Professional Documents
Culture Documents
8 mm 178 x 254 mm
LIN
Provides the latest neuroimaging-based evidence
on the brain mechanisms of oral functions
Dental Neuroimaging: The Role of the Brain in Oral Functions provides an up-to-date
overview of neuroimaging research on the neural mechanisms underlying mastication,
swallowing, sensory processing, and other oral topics. DENTAL
NEUROIMAGING
Divided into three parts, the book first introduces the theoretical framework of the
brain-stomatognathic axis, clinical assessments for oral function, and neuroimaging
methods. The second part presents recent neuroimaging findings of oral sensory and
motor functions such as somatosensation, gustation, and orofacial pain and anxiety.
DENTAL NEUROIMAGING
The book concludes with a review of recent translational research and discussion of
the application of neuroimaging in clinical management. Throughout the text, boxed
sections highlight key information about cognitive neuroscience, imaging techniques, THE ROLE OF THE BRAIN IN ORAL FUNCTIONS
interpreting neuroimaging results, and relating research findings to clinical practice.
Offering a systematic introduction to brain science and how it relates to dental medicine,
Dental Neuroimaging: The Role of the Brain in Oral Functions is essential reading for
students and researchers in disciplines such as neuroscience, neuroanatomy, oral
physiology, dentistry and oral healthcare, speech therapy, and oral rehabilitation.
Chia-Shu Lin, Professor, Department of Dentistry, National Yang Ming Chiao Tung
University (NYCU), Taiwan. Dr Lin is one of the few researchers specializing in both
clinical dentistry and human brain science, focusing on neuroimaging. His current
work explores the brain mechanisms of oral sensorimotor functions and the cognitive-
affective processing of pain and dental anxiety.
www.wiley.com/go/dentistry
0005217271.INDD 6 11/19/2021 09:40:23
Dental Neuroimaging
Chia-Shu Lin
Department of Dentistry
National Yang Ming Chiao Tung University
Taipei, Taiwan, Republic of China
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any
form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice
on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.
The right of Chia-Shu Lin to be identified as the author of this work has been asserted in accordance with law.
Registered Offices
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
Editorial Office
9600 Garsington Road, Oxford, OX4 2DQ, UK
For details of our global editorial offices, customer services and more information about Wiley products, visit us at
www.wiley.com.
Wiley also publishes its books in a variety of electronic formats and by print-on-demand. Some content that appears
in standard print versions of this book may not be available in other formats.
10 9 8 7 6 5 4 3 2 1
Contents
List of Figures x
List of Tables xviii
List of Boxes xx
List of Abbreviations xxi
Preface xxiii
Introduction to Students and Instructors xxiv
Acknowledgements xxv
About the Companion Website xxvi
Index 292
List of Figures
1.1 The association between the brain and the stomatognathic system. The traditional
perspective highlights the brain as a ‘systemic factor’ associated with oral health,
just like the factors related to other body systems. The functional perspective
highlights that the brain and mental functions guided by the brain play an essential
role in stomatognathic functions. 4
1.2 A general view of the neural circuitries of the brain mechanisms of orofacial
functions. The circuitries between the central and peripheral sites (i.e. pathways
labelled in blue and red) are investigated primarily via animal models. Notably, the
circuitries within the brain (i.e. the intracortical pathways labelled in black) have not
been fully elucidated. Source: Avivi-Arber and Sessle (2018).Reproduced with
permission of John Wiley and Sons. 9
1.3 Theoretical frameworks of the association between the brain, oral functions and
behaviour. (a) The oral-to-behaviour (OB) framework, (b) the oral-brain-behaviour
(OBB) framework and (c) the brain–stomatognathic axis (BSA). 19
2.1 The general concept of the blood-oxygen-level-dependent (BOLD) mechanism.
(a) Transportation of oxygenated haemoglobin during a resting condition, when
neural activity is low. (b) Transportation of oxygenated haemoglobin when neural
activity increases. The neurons demand more energy by consuming oxygen provided
by oxygenated haemoglobin. Via a complex haemodynamic process (e.g. an
increasing rate and volume of cerebral flow), the amount of oxygenated
haemoglobin increases (relatively to the amount of deoxygenated haemoglobin),
leading to an over-supply or compensation of the oxygen demand from neurons. 26
2.2 Examples of functional magnetic resonance imaging (fMRI) investigation of
chewing movement. (a) The first-level analysis. In a chewing experiment, the task
conditions (i.e. when subjects are chewing) are contrasted to the baseline conditions
(i.e. when subjects are resting). Brain activation reflects the difference in blood-
oxygen-level-dependent (BOLD) signals in the task vs. the baseline condition. The
first-level analysis focuses on the pattern of brain activation at the individual subject.
(b) The second-level analysis. The second-level analysis focuses on the association
between brain activation and individual variability. The association can be explored
by investigating the correlation between brain activation and individual
performance or comparing brain activation between different clinical groups. 31
2.3 Methodological considerations of a functional magnetic resonance imaging study.
(a) Subjects may show great inter-individual variability in their general conditions,
such as sex, age and general physical/psychological conditions. (b) Subjects may
show great inter-individual variability in their personal trait and performance (e.g.
pain ratings) related to an experimental task. (c) Subjects differ in brain morphology.
