UvA-DARE (Digital Academic Repository)

An introduction to human brain anatomy

Forstmann, B.U.; Keuken, M.C.; Alkemade, A.

DOI
10.1007/978-1-4939-2236-9_4
Publication date
2015
Document Version
Final published version
Published in
An introduction to model-based cognitive neuroscience
License
Article 25fa Dutch Copyright Act
Link to publication

Citation for published version (APA):

Forstmann, B. U., Keuken, M. C., & Alkemade, A. (2015). An introduction to human brain
anatomy. In B. U. Forstmann, & E-J. Wagenmakers (Eds.), An introduction to model-based
cognitive neuroscience (pp. 71-89). Springer. https://doi.org/10.1007/978-1-4939-2236-9_4

General rights
It is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)
and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an open
content license (like Creative Commons).

Disclaimer/Complaints regulations
If you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please
let the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the material
inaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letter
to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. You
will be contacted as soon as possible.

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

Download date:11 apr. 2023

Chapter 4
An Introduction to Human Brain Anatomy

Birte U. Forstmann, Max C. Keuken and Anneke Alkemade

If you want to understand function, study structure.

(Swaab [75])

Abstract This tutorial chapter provides an overview of the human brain anatomy.
Knowledge of brain anatomy is fundamental to our understanding of cognitive
processes in health and disease; moreover, anatomical constraints are vital for neu-
rocomputational models and can be important for psychological theorizing as well.
The main challenge in understanding brain anatomy is to integrate the different levels
of description ranging from molecules to macroscopic brain networks. This chapter
contains three main sections. The first section provides a brief introduction to the
neuroanatomical nomenclature. The second section provides an introduction to the
different levels of brain anatomy and describes commonly used atlases for the visu-
alization of functional imaging data. The third section provides a concrete example
of how human brain structure relates to performance.

4.1 Introduction

The human brain is the most complex and fascinating organ of the human body. Over
centuries, it has been studied by philosophers, physicians, anatomists, biologists,
engineers, psychologists, and in recent times also by neuroscientists. In the Middle
ages, anatomical training played a central role in medical education and knowledge
of human anatomy was highly valued [60]. However, during early years of the last
century and the rapidly developing field of empirical psychology, knowledge of
brain anatomy was considered insufficient to understand cognitive functioning. For

B. U. Forstmann () · M. C. Keuken · A. Alkemade

University of Amsterdam, Cognitive Science Center Amsterdam,
Nieuwe Achtergracht 129, 1018 Amsterdam, The Netherlands
e-mail: buforstmann@gmail.com
M. C. Keuken
e-mail: mckeuken@gmail.com
A. Alkemade
e-mail: jmalkemade@gmail.com
© Springer Science+Business Media, LLC 2015 71
B. U. Forstmann, E.-J. Wagenmakers (eds.), An Introduction
to Model-Based Cognitive Neuroscience, DOI 10.1007/978-1-4939-2236-9_4
72 B. U. Forstmann et al.

many, brain anatomy became largely irrelevant to the development of psychological

models of function and dysfunction [14, 15]. This position was stated succinctly
by the American philosopher and cognitive scientist Jerry Fodor: ‘. . . if the mind
happens in space at all, it happens somewhere north of the neck. What exactly turns
on knowing how far north?’ [16, 32].
In the last decade the advent of ultra-high field 7 Tesla (T) or higher magnetic
resonance imaging (MRI) holds the promise to reinstate anatomy to its former
important position. In vivo structural and functional brain measurements on the
sub-millimeter scale allow researchers to zoom in on fine-grained features such as
the layering of cortex [4, 28, 37, 39, 51, 85] and very small nuclei in the subcortex
[1, 34, 47] without having to lift the skull. Interest in this new technology is rapidly
growing and to date more than 50 ultra-high field MRI scanners are available for
human brain research worldwide. However, the importance of studying anatomy
with ‘expensive’ neuroimaging techniques continues to be criticized. Neuroscep-
tics rightly highlight the deficiencies of implausible and theoretically uninspiring
research findings (e.g., http://blogs.discovermagazine.com/neuroskeptic/) and one
may wonder whether much knowledge has recently been added to the work of the
great neuroanatomists of the past centuries.
Recent developments in neuroimaging have provided a powerful tool to the field
of the neurosciences, which appears to have regained some of brain anatomy’s pop-
ularity in neuroscientific research. Traditional postmortem neuroanatomical studies
are being replaced, at least partially, by advanced non-invasive techniques that now
allow one to study larger numbers, and more importantly, brains of healthy indi-
viduals. In its modern form, anatomical studies will continue to contribute to our
understanding of brain anatomy and function.
In this chapter, we first provide a brief introduction to the nomenclature of neu-
roanatomy because it is all too easy to get lost in localization [20]. Next, we introduce
descriptive, sectional, and connectional neuroantomy and list grey and white mat-
ter atlases that are widely applied to study systematically brain structure. Finally,
a concrete example about function-structure relationships zooming in on individual
differences will be presented.

4.2 Nomenclature

Nomenclature is a normalized system of exactly defined terms arranged according to

specific classification principles and is essential in the field of anatomy [76]. The ori-
gins of the anatomical terminology date back to classical Greek and Roman antiquity
over 2500 years ago, and explain the current use of Greek and Latin spelling. Due
to the early development of terminology and nomenclature, anatomy is considered
to be one of the first exact medical fields [42].
The official anatomical nomenclature that is currently used is in Latin and was
created by the Federative Committee on Anatomical Terminology (FCAT) and ap-
proved by the International Federation of Associations of Anatomists (IFAA). It was
4 An Introduction to Human Brain Anatomy 73

