M., Simmons, A., Scotti, G., Fazio, F., 2010. A
modied damped Richardson-Lucy algorithm to reduce isotropic background effects in spherical deconvolution. Neuroimage 49, 14461458.
Dell'Acqua, F., Simmons, A., Williams, S.C., Catani, M., 2013a. Can spherical deconvolution
provide more information than ber orientations? Hindrance modulated orientational anisotropy, a true-tract specic index to characterize white matter diffusion.
Hum. Brain Mapp. http://dx.doi.org/10.1002/hbm.22080 (in press) (Apr 5, Electronic
publication ahead of print).
Dell'Acqua, F., Bodi, I., Slater, D., Catani, M., Modo, M., 2013b. MR diffusion-based histology and micro-tractography reveal mesoscale features of the human cerebellum. Cerebellum. http://dx.doi.org/10.1007/s12311-013-0503-x (in press).
Denk, W., Briggman, K.L., Helmstaedter, M., 2012. Structural neurobiology: missing link
to a mechanistic understanding of neural computation. Nat. Rev. Neurosci. 13,
351358.
Descartes, R., 1664. L'Homme de Ren Descartes et un Trait de la formation du ftus.
Charles Angot, Paris.
Descartes, R., 1662. De homine. apud Franciscum Moyardum & Petrum Leffen.
Descoteaux, M., Deriche, R., Knsche, T.R., Anwander, A., 2009. Deterministic and probabilistic tractography based on complex bre orientation distributions. IEEE Trans.
Med. Imaging 28, 269286.
Economo, C.F., Koskinas, G.N., 1925. Die Cytoarchitektonik der Hirnrinde des erwachsenen
Menschen. J. Springer.
Egerstedt, M., 2011. Complex networks: degrees of control. Nature 473, 158159.
Felleman, D.J., Van Essen, D.C., 1991. Distributed hierarchical processing in the primate
cerebral cortex. Cereb. Cortex 1, 147.
ffytche, D.H., Catani, M., 2005. Beyond localization: from hodology to function. Philos.
Trans. R. Soc. Lond. B. Biol. Sci. 360, 767779.
Fink, R.P., Heimer, L., 1967. Two methods for selective silver impregnation of degenerating
axons and their synaptic endings in the central nervous system. Brain Res. 4, 369374.
Flechsig, P.E., 1876. Die Leitungsbahnen im Gehirn und Rckenmark des Menschen auf
Grund entwicklungsgeschichtlicher Untersuchungen. Engelmann, Leipzig.
Flechsig, P.E., 1896. Gehirn und seele. Verlag von Veit & Comp, Leipzig.
Fornito, A., Zalesky, A., Breakspear, M., 2013. Graph analysis of the human connectome:
promise, progress, and pitfalls. Neuroimage (this issue).
Galilei, G., 1638. Discorsi e dimostrazioni matematiche: intorno a due nuove scienze,
attenenti alla mecanica e i movimenti locali Con una appendice del centro di
gravit d'alcuni solidi. Appresso gli Elsevirii.
Gall, F., Spurzheim, G., 1810. Anatomie et physiologie du systme nerveux en gnral et
sur celui du cerveau en particulier. Schoell, Paris.
Galvani, L., 1791. De viribus electricitatis in motu musculari: Commentarius.
Glover, J.C., Petursdottir, G., Jansen, J.K., 1986. Fluorescent dextran-amines used as axonal
tracers in the nervous system of the chicken embryo. J. Neurosci. Methods 18, 243254.
Golgi, C., 1873. Sulla struttura della sostanza grigia del cervello. Gazz. Med. Ital. (Lomb.)
33, 244246.
Griesinger, W., 1843. ber psychische Reexaktionen. Mit einem Blick auf das Wesen
der psychischen Krankheiten. Arch. Physiol. Heilk. 2, 76113.

11

Gudden, B.A., 1870. Experimentaluntersuchungen bei das peripherischer und centrale