When individual brains are compared, the individual images are normalized to a
template image, using linear transformation (i.e. translation, rotation, resizing and
shearing) and nonlinear transformation approaches. 33
2.4 Statistical analysis at the individual level and the group level. (a) The analysis at the
individual level focuses on the association between task progression and the
blood‑oxygen-level-dependent (BOLD) time series, as shown in the left panel. For
each voxel, a strong association indicates that the BOLD signals of the voxel can be
predicted by the task condition, in contrast to the baseline condition (e.g. Voxel A),
as shown in the right panel. (b) The analysis at the group level focuses on the
association between brain features (e.g. brain activation of grey matter volume) and
group factors. The association may reflect the difference in brain features between
patient and control groups (the left panel) or the correlation between brain features
and clinical factors (the right panel). (c) A typical image result consists of the
statistical values (e.g. the t-score) from multiple voxels (represented as the grid),
which are visualized by a colour scale, as shown in the left panel. The result can be
thresholded according to intensity (i.e. the t-score). For example, only the voxels
with a t-score >6 are preserved after thresholding, as shown in the middle panel.
The result can be thresholded according to the size of a cluster of voxels. For
example, only the clusters with a size larger than 100 voxels will be preserved after
thresholding, as shown in the right panel. 35
2.5 Experimental design of functional neuroimaging research. (a) Under the assumption
of pure insertion, the difference of brain activation between two experimental
conditions only reflects the mental function contrasted by the conditions (e.g.
perception of pain intensity). However, the contrast may be associated with more than
one mental function (e.g. perception of pain intensity and attention to noxious
stimuli). (b) A factorial design helps to delineate the association between two mental
functions. For example, the light grey area denotes the effect of increased pain on
brain activation, and the dark grey area denotes the effect of increased attention.
(c) A conjunction design focuses on the pattern of brain activation common to two
experimental conditions (e.g. a clenching task and a chewing task). The activation may
reflect the brain mechanisms of a mental function common to both task conditions. 37
2.6 Conceptual differences between functional specialization, functional segregation
and functional integration. (a) Functional specialization highlights the association
between a brain region and a specific mental function. For example, activation at
the occipital lobe is considered mainly for the processing of visual perception.
(b) Functional segregation highlights that a mental function is associated with
multiple brain regions that are functionally connected within a module. For
example, visual cognition is associated with the module consisting of the yellow
regions, and motor control is associated with the module consisting of the blue
regions. (c) Functional integration highlights the pattern of global communication
between multiple brain regions. For example, individual variability in mental
functions may be associated with the efficiency of how information is distributed
in a network. 45
2.7 Analysis of resting-state functional connectivity. (a) The spontaneous blood-oxygen-
level-dependent (BOLD) activity is acquired using resting-state functional magnetic
with a standardized size and shape. The chewed fragments are collected and
photographed. Colour and morphological features (e.g. the area and perimeter) of
each fragment, which reflect individual cutting ability, are assessed by analyzing the
image. Source: Salazar et al. (2020). Reproduced with permission of Elsevier. 60
3.2 Methods of the assessment of oral mixing ability. (a) In the two-colour chewing gum
test, the degree of mixing food can be assessed by the colour hue of chewing gum
with different colours. For example, if a piece of red and a piece of yellow gums are
well mixed, the resulting bolus would in orange homogenously. The hue of the bolus
can be quantified by imaging analysis. A smaller standard deviation of hue
represents a greater homogeneity of colour mixing, i.e. a better mixing ability. (b)
The degree of mixing is assessed according to the pattern of spatial clusters. A piece
of juice chew with red and white portions was chewed by a subject for 20 strokes
and collected, as shown in the left panel. The degree of clustering is assessed based
on the analysis of variogram, which reflects how fine the clusters of different colours
are. A pattern with finer clusters (e.g. the case in the lower-right panel) reflects
better mixing ability. Source: Lo et al. (2020). Reproduced with permission of John
Wiley and Sons. 61
3.3 Experimental design of pain/somatosensory experience. (a) Blood-oxygen-level-
dependent (BOLD) signals are recorded concurrently with discrete ratings of
stimulation. In this design, noxious stimuli with high and low intensities are
followed by a rating phase (‘?’), which requires subjects to rate the pain intensity
they perceive for the stimuli. Brain activation associated with pain can be contrasted
by the phases that subjects feel strong vs. mild pain, according to their ratings (i.e.
the black bars). (b) BOLD signals are recorded concurrently with continuous ratings
of spontaneous pain. Patients with chronic pain continuously rate their pain, which
may increase spontaneously. (c) Brain features and ratings are recorded separately.