Fig. 4.1 (a) Conventions in anatomical terminology; The sagittal plane (red) divides the brain in
a left and right part, the axial or transverse plane (blue) divides the brain in inferior and superior
parts, and the coronal or frontal plane (green) is the vertical plane perpendicular to the sagittal
plane. (b) Connectional neuroanatomy. The corpus callosum and anterior commissure represent the
main commissural tracts connecting the left and right hemisphere (red). The associative pathways
(green) connect specific cortical areas within a hemisphere (ipsilateral connections). The ascending
and descending projection pathways connect cortical and subcortical structures (blue)

published a year later as the Terminologia Anatomica (TA); [30]. Since Latin is
taught increasingly less in high schools [17], the TA has been translated in several
different languages including English, Spanish, Russian, and Japanese.
Another commonly used way of describing the brain is by differentiating between
descriptive, sectional, and connectional aspects of neuroanatomy (see also [16]).
Descriptive neuroanatomy of the nervous system can be defined as the process of
identifying and labeling different parts of the brain and spinal cord [16]. The brain
can be described from its surface including different views. The surface of the brain
can be viewed from above (axial view), from the front (anterior or coronal view), the
back (posterior or coronal view), as well as from the side (lateral or sagittal view;
Fig. 4.1a). The same terminology is used to indicate different regions of the brain
surface (e.g., superior frontal gyrus, Fig. 4.2). The terms distal, proximal, bilateral,
unilateral, ipsilateral, and contralateral are used to indicate the location of an area
relative to another area. In the following we provide some examples to clarify these
terms. Place a finger in your elbow cavity and do the following: Move your finger
towards your wrist. This movement is in distal direction compared to your elbow
cavity. If you move your finger back to the elbow cavity then this movement is in
proximal direction.
The term bilateral indicates that a brain area is represented in both hemispheres.
The term unilateral refers to only one hemisphere. A well-known example for a
unilateral representation is Broca’s area, an area essential for language [23, 49]. In
the majority of people this area is located in the left hemisphere.
Finally, consider a plane that cuts your body symmetrically into a left and
right part. Any connection between two regions that remains on one side is called
ipsilateral. If the connection crosses from the left to the right side or vice versa, this
is called contralateral.
74 B. U. Forstmann et al.

Fig. 4.2 Sectional anatomy. Cortical subdivisions (a) lateral view (b) medial view (c) brain
ontogeny

Sectional neuroanatomy describes the relationship between cortical and subcor-

tical structures, most commonly visualized along orthogonal axial, coronal, and
sagittal planes (Fig. 4.1a). Axial planes divide the brain into an upper and lower part.
In radiological convention, the axial slices are viewed from the feet upwards to the
head. Consequently, on axial slices the left side of the brain is on the right side of the
page, whereas in the neurological convention this mapping is inverted. Axial planes
are also sometimes indicated as horizontal or transverse planes. To allow generaliza-
tion and comparison of results, brains in neuroimaging studies are often displayed
within a space of reference (e.g., [53, 54, 78] see also the section about grey and
white matter atlases). The axial coordinates move in the Z-direction (negative values
indicate slices inferior to the anterior commissure, while positive values indicate
slices superior to the anterior commissure; Fig. 4.1a).
Coronal planes cut through the brain in the anterior to posterior direction. The
coronal planes are conventionally oriented with the left side of the brain on the right
side of the page (radiological convention). Slices anterior to the anterior commissure
are indicated with positive values on theY-axis. Finally, the midsagittal plane divides
the brain into two hemispheres. Parasagittal slices through the left hemisphere are
indicated by negative values on the X-axis.
Connectional neuroanatomy delineates the origin, course, and termination of con-
necting pathways. Post-mortem dissections of white matter tracts, i.e., the paler tissue
of the brain and spinal cord, require special preparation and are particularly difficult
4 An Introduction to Human Brain Anatomy 75

to perform. Recent developments in diffusion MRI tractography have revitalized

the field. The tracts are classified according to their course and terminal projections
(Fig. 4.1b). So-called commissural pathways run along a horizontal axis and connect
the two hemispheres. The majority of projection pathways follow a vertical course
along a dorso-ventral (descending) or ventro-dorsal (ascending) axis and connect the
cerebral cortex to subcortical nuclei, cerebellum, and the spinal cord. Association
tracts run longitudinally along an anterior-posterior axis from front to back and vice
versa and connect cortical areas within the same hemisphere.

4.3 Different Levels of Brain Anatomy

For more than five centuries, the nervous system has been divided into grey matter and
white matter [82]. Grey matter contains cell bodies, neuronal extensions including
axons and dendrites as well as synapses between extensions, glia, and blood vessels
(cf. [77]). Grey matter compartments consist of relatively low numbers of cell bodies
and are composed of mostly unmyelinated axons, dendrites, and glia cell processes
which together form a synaptically dense region which is called the neuropil (cf. [63]).
White matter is indicated as such due to its white appearance as a result of mainly
myelinated axons, although it also contains unmyelinated axons. White matter is used
as a generic term for nervous system volumes where axons predominate connecting
grey matter regions. It can also include neurons when it is in close proximity to grey
matter [63, 72].
In this section we will describe in more detail sectional neuroanatomy includ-
ing grey matter regions such as the cerebral cortex and the basal ganglia. Next
connectional anatomy will be discussed including white matter compartments and
pathways.

4.3.1 Sectional Anatomy

Sectioning of the brain in three orthogonal planes allows us to investigate anatomical

structures located deep inside the brain. Deep brain structures include the ventricular
system, deep grey nuclei, and a number of white and grey matter structures. The
ventricular system is filled with cerebrospinal fluid (CSF) produced by the choroid
plexus which provides a highly protective bath with great buffering capacity in which
the brain is immersed. The ventricular system consists of a number of brain cavities
which are interconnected and allow for a CSF flow towards the submeningeal space.
The mediobasal part of the brain and the areas bordered by the lateral ventricles are
made up by a number of grey matter nuclei. These nuclei are separated by many
fiber tracts which form the cerebral white matter. The outer aspect of the cerebral
white matter forms the cerebral cortex. Below we will describe the aforementioned
structures in more detail.
76 B. U. Forstmann et al.