Nervensystem. Arch. Psychiatr. Nervenkr. 2, 693723.
Hagmann, P., et al., 2007. Mapping human whole-brain structural networks with diffusion
MRI. PLoS One 2 (7), e597.
Hagmann, P., et al., 2010. MR connectomics: principles and challenges. J. Neurosci.
Methods 194, 3445.
Hawrylycz, M.J., Lein, E.S., Guillozet-Bongaarts, A.L., Shen, E.H., Ng, L., Miller, J.A., et al.,
2012. An anatomically comprehensive atlas of the adult human brain transcriptome.
Nature 489, 391399.
Head, H., 1926. Aphasia and Kindred Disorders of Speech. The University Press.
Hubel, D.H., Wiesel, T.N., 1962. Receptive elds, binocular interaction and functional
architecture in the cat's visual cortex. J. Physiol. (Lond.) 160, 106154.
Jbabdi, S., Woolrich, M.W., Andersson, J.L., Behrens, T.E., et al., 2007. A Bayesian framework for global tractography. Neuroimage 37 (1), 116129.
Jones, D.K., Knsche, T.R., Turner, R., 2013. White matter integrity, ber count,
and other fallacies: the do's and don'ts of diffusion MRI. Neuroimage 73,
239254.
Kleinnijenhuis, M., Zerbi, V., Ksters, B., Slump, C.H., Barth, M., van Cappellen van
Walsum, A.M., 2012. Layer-specic diffusion weighted imaging in human primary
visual cortex in vitro. Cortex (doi:pii: S0010-9452(12)00334-6) (this issue).
Klliker, A., 1859. Handbuch der Gewebelehre des Menschen. Wilhelm Engelmann.
Kuypers, H.G., Catsman-Berrevoets, C.E., Padt, R.E., 1977. Retrograde anoxal transport
of uorescent substances in the rat's forebrain. Neurosci. Lett. 6 (23), 127133.
Lashley, K.S., 1950. In search of the engram. Symp. Soc. Exp. Biol. 4, 454482.
Laycock, T., 1845. On the reex function of the brain. BFMR 19, 298311.
Le Gros Clark, W.E., 1952. A note on cortical cyto-architectonics. Brain 75, 96104.
Lee, H.J., Yoo, S.J., Lee, S., Song, H.J., Huh, M.I., Jin, S.U., Lee, K.Y., Lee, J., Cho, J.H., Chang,
Y., 2012. Functional activity mapping of rat auditory pathway after intratympanic
manganese administration. Neuroimage 60, 10461054.
Leuze, C.W., Anwander, A., Bazin, P.L., Dhital, B., Stber, C., Reimann, K., Geyer, S., Turner,
R., 2013. Layer-specic intracortical connectivity revealed with diffusion MRI. Cereb.
Cortex (in press).
Lichtman, J.W., Denk, W., 2011. The big and the small: challenges of imaging the brain's
circuits. Science 334, 618623.
Lichtman, J.W., Sanes, J.R., 2008. Ome sweet ome: what can the genome tell us about
the connectome? Curr. Opin. Neurobiol. 18, 346353.
Malpighi, M., 1666. De cerebri cortice. Montius, Bologna.
Matteucci, C., 1830. Sulla contrazione provata dagli animali all'aprirsi del circolo
elettrico in che trovansi. Casali, Forl, Italy.
McCulloch, W.S., Pitts, W.H., 1943. A logical calculus of the ideas immanent in nervous
activity. Bull. Math. Biophys. 5, 115133.
Mcnab, J.A., Jbabdi, S., Deoni, S.C.L., Douaud, G., Behrens, T.E.J., Miller, K.L., 2009. High
resolution diffusion-weighted imaging in xed human brain using diffusionweighted steady state free precession. Neuroimage 46, 775785.
Mesulam, M., 1982. Tracing Neural Connections with Horseradish Peroxidase. Wiley,
New York.
Mesulam, M., 1998. From sensation to cognition. Brain 121, 10131052.
Mesulam, M., 2012. The evolving landscape of human cortical connectivity: facts and
inferences. Neuroimage 62, 21822189.
Meynert, T., 1868. Neue Untersuchungen ber den Bau der Grosshirnrinde und seine
rtlichen. Verschiedenheiten, Vortrag.
Mishkin, M., 1966. Visual mechanisms beyond the striate cortex. In: Russel, R. (Ed.),
Frontiers in Physiological Psychology. Academic Press, New York.
Mishkin, M., Ungerleider, L.G., Macko, K.A., 1983. Object vision and spatial vision: two
cortical pathways. Trends Neurosci. 6, 414417.
Monakow, C., 1897. Gehirnpathologie. A. Hlder, Vienna.
Monakow, C., 1914. Die Lokalisation im Grosshirn und der Abbau der Funktion durch
kortikale Herde. JF Bergmann, Wiesbaden.
Morecraft, R.J., Ugolini, G., Lanciego, J.L., Wouterlood, F.G., Pandya, D.N., 2009. Classic and
contemporary neural tract tracing techniques. In: Johansen-Berg, H., Behrens, T.E.J.
(Eds.), Diffusion MRI. Elsevier, pp. 273308.
Mountcastle, V.B., Powell, T.P., 1959. Neural mechanisms subserving cutaneous sensibility, with special reference to the role of afferent inhibition in sensory perception
and discrimination. Bull. Johns Hopkins Hosp. 105, 201232.
Nauta, W.J., Gygax, P.A., 1951. Silver impregnation of degenerating axon terminals in the
central nervous system: (1) technic. (2) Chemical notes. Stain Technol. 26 (1), 511.
Nieuwenhuys, R., 2013. The myeloarchitectonic studies on the human cerebral cortex
of the Vogt-Vogt school, and their signicance for the interpretation of functional
neuroimaging data. Brain Struct. Funct. 218, 303352.
Pavlov, I.D., 1903. The experimental psychology and psychopathology of animals. The
14th International Medical Congress, Madrid, Spain.
Pevsner, J., 2002. Leonardo da Vinci's contributions to neuroscience. Trends Neurosci.
25, 217220.
Piccolomini, A., 1586. Anatomicae Praelectiones, explicantes miricam corporis humani
fabricam. Bonfadini.
Raichle, M.E., Snyder, A.Z., 2007. A default mode of brain function: a brief history of an
evolving idea. Neuroimage 37, 10831090 (discussion 10971099).
Raichle, M.E., MacLeod, A.M., Snyder, A.Z., Powers, W.J., Gusnard, D.A., Shulman, G.L., 2001.
A default mode of brain function. Proc. Natl. Acad. Sci. U. S. A. 98, 676682.
Reil, J.C., 1809. Die Sylvische Grube oder das Thal, das gestreifte groe Hirnganglium,
dessen Kapsel und die Seitentheile des groen Gehirns. Arch. Physiol. 9,
195208.
Reil, J.C., 1812. Die vrdere Commissur im groen Gehirn. Arch. Physiol. 11, 89100.
Rogalski, E., Cobia, D., Harrison, T.M., Wieneke, C., Thompson, C.K., Weintraub, S.,
Mesulam, M.-M., 2011. Anatomy of language impairments in Primary Progressive
Aphasia. J. Neurosci. 31, 33443350.