In this design, the rating of pain or somatosensory experience is conducted outside a
scanner. Association between the individual ratings (e.g. pain) and their brain
features (e.g. grey matter volume of the insula), which are collected separately, can
be investigated by correlational analysis. 75
4.1 An overview of brain regions associated with motor control. The figure only displays
the relative position and size of the brain regions, not depicting the anatomical details. 94
4.2 Sensorimotor control of the basal ganglia and the cerebellum. (a) The basal ganglia
consist of a direct and an indirect pathway of motor control. In both pathways, the
striatum is activated by the cortex, which forms a loop of control with the thalamus
(the grey arrow). In the direct pathway (the solid black arrow), the activation of the
striatum inhibits the activity of the internal segment of the globus pallidus (GPi) and
the substantia nigra (SNr), which further inhibits thalamic functions. Therefore,
cortical activation is associated with an increased thalamic activity via the direct
pathway. In the indirect pathway (the black dashed line), the activation of the
striatum inhibits the activity of the external segment of the globus pallidus (GPe),
which further inhibits the activity of the subthalamic nucleus (STN). Notably, the
activity of the STN further activates GPi/SNr, which decreases thalamic activity.
Therefore, cortical activation is associated with a decrease in thalamic activity via
the indirect pathway. (b) The cerebellum plays a key role as an internal model of
motor learning. A forward model predicts the sensory outcomes when motor
commands are executed. It adjusts sensorimotor processing via feedback of the
predicted sensory outcomes. An inverse model calculates the motor commands that
would produce the sensory outcomes from desired actions.
According to Wolpert et al. (2001), both models are crucial to motor control. 96
4.3 Experimental design for neuroimaging of the brain mechanisms of chewing. (a) The
basic concept of study design. The study consists of multiple blocks of a chewing
task and a baseline (no-chewing) block. (b) Variations of the study design. Different
studies may differ in the number of blocks of tasks and the definition of the baseline
block (e.g. resting or clenching the teeth). The variations lead to a different
interpretation of imaging results. 100
4.4 Brain activation associated with chewing and clenching. Source: Lin (2018).
Reproduced with permission of John Wiley and Sons. 103
4.5 Experimental design for neuroimaging of the brain mechanisms of swallowing. (a)
Study design of the swallowing tasks, including the water swallowing task and the
saliva swallowing task. (b) Examples of the study design for investigating brain
mechanisms of swallowing. An overt swallowing task (with either water or saliva
swallowing) is associated with the execution of swallowing movement. A covert
swallowing task is associated with the motor planning of swallowing. 111
4.6 Brain activation associated with water (the upper panel) and saliva (the lower panel)
swallowing. Source: Sörös et al. (2009). Reproduced with permission of John Wiley &
Sons, Inc. 112
5.1 Processing of oral somatosensory information. Information from individual sensory
modalities is transduced via peripheral receptors at the level of somatosensation and
integrated to form a holistic perceptual experience (e.g. oral stereognosis) at the level
of somatoperception. Information is further integrated, with knowledge and
affective–motivational experience, to form a feeling of intraoral condition at the
level of somatorepresentation. 129
5.2 Experimental methods of investigating oral mechanoreceptors. (a) Recording signals
from periodontal mechanoreceptors. Source: Trulsson (2006). Reproduced with
permission of John Wiley and Sons. (b) The food splitting task. Source: Grigoriadis
et al. (2017) with permission of Springer Nature under the terms of the Creative
Commons CC BY 4.0 License. 130
5.3 Brain activation associated with gustation at the insula. A consistent pattern of brain
activation is identified in the insular cortex for studies focusing on the quality,
intensity and affective value of taste stimuli, respectively. Source: Yeung et al. (2018).
Reproduced with permission of Elsevier. 137
5.4 Basic concepts of perceptual processing. (a) Top-down processing highlights the
neural processing of intrinsic (personal) factors, such as one’s goal planning, on the
formation of perceptual experience. The bottom-up processing highlights the neural
processing of extrinsic (environmental) factors, such as the physical features of
stimuli, on the formation of perceptual experience. (b) In predictive coding, the
sensory inputs that we receive from the real world are compared with our prediction
of the sensory outcomes. A mismatch (i.e. ‘prediction error’) occurs when our
prediction does not fit the outcome we perceive. The prediction error is associated
with attentional bias and learning. For example, we may pay more attention to an
unexpected event compared to an expected one. 142
5.5 Experimental design of manipulating threat value associated with pain. (a) The
presence of noxious stimuli is associated with visual cues, which predict
from the cross-sectional design only suggest, but not confirm, the causal direction of
a plastic effect. The difference in brain features (as observed by neuroimaging) may
be attributed to prior experience of training (the middle panel). However, it is also
possible that the individual differences in the brain features predispose their
performance of a skill (the right panel). (b) To better elucidate the plastic effect of
training, a longitudinal design is used to assess the performance and brain features
at different stages of training. Importantly, one can assess whether the plastic effect,
i.e. changes in brain features during the training session, can last for a period or
vanish right after the termination of training. (c) The same longitudinal design also
helps to elucidate individual differences in their performance. For example, the
variation in brain features may account for the difference in learning speed. 258
9.1 Major aims of brain–stomatognathic integrative assessment (BSIA). (a) Prediction of
long-term changes in oral functions. Individual differences in brain reserve and
cognitive reserve may play a key role in their susceptibility to diseases. The BSIA,
which includes the assessment of cognitive functions of older people, would help to
predict the future condition of individual oral health. (b) Classification of patients
with different risks of oral diseases. The BSIA helps to classify the patients for their
risk of oral diseases and the prognosis of treatment, based on a full-scale assessment