Fig. 4.3 Reproduction of the classical Brodmann map. (a) lateral view, (b) medial view. Individual
numbers indicate Brodmann area numbers

4.3.1.1 Cerebral Cortex

The outer surface of the cerebral hemispheres is called cerebral cortex which consists
of a densely folded strip of grey matter. The cerebral cortex displays a multi-layered
organizational pattern. The architecture of the cortical layers varies across the brain.
The majority of the cortex is six-layered (neocortex), and can be further parcellated
in distinct areas or fields. The cytoarchitecture reveals variations in cell shape, size,
and density; characteristics which are used as anatomical landmarks to divide the
cortex into distinct areas.
According to the work of individual groups and authors, the number of areas that
are distinguished varies from a minimum of 17 [13] to a maximum of 107 [26]. The
most widely used maps are provided by Brodmann [9] (Fig. 4.3) and his original
atlas contained 52 areas.
According to Brodmann’s classification, the frontal lobe consists of eleven fields,
which are grouped in five main regions. (1) Area 4 corresponds to the primary motor
cortex, containing neuronal bodies, as well as cortico-spinal projection fibers which
show a somatotopical organization; (2) Area six contains the premotor region and is
subdivided into the lateral premotor cortex (PMC), and the pre-supplementary motor
area (pre-SMA); (3) Area 44 and 45 correspond to Broca’s area; (4) Area 8–10, and
46 include dorsolateral prefrontal areas, and 47 to the ventrolateral prefrontal corex;
(5). Finally, areas 11 and 47 represent the main parts of the orbitofrontal cortex.
The parietal lobe is divided into four regions consisting of a total of nine separate
fields by Brodmann. (1) Areas 1–3 correspond to the somatosensory cortex and
its cytoarchitecture strongly resembles that of the primary motor cortex; (2) The
more laterally located areas 5 and 7 together form the superior polymodal parietal
cortex; (3) Areas 39 and 40 are located in the inferior polymodal parietal cortex,
corresponding to the Geschwind’s territory; (4) The medial parts of areas 31, 5, and 7
form the precuneus, and area 43 is considered a transition region of the fronto-parietal
operculum.
4 An Introduction to Human Brain Anatomy 77

Fig. 4.4 Probabilistic Basal Ganglia atlas. (a) Probability maps for the striatum, globus pallidus
external segment, globus pallidus internal segment, subthalamic nucleus, substantia nigra, and red
nucleus of the left hemisphere. (b) Bilateral representation. (c) Axial presentation, level of the
globus pallidus. The color intensity reflects the percentage overlap across the 30 participants. (d)
Axial presentation, level of the red nucleus. The color intensity reflects the percentage overlap
across the 30 participants. (adapted from [46])

The temporal lobe is also divided into four main regions consisting of seven sep-
arate fields. (1) Area 41 is the primary auditory cortex; (2) Adjacent is the auditory
association cortex consisting of area 42 and 22, which in part overlays with Wer-
nicke’s area; (3) The temporal visual association cortex is formed by areas 20, 21, and
37; (4) Finally, area 38, one of the paralimbic areas, occupies the temporopolar cortex.
The occipital lobe consists of three areas (17–19). The primary visual cortex
corresponds to area 17, and area 18 and 19 form the visual association cortex.
The limbic lobe includes as much as 15 fields. Areas 11, 24, 25, 32, 33, 36,
and 47 form an olfactocentric group and areas 23, 26–31 form a hippocampocentric
group. Additionally, many white matter tracts are present in the cerebral cortex with
two main orientations: tangential and radial. Further myeloarchitectonic subdivision
within the cerebral cortex reveals over 200 additional areas [84].
Basal Ganglia The basal ganglia are located deep in the white matter of the cere-
bral hemispheres anterior to the thalamus positioned medial to the lateral ventricles
(Fig. 4.4). They consist of two major functional divisions, the striatum and, more
medially located, the globus pallidus. The striatum in turn is composed of two highly
interconnected masses, the putamen and caudate nucleus.
78 B. U. Forstmann et al.

The caudate nucleus shows a somewhat elongated shape with a large anterior head,
which becomes more narrow towards the thalamus and continues to narrow down
and forward towards the temporal lobe along the wall of the lateral ventricle. Towards
the antero-ventral part of putamen, distinguishing between the caudate nucleus and
putamen around the inferior border of the internal capsule is difficult.
The globus pallidus is further divided into an external and internal segment.
The putamen and globus pallidus together form the lentiform nucleus and are
located lateral to the internal capsule. Finally, two areas that are functionally but not
ontologically considered to be part of the basal ganglia are the subthalamic nucleus
(STN) and the substantia nigra. The STN forms a small lens-shaped nucleus, which
is high in iron content and a main target region for deep-brain stimulation for the
treatment of refractory movement disorders, e.g., Parkinson’s disease, and is located
adjacent in anterior and lateral direction of the substantia nigra. The substantia nigra
is a crescent shaped region that is one of the production sites of the neurotransmitter
dopamine and the degeneration of this nucleus is a main hallmark of Parkinson’s
disease. The substantia nigra is named after its pigmented, melanin containing
neurons, which allows it to be clearly distinguished in histological sections.

4.3.1.2 Cerebellum

The cerebellum, or “little brain”, is located inferior of the occipital lobe and posterior
to the brain stem. The functions of the cerebellum range from maintaining balance,
coordination of half automatic movements to cognitive functions such as attention,
language, and working memory [6, 38, 73]. Similar to the cerebrum, the cerebellum
is distributed across the two hemispheres, which are connected through the vermis.
Each hemisphere has three lobes: the anterior, posterior, and flocculonodular lobe.
The cerebellum is connected to the brain stem via three peduncles namely the superior
cerebellar peduncle, the middle cerebellar peduncle, and the inferior cerebellar pe-
duncle connecting respectively to the midbrain, pons, and medulla oblongata. While
the cerebellum occupies only 10 % of the total volume of the entire human brain,
it contains many more neurons than the rest of the brain combined, such that for
each neuron in the cerebral cortex there are on average 3.6 neurons in the cerebellum
[40, 43].

4.3.2 Connectional Neuroanatomy

White matter tracts can be classified in different groups: tracts in the brainstem and
projections, association, and commissural tracts in the cerebral hemispheres. Pro-
jection fibers connect cortical and subcortical gray matter, association tracts connect
two cortical areas, and commissural tracts connect the brain’s hemispheres.
Brain regions are connected by white matter tracts which vary across subjects in
terms of their position, extent, and course [12]. Even though the regional pattern of
connectivity is largely preserved across individuals and these afferent, i.e., conduct-
ing incoming, and efferent, i.e., conducting outgoing, pathways strongly influence
4 An Introduction to Human Brain Anatomy 79

the information processing properties of individual brain regions [61]. Consequently,