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109

12

M. Catani et al. / NeuroImage xxx (2013) xxxxxx

Rosenblatt, Frank, 1958. The perceptron: a probabilistic model for information storage
and organization in the brain, Cornell Aeronautical Laboratory. Psychol. Rev. 65,
386408.
Schwann, T., 1839. Mikroskopische Untersuchungen ber Die bereinstimmung
in Der Struktur und Dem Wachstum Der Tiere und Panzen. G. E. Reimer,
Berlin.
Seung, S., 2012. Connectome: How the Brain's Wiring Makes Us Who We Are.
Houghton Mifin Harcourt Trade.
Sporns, O., 2011. Networks of the Brain. MIT Press.
Sporns, O., 2013. The human connectome: origins and challenges. Neuroimage
S1053S8119.
Sporns, O., Tononi, G., Kotter, R., 2005. The human connectome: a structural description
of the human brain. PLoS Comput. Biol. 1, e42.
Steno, N., 1669. Discours de Monsieur Stenon, sur l'anatomie du cerveau Messieurs de
l'Assemble, qui se fait chez Monsieur Theuenot. chez Robert de Ninuille.
Stephan, K.E., et al., 2009. Tractography-based priors for dynamic causal models.
Neuroimage 47, 16281638.
Strogatz, S.H., 2001. Exploring complex networks. Nature 410, 268276.
Takahashi, E., Song, J.W., Folkerth, R.D., Grant, P.E., Schmahmann, J.D., 2013. Detection
of postmortem human cerebellar cortex and white matter pathways using high
angular resolution diffusion tractography: a feasibility study. Neuroimage 68,
105111.
Testut, L., 1897. Trait d'anatomie humaine: anatomie descriptive, histologie,
dveloppement. O. Doin.

Thiebaut de Schotten, M., Dell'Acqua, F., Forkel, S., Simmons, A., Murphy, D.G., Catani,
M., 2011. A lateralized brain network for visuospatial attention. Nat. Neurosci.
14, 12451246.
Thiebaut de Schotten, M., Dell'Acqua, F., Valabregue, R., Catani, M., 2012. Monkey to
human comparative anatomy of the frontal lobe association tracts. Cortex 48,
8296.
Tournier, T.J.D., Calamante, F., Connelly, A., 2007. Robust determination of the bre orientation distribution in diffusion MRI: non-negativity constrained superresolved
spherical deconvolution. Neuroimage 35, 14591472.
Vesalius, A., 1543. De Humani Corporis Fabrica. Oporini, Basel.
Vieussens, R., 1685. Neurographia universalis. Jean Certe, Lyon.
Vogt, C., Vogt, O., 1926. Die vergleichend-architektonische und vergleichendreizphysiologische Felderung der Grosshirnrinde unter besonderer Bercksichtigung
der menschlichen. Naturwissenschaften 14, 11901194.
Wedeen, V.J., Rosene, D.L., Wang, R., Dai, G., Mortazavi, F., Hagmann, P., Kaas, J.H.,
Tseng, W.-Y.I., 2012. The geometric structure of the brain ber pathways. Science
335, 16281634.
Wernicke, C., 1906. Grundrisse der Psychiatrie. Thieme, Leipzig, Germany.
White, J.G., Southgate, E., Thomson, J.N., Brenner, S., 1986. The structure of the nervous system of the nematode Caenorhabditis elegans. Philos. Trans. R. Soc. Lond. B 314, 1340.
Willis, T., 1664. Cerebri anatome: cui accessit nervorum descriptio et usus. typis Ja. Flescher,
impensis Jo. Martyn & Ja. Allestry apud insigne Campanae in Coemeterio D. Pauli.
Zilles, K., Amunts, K., 2009. Receptor mapping: architecture of the human cerebral
cortex. Curr. Opin. Neurol. 22, 331339.

Please cite this article as: Catani, M., et al., Connectomic approaches before the connectome, NeuroImage (2013), http://dx.doi.org/10.1016/
j.neuroimage.2013.05.109