of general physical and mental status. 282
9.2 Key elements for implementing the brain–stomatognathic integrative assessment.
(a) The multidisciplinary investigation can be undertaken in a hospital, where
different departments are in charge of different assessments. Critically, the results of
an assessment from one discipline (e.g. the score of cognitive tests from
neurologists) are distributed for the use of other disciplines (e.g. dentistry). For
example, the cognitive performance of older patients, as assessed by neurologists, is
available for prosthodontists to evaluate if patients can adapt well to their new
dentures. (b) In contrast to the hospital setting, a home-based assessment can be
facilitated by the use of teledentistry approaches and digital technology. For
example, in long-term care institutes or at home, patients can record their own oral
status by photographing (via a smartphone). The image record may include a photo
about their intraoral conditions (e.g. bleeding gum) or performance of oral functions
(e.g. the food bolus after chewing). The images are sent to the cloud storage service
for further analysis. A preliminary assessment is performed automatically by
machine-learning-based algorithms, and critical problems (e.g. a poor oral mixing
ability) are screened and forwarded to dental professionals for further evaluation. 284
9.3 An example of neuroimaging investigation combined with animal research. Using
structural MRI, Avivi-Arber et al. (2017) compared the post-mortem brain volume
between the mice that received molar extraction and those who received sham
operation. Tooth extraction was associated with a widespread reduction in volumes
of brain regions of sensorimotor and cognitive–affective processing. Source: Avivi-
Arber et al. (2017) with permission of Frontiers Media S.A. under the terms of the
Creative Commons Attribution 4.0 License. 289
List of Tables
8.1 Neuroimaging research on brain mechanisms of brain plasticity and oral functions
(since 2010). 269
8.2 Neuroimaging research on brain mechanisms of training/intervention to improve
swallowing (since 2015). 273
9.1 Proposed components of the brain–stomatognathic integrative assessment (BSIA). 283
9.2 Recent findings on neuroimaging research on oral functions based on animal
models (since 2015). 287
List of Boxes
List of Abbreviations
Preface
About 20 years ago when I just finished my brain, the stomatognathic system and oral
DDS programme, new frontiers were rapidly health-related behaviour. As shown in the book,
expanding in dentistry and brain science. neuroimaging studies have provided many
Dentistry was markedly revolutionized by insights on the mechanisms of oral diseases,
innovations in biological, material and digital such as chronic orofacial pain and swallowing
technology. Brain science, with the use of and masticatory dysfunctions. Moreover, as I
non-invasive neuroimaging approaches, has have emphasized in this book, the knowledge
risen as a mainstream field in clinical science. from neuroimaging research also contributes to
At the beginning of the twenty-first century, translational application in clinical manage-
we have witnessed the advancement of neuro- ment, such as the management of geriatric and
imaging in various health-related fields, such special needs patients.
as neurodegenerative disorders and psychiat- Undeniably, a single book will not cover all
ric diseases. In contrast, our knowledge of the the progress of neuroimaging research related
stomatognathic system is largely based on ani- to oral health. The book focuses on the future
mal and clinical research, and the brain’s role directions and challenges of dental neuroim-
in oral functions has not been fully elucidated. aging research rather than merely a conclusion
During my college days, the topic of the human of past knowledge about the brain and oral
brain and mental functions has been rarely functions. In this book, while the key results
discussed in dental textbooks. from classic research works are summarized,
This book responds to the evolution of using more emphasis is put on recent findings and
neuroimaging approaches to investigate the the current trend in neuroimaging research of
human brain and oral functions, which has oral topics.
been lasting for decades and become gradu- As an investigator who is lucky enough to
ally popular in recent years. As a relatively witness the rise of this cross-disciplinary field,
uncharted field, ‘dental neuroimaging’ is not I hope you enjoy this book and that you would
merely an adoption of a new method in dental appreciate how neuroimaging helps to unravel
research. It also focuses on the cross-disciplinary the mysterious connection between the brain
investigation about the connection between the and oral functions.
This book consists of nine inter-related chap- ageing and oral function (Chapter 7), the brain
ters, which are organized into three parts. Part I mechanisms of the plasticity and adaptation of
Methods of Neuroimaging and Assessment of oral functions (Chapter 8) and future direc-
Oral Functions focuses on the methodological tions about the translation application of neu-
issues of the research of the association between roimaging in clinical management (Chapter 9).
the human brain and the stomatognathic func- To readers who explore the field of brain
tions, which covers an introduction to theoreti- imaging for the first time, key information
cal frameworks of the brain–stomatognathic regarding imaging and brain science can be
connection (Chapter 1), an introduction to found in the following supplementary materials.
neuroimaging methods, with a focus on
●● From the Brain to Behaviour: These text boxes
magnetic resonance imaging (Chapter 2), and
provide further information regarding the
an introduction to clinical assessments of oral
basic information in cognitive neuroscience.
functions (Chapter 3).
●● From Research to Practice: These text boxes
Part II Neuroimaging Research of Brain
provide suggestions about how to properly
Mechanisms of Oral Functions focuses on recent
interpret a neuroimaging finding and how
neuroimaging findings of oral sensory and motor
the findings are related to clinical questions.
functions, including the brain mechanisms of
●● From Tools to Discovery: These online mate-
oral motor functions (e.g. mastication and swal-
rials include video courses that demon-
lowing) (Chapter 4), oral sensory functions (e.g.
strate how to use basic research tools in
somatosensation and gustation) (Chapter 5) and
neuroimaging.
orofacial pain and anxiety (Chapter 6).