it is often desirable to understand regional activation patterns in terms of an underly-
ing system of neural regions and their interactions [35, 36]. To this end, advances in
diffusion weighted tractography, a technique that allows the measurement of white
matter fibers, offer the potential for explicitly relating functional and connectional
anatomy. For instance, regional connectivity patterns can be used to reliably identify
both cortical [3, 41] and subcortical regions [5, 22, 50] even when these areas lack
clear macroanatomical borders. Consequently, by collecting connectivity informa-
tion along with functional data, it is possible to directly relate structure to function
in individuals. The final section of this chapter provides a concrete example relating
white matter connectivity to function by zooming in on interindividual differences.
Finally, the brainstem consists of five major white matter tracts: the superior,
middle, and inferior cerebellar peduncles, the corticospinal tract, and the medial
lemniscus.
Projection fibers connect the cortex with distant regions located in the lower parts
of the brain and spinal cord. These fibers can be differentiated into two classes: the
corticothalamic/thalamocortical fibers (collectively called thalamic radiations) and
the long corticofugal (corticoefferent) fibers. The corticofugal fibers include such
fibers as the corticopontine, corticoreticular, corticobulbar, and corticospinal tracts.
These fibers all penetrate the internal capsule either between the thalamus and the
putamen or between the caudate nucleus and the putamen. When approaching the
cortex, they fan out forming the corona radiata. The thalamus is known to have
reciprocal connections to many areas in cortex.
Association fibers connect different ipsilateral cortical areas and are classified
into short and long association fibers. The former connect areas within the same
cerebral hemisphere and include the fibers connecting adjacent gyri, so-called U-
fibers. The long association fibers connect different lobes, forming prominent fiber
bundles. Some of the main association fibers are (1) the superior longitudinal or
arcuate fasciculus connecting the frontal, temporal, and parietal lobe; (2) the inferior
longitudinal fasciculus connecting the temporal and occipital lobe; (3) the superior
fronto-occipital fasciculus, connecting the frontal and parietal lobe; (4) the inferior
fronto-occipital fasciculus connecting the orbital cortex to the ventral occipital lobe,
and; (5) the uncinate fasciculus connecting the anterior temporal lobe and lateral or-
bital frontal regions. Three major fibers connecting the limbic system (hippocampus),
namely (1) the cingulum connecting the prefrontal, parietal and occipital cortex to
the temporal lobe and hippocampus; (2) the stria terminalis connecting the hypotha-
lamus to the amygdala, and; (3) the fornix, which is a projection fiber, connecting
the medial temporal lobe to the mamillary bodies and hypothalamus.
Commissural fibers interconnect the two cerebral hemispheres and contain more
than 300 million axons [58]. Most of the commissures interconnect homologous cor-
tical areas in roughly mirror-image sites, but a substantial number have heterotopic
connections ending in asymmetrical areas. The largest of the commissural fiber tracts
is the corpus callosum, which connects the neocortical hemispheres. Inferior to the
corpus callosum are two other commissures, namely the anterior and the posterior
commissure (AC and PC respectively). The anterior commissure connects the bilat-
eral anterior and ventral temporal lobes. An imaginary line between connecting these
80 B. U. Forstmann et al.

two points, the AC-PC line, is often used for MRI-analyses. See Catani and de Schot-
ten [15] for an in vivo atlas of the projection, association, and commissural fibers.

4.4 Neuroanatomical Atlases

Neuroanatomical atlases are a useful tool because they provide a common reference
framework for structural and functional MRI studies. This is important in light of the
substantial variability between individual healthy brains [44, 79], even in identical
twins [80]. For example, all humans have a transverse gyrus (known as Heschl’s
gyrus) in the superior temporal lobe that is associated with the primary auditory
cortex, but the number and size of these transverse gyri varies across individuals [62,
64]. Similar variability is observed in other cortical regions [2, 19, 59, 65, 66]. The
shape and relative locations of subcortical structures also differ between individuals
and change with age [24, 47, 48]. The same can be observed for commissural fibers.
This variability in brain structure across individuals is of critical importance because
it implicates that no two twin brains are identical at a macroscopic level and therefore
no single brain can be considered representative of a population. Consequently, any
atlas based on a single ‘template’ brain will necessarily provide limited accuracy.
In theory, it may be possible to determine a topological transformation that could
morph one brain into precise correspondence with another, although current spatial
normalization procedures tend to correct only for gross anatomical variability.

4.4.1 Grey Matter and White Matter Atlases

Prior to the development of MRI scanners, atlases were solely based on the analysis
of post-mortem tissue. Classical examples include for instance the work by Brod-
mann [9]. As described earlier, Brodmann’s atlas is based on cyotarchitectonics, i.e.,
the density and types of cells present in different cortical layers. Other cytoarchitec-
tonic atlases have been published by von Economo and Koskinas [26], Ngowyang
[57], Sarkisov and colleagues [69], and Braak [8]. Strikingly, all these atlases vary
substantially in the number of areas, the size of the areas, and the exact location of
delineated areas. The most recent cytoarchitectonic atlas is provided by Duvernoy
which covers both cortical and subcortical brain areas including the brain stem [25].
A different class of atlases is based on the myeloarchitecture, i.e., the distribution
of myelinated axons in the brain. The most prominent myeloarchitecture atlases were
published by Reil [67, 68], Burdach [11], Smith [71], Flechsig [31], Vogt [83], and
Strasburger [74].
Other classes of atlases that are used in the cognitive neurosciences are based on
non-human primate data. An example of such an atlas is provided by Schmahmann
and Pandya [70] which includes tracer sections. Postmortem sections were analyzed
after the in vivo injection with chemical tracers.
4 An Introduction to Human Brain Anatomy 81

4.4.2 Standard Space Atlases

With the advent of MRI, the need for a common reference space increased because
a common reference allows the comparison across individuals’ functional and struc-
tural brain maps. A first attempt at a standard 3D space was developed by Talairach
and Tournoux [78] which was based on a single post-mortem brain from a 60-year
old woman. This brain template, together with the anatomical labels, provided the
opportunity to compare individuals’ functional brain maps in a common reference
space. However, while this approach was considered state of the art for many years,
further progression of the research field now allows for better alternatives. In the be-
ginning of the 1990s, a template was developed which has now become the standard
space in MRI research: the Montreal NeuroImaging (MNI) template. Several versions
of this template exist with the most recent one including 452 subjects’ anatomical
MRI scans [29] specifying 142 brain areas. Other MRI atlases are based on macro-
scopic delineations of cortical and subcortical areas, such as the Harvard-Oxford atlas
which includes 96 brain areas and is created based on 37 subjects [21, 52]. Eickhoff
and colleagues [27] created an atlas which is based on cytoarchitectonic properties
and includes ten ex vivo brains. More recently, specialized atlases partly relying
on ultra-high resolution 7 T MRI have been developed including the STN [34, 47]
(http://www.nitrc.org/projects/atag), the locus coeruleus [46], and the thalamus [56].

4.5 Structure-Function Relationships

In the final section of this chapter, we will discuss a concrete example how structure-
function relationships can be established in a model-based framework. This example
focuses on interindividual differences in the speed-accuracy tradeoff (SAT), both
in behavior and in structural features of the human brain. The SAT describes a
behavioral regularity that explains why the temporal benefits of responding more
quickly come at the cost of making more errors. A possible mechanism for the
SAT is that the response threshold can be adjusted depending on the strategy. If the
response needs to be very accurate, a high response threshold is adopted so that
sufficient information can be accumulated. If a speeded response is required, a low
response threshold ensures responses that are quick but error-prone.
Non-human primate and human studies have provided handles on how the mech-
anism of flexibly adjusting response thresholds could be implemented in the human
brain (for a review see [7]). Specifically, the anatomical substrate that could subserve
the SAT mechanism is the interplay between cortex and the basal ganglia. There are
two main pathways that connect the cortex to the basal ganglia namely the cortical-
striatal pathway and the cortical-subthalamic pathway. The striatal hypothesis poses
that an emphasis on speed promotes excitatory input from cortex to striatum; the
increased baseline activation of the striatum acts to decrease the inhibitory control
that the output nuclei of the basal ganglia exert over the brain, thereby facilitating
faster but possibly premature responses. Alternatively, the STN hypothesis posits
82 B. U. Forstmann et al.

Fig. 4.5 Visualization of the STN with different MRI contrasts at different field strengths. (a) The
left panel shows a 7 T T1 weighted anatomical MRI scan. Note the low contrast in the midbrain.
The right panel shows a 7 T T2* weighted anatomical scan. Note the improved contrast in the
subcortical regions. (b) The lower left panel shows a 7 T T2* weighted anatomical scan. The use of
high field strength yields excellent contrast allowing differentiation between individual subcortical
nuclei including the STN and SN. The middle panel shows a 3 T T2* weighted anatomical scan and
the lower right panel shows a 1.5 T T2* weighted anatomical scan for comparison. (adapted from
[18])

that an emphasis on accuracy promotes excitatory input from cortex (e.g., anterior
cingulate cortex) to the STN; increased STN activity may lead to slower and more
accurate choices.
These two hypotheses result in competing anatomical predictions. In particular,
the striatal hypothesis predicts that participants who display better control of the SAT
have stronger white matter tract connections between cortex and striatum, whereas
the STN hypothesis predicts stronger white matter connections between cortex and
STN. Testing of these distinct hypotheses has been technically challenging, but a first
step has been made by visualization of the STN using 7 T MRI [33, 34, 47]. Since
the STN is a very small lens-shaped nucleus in the subcortex, it is barely visible on 3
T anatomical MRI scans (Fig. 4.5). From an anatomical point of view, the technical
development of ultra-high resolution MRI has therefore already been proven crucial.
4 An Introduction to Human Brain Anatomy 83

Fig. 4.6 Structural differences in brain connectivity predict individual differences in decision mak-
ing. (a) Connectivity-based seed classification for the pre-SMA projecting into the striatum (green)
and STN (red). (b) Individual differences in tract strength between right pre-SMA and right striatum
are associated with flexible adjustments of the speed-accuracy tradeoff. (adapted from [33])

The STN contains iron, yielding hypointensities which in turn results in a loss
of the MR signal. Moreover, there are large interindividual differences in the STN’s
location [47]. (Fig. 4.5)
Next, diffusion weighted imaging (DWI) was applied to calculate the tract strength
between cortex and striatum and cortex and STN, respectively. Additionally, behav-
ioral data from a moving dots kinematogram with a SAT manipulation was analyzed.
The behavioral data were then modeled using the linear ballistic accumulator (LBA)
model to account for latent cognitive processes such as individuals’ response caution
([10] see also the chapter 1B Cognitive models in action). Statistical model selection
techniques confirmed that the LBA explained the data best when only the response
caution parameter was free to vary between the speed and the accuracy condition.
Interindividual differences in efficacy of changing response caution were quantified
by the differences in LBA threshold estimates between both conditions. These dif-
ferences were then related to participants’ tract strength measures between pre-SMA
and striatum and pre-SMA and STN, respectively.
Results showed that individual tract strength between pre-SMA and striatum trans-
late to individual differences in the efficacy with which people adjust their response
thresholds (Fig. 4.6). This supports the striatal hypothesis of how the brain regulates
the competing demands for speed vs. accuracy and show that individual differences
in brain connectivity affect decision making even in simple perceptual tasks. The
findings also show that, inconsistent with the STN hypothesis of SAT, there is little
or no evidence that strong connections from any cortical region to STN lead to more
flexibility in threshold settings (see [33] for more details regarding these findings).
84 B. U. Forstmann et al.

Importantly, this research could only be performed using ultra-high resolution 7 T

MRI, which allowed accurate localization of the STN.
In sum, this example highlights structure-function relationships as revealed
through individual differences in brain connectivity. This approach provides a
window to human cognition that exploits individual anatomical variability and
complements popular methods such as functional MRI.

4.6 Concluding Comments

In contemporary neuroscience, the anatomy of the central nervous system has essen-
tially become the anatomy of the connections between brain regions. Understanding
patterns of connectivity and how neurons communicate is probably one of the next
frontiers in the neurosciences [55]. It is important to delineate fine-grained cere-
bral and subcortical structures using ultra-high resolution MRI [81]. Ultimately,
by linking information from large-scale networks with macro- and microanatomi-
cal knowledge we can reach a deeper understanding of descriptive, sectional, and
connectional neuroanatomy.
Finally, capturing the interindividual variability in the form of probabilistic atlases
and maps both from in vivo and postmortem brains holds a great promise. It will fa-
cilitate our understanding of functional neuroanatomy including changes associated
with development, aging, learning, and disease. Informing mathematical and neu-
rocomputational models of the human brain with the anatomical knowledge gained
by using ultra-high resolution MRI promises exciting new avenues of understanding
human cognition.