Finally, in Part III Translational Research of Further correspondence from students, faculty
Dental Neuroimaging, the issues related to clin- and clinicians is welcome.
ical applications of dental neuroimaging are Please send the message to
reviewed, including the association between Chia-Shu Lin; e-mail: winzlin@nycu.edu.tw.
Acknowledgements
Writing this book echoes my career as a neuroscientist and a dentist. I am grateful to Dr Shin-
Fang Yang and Dr. Mei-Yu Chen, who showed me the road of being a scientist, Dr Irene Tracey
and Dr Katja Wiech, who introduced functional neuroimaging to me during my DPhil study at
the University of Oxford, and Dr Jen-Chuen Hsieh, who inspired me with many insights on
brain science during my postdoctoral research at Taipei Veterans General Hospital.
My exploration in dental neuroimaging, a new frontier in dentistry, is greatly benefited from
the discussion with my colleagues from dental research: Dr Ming-Lun Hsu, Dr Shyh-Yuan Lee,
Dr Kuo-Wei Chang and Dr Ching-Yi Wu, and colleagues in from brain research: Dr David Niddam
and Dr Li-Fen Chen, at National Yang Ming Chiao Tung University. The exploration cannot go
further without the professional support from the 3T MRI Core Facility at National Yang Ming
Chiao Tung University.
I am grateful to the managers and editors from Wiley Publishing: Associate Commissioning
Editor Loan Nguyen, Managing Editor Tanya McMullin, Content Refinement Specialist Samras
Johnson Vanathaiya, Copyeditors Aravind Kannankara and Rathi Aravind, and Editorial
Publication Coordinator Susan Engelken. I have been much benefited from their professional
assistance in preparing this book.
Finally, I am lucky to see things just because I stand on the shoulders of many giants – the
researchers and clinicians who devote themselves in oral medicine and neuroscience. This book
will just be empty without their pioneering works. They are the true explorers in this new frontier.
www.wiley.com/go/lin/dental-neuroimaging
Part I
Dental Neuroimaging: The Role of the Brain in Oral Functions, First Edition. Chia-Shu Lin.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
Companion website: www.wiley.com/go/lin/dental-neuroimaging
Figure 1.1 The association between the brain and the stomatognathic system. The traditional perspective
highlights the brain as a ‘systemic factor’ associated with oral health, just like the factors related to other
body systems. The functional perspective highlights that the brain and mental functions guided by the
brain play an essential role in stomatognathic functions.
pain. Subsequently, the author put forward epileptic signs and symptoms and irregular
some insightful speculation on orofacial pain: behaviour in eating and drinking
(Anderson 1790). The finding echoes the link
Cerebral diseases, for example, insanity, between the brain and oral sensorimotor func-
softening of the brain, tumors and tions, extensively studied in animal research
inflammation may produce odontalgia, (Lund 1991). New issues have emerged in mod-
but clinical reports reveal comparatively ern days. For example, can older individuals be
few, inasmuch owing to their obscurity benefited from oral functional training to
positive diagnosis is often rendered dif- improve mastication and swallowing
ficult (Hayes 1889). (Sessle 2019)? Can patients with neurodegenera-
tive disorders, who have deficits in mental func-
Though not scientifically accurate from the tions, also improve their oral functions? There
modern view, the statement points out the com- are more challenges to meet for the old alliance
plex association between the brain and orofacial between dentistry and brain science.
pain, which has confused dentists for more than
one century. The alliance becomes cemented
due to the challenge of treating orofacial pain, 1.1.3 Dental Education: The Role
and new technologies, including neuroimaging, of Neuroscience and the Brain
have provided new insights into this field (see In the previous section, we have briefly dis-
Chapter 6). Our second evidence comes from cussed how the research of the brain has been
the issues of infection control, especially the linked to issues of oral health. However, the
brain abscess secondary to dental infection. At discussion may not be complete without look-
present, dentists have been highly aware of ing into dental education for the following
infection control within the oral cavity. However, questions: has the role of brain science been
new challenges have emerged, such as the recent recognized in dental education?
debates on the neuroinflammatory mechanisms
that may underlie the link between neurodegen- 1.1.3.1 The Tradition of ‘Dentists as Surgeons’
erative disorders and periodontal diseases (see A discussion of early dental education will not
Chapter 7). Finally, the third evidence of the old be complete without mentioning the contribu-
alliance has an even longer history. Back in 1790, tion from Pierre Fauchard, widely recognized
when the terms ‘brain science’ and ‘dentistry’ as the Father of Modern Dentistry, with the
have not yet popularized, in an article entitled first textbook of dentistry Le Chirurgien Dentiste
‘Pathological Observations on the Brain’, the (‘The Surgeon Dentist’) published in 1728. As
author reported a potential association between the name suggests, dentistry is the discipline of
managing dental diseases with a surgeon’s dental schools in North American. In 2014,
training. Notably, in this book, Fauchard has among 66 dental schools, 31 (47%) offered neu-
extended the professional domain of dentists roscience as a standalone course, with the oth-
from ‘teeth’ to the oral cavity (including the ers integrated the neuroscience topics into
soft tissue). The new profession, a ‘surgeon other courses (Gould et al. 2014). It is also
dentist’, is different from a ‘toothpuller’ in the noteworthy that in most dental schools, the
seventeenth to eighteenth centuries (Lynch course was delivered by teachers from medical
et al. 2006). Though he also emphasized the schools, who may not tailor-make the course
relationship between oral and systemic dis- for dental students (Gould et al. 2014). The
eases (Lynch et al. 2006), the primary task for average year of teaching of the teachers is
dentists is to fix the structural deficits of the relatively high (23.1 years), suggesting fewer
oral cavity, such as restoring a decayed tooth or younger teachers are involved in the field
replacing the missing teeth with a denture. All (Gould et al. 2014). Critically, the topics to be
the jobs require dentists to be capable of per- delivered significantly varied between courses.