Exercises

Q(1)
Step (1) Imagine a human brain floating in front of you with the eyes pointing
towards the cutter.
Step (2) Take out your imaginary knife and make the following cuts:
a. Cut the midline.
b. Take the left part and cut it in half again so that you have a superior and inferior
part.
c. Take the inferior part and cut it again in half so that you have an anterior and
posterior part.
Q1(a) Which two parts did you have after step 2a?
Q1(b) In which part is the cerebellum located after step 2b?
Q1(c) In which of the two parts is the temporal pole located after step 2c?
4 An Introduction to Human Brain Anatomy 85

Q(2)
Q2(a) Which panel shows the brain in an axial view?
Q2(b) Which panel shows the brain in a sagittal view?
Q2(c) Name all the different anatomical structures in the following figure:

14

11
19 6
16
18 8
20
14 9 1 10
3
7
5 4 12
2
23 22
16 21
10
17

7 6
9 5

2 4 8
3
1
12 b
33

16

11

16
18 19
19
15 1
26
a 13
13

29
25 2

30 27 32

28 31
24

c
34

Q(3) What is the difference between myeloarchitecture and cytoarchitecture?

Further Reading

1. For a more in depth textbook on anatomy see Catani and de Schotten [16]. Atlas
of human brain connections (1st ed., pp. 1–515). Oxford University Press.
2. See http://brancusi.usc.edu for an extensive resource for information about neural
circuitry.
3. See www.brain-map.org for a large online collection of neuroanatomical atlases,
gene expression and cross-species comparisons.
4. See http://www.unifr.ch/ifaa/Public/EntryPage/ViewTAOnLine.html for the ana-
tomical nomenclature.
5. See https://bigbrain.loris.ca/main.php for an ultra-high resolution image of a
post-mortem stained brain.
6. See http://www.appliedneuroscience.com/Brodmann.pdf for the English trans-
lation of the seminal work by Korbinian Brodmann.
86 B. U. Forstmann et al.

References

1. Alkemade A, Keuken MC, Forstmann BU (2013) A perspective on terra incognita: uncovering

the neuroanatomy of the human subcortex. Front Neuroanat 3:7–40
2. Amunts K, Zilles K (2012) Architecture and organizational principles of Broca’s region. Trend
Cogn Sci 16(8):418–426
3. Anwander A, Tittgemeyer M, Cramon von D, Friederici A, Knosche T (2006) Connectivity-
based parcellation of Broca’s area. Cereb Cortex 17(4):816–825
4. Bazin P-L, Weiss M, Dinse J, Schäfer A, Trampel R, Turner R (2013) A computational
framework for ultra-high resolution cortical segmentation at 7 Tesla. Neuroimage 2:201–209
5. Behrens T, Johansen-Berg H, Woolrich MW, Smith SM, Wheeler-Kingshott C, Boulby PA
et al (2003) Non-invasive mapping of connections between human thalamus and cortex using
diffusion imaging. Nat Neurosci 6(7):750–757
6. Berquin PC, Giedd JN, Jacobsen LK, Hamburger SD, Krain AL, Rapoport JL, Castellanos
FX (1998) Cerebellum in attention-deficit hyperactivity disorder: a morphometric MRI study.
Neurology 50(4):1087–1093
7. Bogacz R, Wagenmakers E-J, Forstmann BU, Nieuwenhuis S (2010) The neural basis of the
speed-accuracy tradeoff. Trend Neurosci 33(1):10–16
8. Braak H (1980) Architectonics of the human telencephalic cortex. Springer, Berlin
9. Brodmann K (1909) Vergleichende Lokalisationslehre der Grosshirnrinde in ihren Prinzipien
dargestellt aufgrund des Zellenbaues. Johann Ambrosius Barth, Leipzig
10. Brown SD, Heathcote A (2008) The simplest complete model of choice response time: linear
ballistic accumulation. Cogn Psychol 57(3):153–178
11. Burdach KF (1819) Vom Baue und Leben des Gehirns. Dyk’schen Buchhandlung, Leipzig, pp
1–285
12. Bürgel U, Amunts K, Hoemke L, Mohlberg H, Gilsbach JM, Zilles K (2006) White matter fiber
tracts of the human brain: three-dimensional mapping at microscopic resolution, topography
and intersubject variability. Neuroimage 29(4):1092–1105
13. Campbell AW (1905) Histological studies on the localisation of cerebral function. Cambridge
University Press, Cambridge
14. Catani M, Ffytche DH (2010) On the study of the nervous system and behaviour. Cortex
46(1):106–109
15. Catani M, de Schotten MT (2008) A diffusion tensor imaging tractography atlas for virtual in
vivo dissections. Cortex 44(8):1105–1132
16. Catani M, de Schotten MT (2012) Atlas of human brain connections, 1st edn. Oxford University
Press, Oxford, pp 1–515
17. Cha Y-K (1991) Effect of the global system on language instruction, 1850–1986. Sociol Educ
64(1):19–32
18. Cho ZH, Min HK, Oh SH, Han JY, Park CW, Chi JG et al (2010) Direct visualization of
deep brain stimulation targets in Parkinson disease with the use of 7-tesla magnetic resonance
imaging. J Neurosurg 113:1–9
19. Choi H-J, Zilles K, Mohlberg H, Schleicher A, Fink GR, Armstrong E, Amunts K (2006)
Cytoarchitectonic identification and probabilistic mapping of two distinct areas within the
anterior ventral bank of the human intraparietal sulcus. J Compar Neurol 495(1):53–69
20. Derrfuss J, Mar RA (2009) Lost in localization: the need for a universal coordinate database.
Neuroimage 48(1):1–7
21. Desikan RS, Ségonne F, Fischl B, Quinn BT, Dickerson BC, Blacker D et al (2006) An au-
tomated labeling system for subdividing the human cerebral cortex on MRI scans into gyral
based regions of interest. Neuroimage 31(3):968–980
22. Devlin JT, Sillery EL, Hall DA, Hobden P, Behrens TEJ, Nunes RG et al (2006) Reliable
identification of the auditory thalamus using multi-modal structural analyses. Neuroimage
30(4):1112–1120
23. Dronkers NF, Plaisant O, Iba-Zizen MT, Cabanis EA (2007) Paul Broca’s historic cases: high
resolution MR imaging of the brains of Leborgne and Lelong. Brain 130(5):1432–1441
4 An Introduction to Human Brain Anatomy 87