forming complicated surgical skills. Some topics, such as the knowledge of cranial
An over-focus on the surgical skills of dental nerves, were taught averagely for three hours.
treatment, however, had gradually received In contrast, issues of the neuropathic mecha-
criticism since the early days when dental edu- nisms of pain, including nerve regeneration,
cation became an independent discipline. As neuralgia, allodynia and hyperalgesia, were
pointed by Eugene Talbot early in 1900: taught less than half an hour (Gould
et al. 2014). Topics related to the human brain
The result is that study of the general dis- were taught in most of the courses.
eases which affect the mouth, jaws and Nevertheless, among the 31 independent neu-
teeth have been neglected. Limitations of roscience courses, almost half of them focused
a dental education have prevented the on neuroanatomy, which emphasizes the
dentist from associating local diseases knowledge of brain structure rather than the
with systemic causes (Talbot 1900). link between the brain and oral functions. This
alienation reflects that many courses were
The statement corresponds to the degree taught by personnel outside the dental schools
delivered for this new profession, namely and may not provide what dentists need to
Doctor of Dental Surgery (DDS), at Talbot’s know for their clinical careers.
time. He further showed the concern that ‘. . . Therefore, for teaching neuroscience and
the graduate of dental surgery is not competent brain science in dental schools, the real chal-
to associate systemic diseases with their effects lenge is not the time and classes allocated for
on the teeth, nor is he capable of appreciating teaching, but how these materials are taught.
systemic lesions due to overtreatment of patho- Non-dental school faculties mainly taught the
logic conditions of the teeth’ (Talbot 1900). The courses, and the topics were less tailored for
gap between a dentist and medical knowledge dentistry. For example, in some syllabi of the
would make dentists ignore the systemic condi- neuroscience courses, the issue ‘pain’ is taught
tion of patients – moreover, the ignorance may alongside somatosensation. Nowadays, we
further exacerbate systemic health when den- have much evidence showing that pain, as a
tists ‘overtreat’ patients (Talbot 1900). more generalized cognitive–affective experi-
ence, is associated with the brain mechanisms
1.1.3.2 Brain and Neuroscience: Is It of attention, emotional and cognitive process-
Neglected in Dental School? ing (see Chapter 6). In this case, a focus on
According to the Basic Science Survey Series of the brain and mental functions, such as the
the American Dental Education Association modulatory effect of attention and cognitive
(ADEA), neuroscience is widely taught in most appraisal on pain, should be tailored for dental
students since it is highly associated with the the National Institute of Dental and
clinical management of patients. Craniofacial Research (NIDCR). This number
is almost twice the number of sponsored pro-
1.1.4 The ‘New Engagement’: Modern jects (53) in the whole 1990s when the NIDCR
Cross-Disciplinary Research of Dentistry funded 29 projects. The results suggest an
and Brain Science increasing trend of cross-disciplinary research
between oral and brain sciences. Critically,
Instead of being a comprehensive textbook
not all the projects were granted by the
on the neurobiology of dentistry, this book
NIDCR, which specializes in orofacial medi-
aims to outline the ‘new engagement’
cine. Several projects were supported by the
between dentistry and brain science, with
National Institute of Mental Health and the
neuroimaging as a critical approach to bridge
National Institute on Aging, highlighting
the two fields. Here, we discuss the trend of
the importance of oral issues in cognitive def-
cross-disciplinary research between dentistry
icits and aging.
and brain science, according to two brief bib-
liometric surveys. Firstly, a survey based on
PubMed was performed by the keywords 1.1.5 Summary
‘dental’ and ‘brain’ and the search was lim-
●● From a functional perspective, the brain,
ited to titles and abstracts of the literature
behaviour and oral health are directly linked
(tooth[mesh] OR oral[mesh] OR dental[mesh]
because the brain plays a crucial role in
OR dentistry[mesh] OR teeth[tiab] OR
maintaining oral functions, and the integrity
tooth[tiab] OR oral[tiab] OR dental[tiab] OR
of mental functions is critical to maintaining
dentistry[tiab] AND brain[tiab]). The find-
oral health.
ings revealed that by December 2020, 20261
●● The alliance between the research on den-
research papers had been documented in
tistry and the brain has a long history. It con-
PubMed. The number of publications shows
tributes to tackling unsolved challenges (e.g.
a pronounced rise in recent years, which
orofacial pain) and new challenges (e.g.
almost doubled within 10 years. For example,
aging and oral functions).
between 1980 and 1989, the number of publi-
●● Topics of neuroscience and the brain are not
cations n = 1566. This number rose from 1990
neglected in dental education. However,
to 1999 (n = 2684) and almost doubled from
many courses are taught by non-dental
2000 to 2009 (n = 4721). From 2010 to 2019,
school faculties, and the topics were less
the number doubled again to n = 9331.
tailored for dentistry.