24. Dunnen DWF, Staal MJ (2005) Anatomical alterations of the subthalamic nucleus in relation
to age: a postmortem study. Mov Disord 20(7):893–898
25. Duvernoy MH (1999) The human brain, 2nd edn. Springer, Wien
26. Economo C, Koskinas GN (1925) Die Cytoarchitektonik der Hirnrinde des erwachsenen
Menschen. Springer, Wien
27. Eickhoff SB, Stephan KE, Mohlberg H, Grefkes C, Fink GR, Amunts K, Zilles K (2005a) A
new SPM toolbox for combining probabilistic cytoarchitectonic maps and functional imaging
data. Neuroimage 25(4):1325–1335
28. Eickhoff S, Walters NB, Schleicher A, Kril J, Egan GF, Zilles K et al (2005b) High-resolution
MRI reflects myeloarchitecture and cytoarchitecture of human cerebral cortex. Hum Brain
Mapp 24(3):206–215
29. Evans AC, Janke AL, Collins DL, Baillet S (2012) Brain templates and atlases. Neuroimage
62(2):911–922
30. Federative Committee on Anatomical Terminology (1998) Terminologia Anatomica. (Federa-
tive Committee on Anatomical Terminology, Ed.). Thieme, New York, pp 1–292
31. Flechsig P (1920)Anatomie des menschlichen Gehirns und Rückenmarks auf myelogenetischer
Grundlage. Thieme, Leipzig, pp 1–121
32. Fodor J (1999) Diary. London Rev Books 21(9):68–69. http://www.lrb.co.uk/v21/n19/jerry-
fodor/diary. Accessed 1 April 2013
33. Forstmann BU, Anwander A, Schafer A, Neumann J, Brown S, Wagenmakers E-J et al (2010)
Cortico-striatal connections predict control over speed and accuracy in perceptual decision
making. Proc Natl Acad Sci U S A 107(36):15916–15920
34. Forstmann BU, Keuken MC, Jahfari S, Bazin PL, Neumann N, Schafer A et al (2012) Cortico-
subthalamic white matter tract strength predict interindividual efficacy in stopping a motor
response. Neuroimage 60:370–375
35. Friston K (2002a) Beyond phrenology: what can neuroimaging tell us about distributed
circuitry? Ann Rev Neurosci 25(1):221–250
36. Friston KJ (2002b) Bayesian estimation of dynamical systems: an application to fMRI.
Neuroimage 16(2):513–530
37. Geyer S, Weiss M, Reimann K, Lohmann G, Turner R (2011) Microstructural parcellation
of the human cerebral cortex-from Brodmann’s post-mortem map to in vivo mapping with
high-field magnetic resonance imaging. Front Hum Neurosci. doi:10.3389/fnhum.2011.00019
38. Gottwald B, Mihajlovic Z, Wilde B, Mehdorn HM (2003) Does the cerebellum contribute to
specific aspects of attention? Neuropsychologia 41(11):1452–1460
39. Heidemann RM, Ivanov D, Trampel R, Fasano F, Meyer H, Pfeuffer J, Turner R (2012) Isotropic
submillimeter fMRI in the human brain at 7 T: combining reduced field⣳of⣳view imaging
and partially parallel acquisitions. Magn Reson Med 68(5):1506–1516
40. Herculano-Houzel S (2010) Coordinated scaling of cortical and cerebellar numbers of neurons.
Front Neuroanat. doi:10.3389/fnana.2010.00012
41. Johansen-Berg H, Behrens T, Robson MD, Drobnjak I, Rushworth M, Brady JM et al (2004)
Changes in connectivity profiles define functionally distinct regions in human medial frontal
cortex. Proc Natl Acad Sci U S A 101(36):13335–13340
42. Kachlik D, Baca V, Bozdechova I, Cech P, Musil V (2008) Anatomical terminology and
nomenclature: past, present and highlights. Surg Radiol Anat 30(6):459–466
43. Kandel ER, Schwartz JH, Jessell T (2000) Principles of neural science, 4th edn. McGraw-Hill,
New York, pp 1–1414
44. Kennedy DN, Lange N, Makris N, Bates J, Meyer J, Caviness VS (1998) Gyri of the human
neocortex: an MRI-based analysis of volume and variance. Cerebral Cortex 8(4):372–384 (New
York: 1991)
45. Keren NI, Lozar CT, Harris KC, Morgan PS, Eckert MA (2009) In vivo mapping of the human
locus coeruleus. Neuroimage 47(4):1261–1267
46. Keuken MC, Bazin P-L, Crown L, Hootsmans J, Laufer A, Muller-Axt C, Sier R, van der
Putten EJ, Schafer A, Turner R, Forstmann BU (2014) Quantifying inter-individual anatomical
variability in the subcortex using 7T structural MRI. Neuroimage 94:40–46
88 B. U. Forstmann et al.