As discussed in Section 1.2, the increasing
●● Recently, cross-disciplinary research on oral
number of publications on the brain topic
and brain sciences has quickly emerged
corresponds to the increasing number of
in the number of publications and research
publications on neuroimaging, which has
grants.
become a pivotal method in studying the
human brain.
A second survey was conducted by search-
1.2 What Is Neuroimaging?
ing for the past and current research projects
funded by the National Institutes of Health
1.2.1 Introduction
(NIH), USA, using the online platform of
Research Portfolio Online Reporting Tools In Section 1.1, we have highlighted a significant
(RePORT) report. From 2020 to February overlap between dentistry and brain science.
2021, the keywords ‘dental’ and ‘brain’ have Though the two fields are closely linked from a
led to 106 projects, with 39 projects funded by functional perspective, there exists a vast
difference between research approaches of the their roles in brain science and practical implica-
oral cavity and those of the brain. Dentists can tions in dentistry are highlighted.
visually examine the oral structure, and oral
functions can be quantified with a chairside set
1.2.2 What Is the Role of Neuroimaging
of assessments. In contrast, brain functions and
Research in Dentistry
mental status, sometimes metaphorized as a
‘black box’, can hardly be examined directly at 1.2.2.1 Trends of Research Publications
the chairside. Therefore, a pivotal step to facili- in Dental Neuroimaging Research
tate the investigation of the brain is to develop Table 1.1 summarizes the number of dental
the technology for quantifying brain structure research publications combined with research
and functions. Neuroimaging, defined as a non- on the brain and neuroimaging, according to a
invasive approach of ‘visualizing the central PubMed-based survey. The trend of publica-
nervous system, especially the brain, by various tion of dental research related to neuroimag-
imaging modalities’ (MeSH 2012), is such a ing was strikingly similar to that related to the
technological breakthrough that revolutionizes brain. Before 1970, only a few publications
the research approaches of the brain. were found in these fields. In contrast, after the
A common myth is that neuroimaging or 1990s, the percentage of these studies showed
‘brain scan’ would be a kind of ‘modern magic’. a pronounced increase. Notably, this trend in
The impression is strengthened by some sci-fi publication can be compared with the dental
movies, where ‘peeking into the brain’ is taken research related to neuroscience in general, as
as an icon of something futuristic. Contrary to recently reported by Iwata and Sessle (2019).
the popular myth, the term ‘neuroimaging’ has In terms of brain and neuroimaging research,
been adopted as a common method for regular almost half of the dental research on brain and
clinical investigation and research. In the follow- neuroimaging was published last decade
ing sections, we outline the primary methods of (2011–2020) (Table 1.1). This trend is very dif-
neuroimaging approaches in brain science, and ferent from the general field of neuroscience. It
Table 1.1 Trends of the academic publicationa in dental research related to brain and neuroimaging.
Dentistry + Brain Dentistry + Neuroimaging
is also noteworthy that more than 90% of Figure 1.2). On the other hand, neuroimag-
the dental research on neuroimaging was ing findings disclose new knowledge about
published after the mid-1990s, about the the role of learning and cognitive control in
same time when functional magnetic reso- oral motor functions (Table 1.3). As shown
nance imaging (fMRI) came into practice in Table 1.2, many examples reveal how
(Bandettini 2012). The trend reflects that tech- neuroimaging, as an exploratory tool,
nological innovation of biomedical imaging broadens the frontier of orofacial neurosci-
may facilitate cross-disciplinary research on ence into the uncharted area.
dentistry and the brain.
Thalamus
7
Basal ganglia
1. Motivation, emotion 9 8 6
2. Cognition, attention, memory Brainstem
3. Motorplanning Hippocampus Cerebellum
4. Sensorimotor integration,
execution, control and learning Other subcortical areas Central Afferents
5. Multisensory processing and Pattern Generators Central
(e.g. red nucleus, reticular
integration formation) Peripheral
6. Sensorimotor integration, control, Efferents
timing and learning Central
7. Sensorimotor control and learning Muscle Motor V, VII, XII Peripheral
8. Initiation and modulation of contraction Intracortical
semi-automatic movements
9. Memory and spatial coding Sensory V, STN
Sensory receptors
Figure 1.2 A general view of the neural circuitries of the brain mechanisms of orofacial functions. The circuitries between the central and peripheral sites
(i.e. pathways labelled in blue and red) are investigated primarily via animal models. Notably, the circuitries within the brain (i.e. the intracortical pathways
labelled in black) have not been fully elucidated. Source: Avivi-Arber and Sessle (2018). Reproduced with permission of John Wiley and Sons.
Table 1.2 Selected findings (since 2010)a of neuroimaging research, which are related to the issues of the
‘landmark discoveries or concepts’ of oral neuroscience (Iwata and Sessle 2019), as quoted in field (A) to (G).