47. Keuken MC, Bazin PL, Schafer A, Neumann J, Turner R, Forstmann BU (2013) Ultra-High 7
T MRI of structural age-related changes of the subthalamic nucleus. J Neurosci 33(11):4896–
4900
48. Kitajima M, Korogi Y, Kakeda S, Moriya J, Ohnari N, Sato T et al (2008) Human subthalamic
nucleus: evaluation with high-resolution MR imaging at 3.0 T. Neuroradiology 50(8):675–681
49. Knecht S, Dräger B, Deppe M, Bobe L, Lohmann H, Flöel A et al (2000) Handedness and
hemispheric language dominance in healthy humans. Brain 123(12):2512–2518
50. Lehéricy S (2004) 3-D diffusion tensor axonal tracking shows distinct SMA and Pre-SMA
projections to the human striatum. Cereb Cortex 14(12):1302–1309
51. Leuze CWU, Anwander A, Bazin PL, Dhital B, Stuber C, Reimann K et al (2012)
Layer-specific intracortical connectivity revealed with diffusion MRI. Cerebral Cortex.
doi:10.1093/cercor/bhs311
52. Makris N, Goldstein JM, Kennedy D, Hodge SM, Caviness VS, Faraone SV et al (2006)
Decreased volume of left and total anterior insular lobule in schizophrenia. Schizophr Res
83(2–3):155–171
53. Mazziotta JC, Toga AW, Evans A, Fox P, Lancaster J (1995) A probabilistic atlas of the human
brain: theory and rationale for its development the international consortium for brain mapping
(ICBM). Neuroimage 2(2PA):89–101
54. Mazziotta J, Toga A, Evans A, Fox P, Lancaster J, Zilles K et al (2001) A probabilistic atlas and
reference system for the human brain: international consortium for brain mapping (ICBM).
Philos Trans R Soc B Biol Sci 356(1412):1293–1322
55. Mesulam M (2005) Imaging connectivity in the human cerebral cortex: the next frontier? Ann
Neurol 57(1):5–7
56. Morel A, Magnin M, Jeanmonod D (1997) Multiarchitectonic and stereotactic atlas of the
human thalamus. J Compar Neurol 387(4):588–630
57. Ngowyang G (1934) Die Cytoarchitektonik des menschlichen Stirnhirns I. Cytoarchitektonis-
che Felderung der Regio granularis und Regio dysgranularis. Monogr Natl Res Inst Psychol
Acad Sin (Shanghai) 7:1–68
58. Oishi K, Faria AV, van Zijl PC, Mori S (2010) MRI atlas of human white matter, 2nd ed.
Academic, Waltham, pp 1–266
59. Ono M, Kubik S, Abernathey CD (1990) Atlas of the cerebral sulci. Thieme, New York, pp
1–232
60. Paluzzi A, Belli A, Bain P, Viva L (2007) Brain “imaging”in the Renaissance. J R Soc Med
100(12):540–543
61. Passingham RE, Stephan KE, Kötter R (2002) The anatomical basis of functional localization
in the cortex. Nat Rev Neurosci 3(8):606–616
62. Penhune VB, Zatorre RJ, MacDonald JD, Evans AC (1996) Interhemispheric anatomical dif-
ferences in human primary auditory cortex: probabilistic mapping and volume measurement
from magnetic resonance scans. Cerebral Cortex 6(5):661–672 (New York: 1991)
63. Purves D,Augustine GJ, Fitzpatrick D, Hall WC, LaMantiaAS, White LE (2012) Neuroscience,
5th edn. Sinauer Associates Inc, Massachusetts, pp 1–833
64. Rademacher J, Morosan P, Schormann T, Schleicher A, Werner C, Freund HJ, Zilles K
(2001) Probabilistic mapping and volume measurement of human primary auditory cortex.
Neuroimage 13(4):669–683
65. Rajkowska G, Goldman-Rakic PS (1995a) Cytoarchitectonic definition of prefrontal areas in
the normal human cortex: I. Remapping of areas 9 and 46 using quantitative criteria. Cerebral
Cortex 5(4):307–322 (New York: 1991)
66. Rajkowska G, Goldman-Rakic PS (1995b) Cytoarchitectonic definition of prefrontal areas in
the normal human cortex: II. Variability in locations of areas 9 and 46 and relationship to the
Talairach coordinate system. Cerebral Cortex 5(4):323–337 (New York: 1991)
67. Reil JC (1809) Die Sylvische Grube oder das Thal, das gestreifte große Hirnganglium, dessen
Kapsel und die Seitentheile des großen Gehirns. Arch Physiol 9:195–208
68. Reil JC (1812) Die vördere Commissur im großen Gehirn. Arch Physiol 11:89–100
4 An Introduction to Human Brain Anatomy 89

69. Sarkisov SA, Filimonoff IN, Kononowa EP, Preobraschenskaja IS, Kukuew EA (1955) Atlas
of the cytoarchitectonics of the human cerebral cortex. Medgiz, Moskow
70. Schmahmann JD, Pandya DN (2009) Fiber pathways of the brain. Oxford University Press,
Oxford, pp 1–654
71. Smith G (1907) A new topographical survey of the human cerebral cortex, being an account of
the distribution of the anatomically distinct cortical areas and their relationship to the cerebral
sulci. J Anat Physiol 41(Pt 4):237
72. Standring S (2008) Gray’s anatomy, 40 edn. Elsevier, Amsterdam, pp 1–1576
73. Stoodley CJ, Schmahmann JD (2009) Functional topography in the human cerebellum: a
meta-analysis of neuroimaging studies. Neuroimage 44(2):489–501
74. Strasburger EH (1937) Die myeloarchitektonische Gliederung des Stirnhirns beim Menschen
und Schimpansen. Journal fur psychologie und neurologie 47(6):565–606
75. Swaab DF (2003) The human hypothalamus: basic and clinical aspects. Part 1: nuclei of
the human hypothalamus. In: Aminoff MJ, Boller F, Swaab DF (eds) Handbook of clinical
neurology, vol 79. Elsevier, Amsterdam, p 9
76. Swanson LW (2000) What is the brain? Trends Neurosci 23(11):519–527
77. Swanson LW, Bota M (2010) Foundational model of structural connectivity in the nervous
system with a schema for wiring diagrams, connectome, and basic plan architecture. Proc Natl
Acad Sci U S A 107(48):20610–20617
78. Talairach J, Tournoux P (1988) Co-planar stereotaxic atlas of the human brain. Thieme, New
York, pp 1–122
79. Thompson PM, Schwartz C, Lin RT, Khan AA, Toga AW (1996) Three-dimensional statistical
analysis of sulcal variability in the human brain. J Neurosci 16(13):4261–4274
80. Thompson PM, Cannon TD, Narr KL, Van Erp T, Poutanen V-P, Huttunen M et al (2001)
Genetic influences on brain structure. Nat Neurosci 4(12):1253–1258
81. Turner R (2012) Neuroscientific applications of high-field MRI in humans. In: Hennig J, Speck
O (eds) High-Field MR imaging. Springer, Berlin
82. Vesalius A (1543) De humani corporis fabrica libri septem. School of medicine, Padua
83. Vogt O (1910) Die myeloarchitektonische Felderung des menschlichen Stirnhirns. J Psychol
Neurol 15(4/5):221–232
84. Vogt C, Vogt O (1926) Die vergleichend-architektonische und die vergleichend-
reizphysiologische Felderung der Großhirnrinde unter besonderer Berücksichtigung der
menschlichen. Naturwissenschaften 14(50):1190–1194
85. Waehnert MD, Dinse J, Weiss M, Streicher MN, Waehnert P, Geyer S et al (2013) Anatomically
motivated modeling of cortical laminae. NeuroImage. doi:10.1016/j.neuroimage.2013.03.078