Table 1.2 (Continued)
(G) ‘Delineation of peripheral processes and CNS circuits underlying touch, temperature, taste and
salivation, including the discovery of a fifth taste, umami’
Trulsson et al. (2010) Healthy adults fMRI ‘. . . PDLMs, and SA II-type receptors in general, may be
involved in one aspect of the feeling of body ownership’.
Nakamura Healthy adults fMRI ‘The peaks of the activated areas in the middle insular
et al. (2011) cortex by umami were very close to another prototypical
taste quality (salty)’.
ASL-MRI: arterial spin labelling magnetic resonance imaging; CBF: cerebral blood flow; CD: complete denture; CNV:
the trigeminal nerve; DTI: diffusion tensor imaging; fMRI: functional magnetic resonance imaging; sMRI: structural
magnetic resonance imaging; IOD: implant-supported denture; MRS: magnetic resonance spectroscopy; PDLM:
periodontal ligament mechanoreceptor; SA: slowly adapting; TMD: temporomandibular disorders; TN: trigeminal
neuropathy; TNP: trigeminal neuropathic pain; WM: white matter.
a
The survey is performed using Google Scholar, with the date of publication ranged from 1 January 2010 to 30 May 2021.
Source: Field (A) to (G) based on Iwata and Sessle (2019).
Table 1.3 Selected findings (since 2010) of brain imaging research related to the clinical disciplines
of dentistrya.
Prosthodontic For edentulous patients, reduced prefrontal activation associated Kamiya et al. (2016)
treatment with tooth loss may be prevented by chewing with a denture.
Prosthodontic The adaptation to replacement of dentures may be associated with Luraschi et al. (2013)
treatment changes in brain activity during oral motor tasks.
Prosthodontic Adaptative chewing experience induced by palate coverage was Inamochi et al. (2017)
treatment associated with changes in brain activity associated with motor
learning.
Dental implant In rats, tooth loss and installing dental implants may be associated Avivi-Arber
with neuroplasticity at the facial somatosensory/motor region. et al. (2015)
Dental implant Osseoperception may be associated with the brain and the Habre-Hallage et al.
processing of primary and secondary somatosensory areas. (2012)
Orthodontic Functional appliances may work as exercise devices for Ozdiler et al. (2019)
treatment neuromuscular changes associated with muscle adaptation and
brain activation.
Orthodontic In rats, inflammation induced by tooth movement may relate to the Horinuki
treatment activity of the somatosensory cortex and insula, which may be et al. (2015)
associated with higher sensitivity to pain.
Occlusion Occlusal discomfort may be associated with attention and/or Ono et al. (2015)
self-regulation of the uncomfortable somatosensory experience.
Occlusion Regulation of occlusal force and periodontal sensation was Kishimoto
modulated by prefrontal activity. et al. (2019)
Periodontal In rats, mechanical and electrical stimuli may respectively excite Kaneko et al. (2017)
treatment activation at the primary and secondary somatosensory cortices.
Periodontal and Periodontal inflammatory/infectious burden is associated with the Kamer et al. (2015)
systemic health accumulation of amyloid-β plaques, a key feature of Alzheimer’s
disease.
Periodontal and Poor periodontal health may be associated with lacunar infarction, Taguchi et al. (2013)
systemic health a potential cause of dementia.
a
All the search was performed using PubMed, with date of publication ranged from 1 January 2010 to 31 May 2021.
mental functions of research interest (see Some factors other than neural activity, e.g.
Section 2.2). A resting-state fMRI study investi- CBF or vessel volume, may influence the
gates the intrinsic activity of the brain, i.e. the BOLD signal. Perfusion MRI, in contrast,
spontaneous activity when subjects are not assesses the delivery of cerebral blood and pro-
perturbed by external stimuli (see Section 2.4). vides a quantitative measure that can be linked
The fMRI methods can be further categorized to the actual state of blood perfusion by the
according to the nature of the brain signals unit ml/100 g/min for the volume of blood
detected, as discussed below. passing 100 g of tissue within one minute
(Jenkinson and Chappell 2018). The basic con-
1.2.5.1 Blood‑Oxygen-Level-Dependent fMRI cept of perfusion MRI is to label part of the
The blood‑oxygen-level-dependent (BOLD) blood flow and detect the labelled marker after
fMRI detects the changes in the proportion of a fixed time delay. Then, the change of the
deoxygenated and oxygenated haemoglobin in labelled content against time can be quanti-
the brain. This metabolic event (i.e. oxygena- fied. In arterial spinning labelling (ASL), water
tion of haemoglobin) is further associated with molecules are used as an intrinsic marker. In
neural activity. Firstly, the brain region with ASL-MRI, labelling is achieved by altering the
increased neural activity is associated with magnetic properties of the hydrogen nuclei
more energy consumption, i.e. for synaptic (i.e. their spinning behaviour) using different
activity. Secondly, oxygen consumption is asso- radiofrequency. Because changes in CBF can
ciated with increased CBF and changes in cer- be a critical characteristic of neurological dis-
ebral vessel volume, leading to an over-supply orders, perfusion MRI has become an impor-
of the oxygenated vs. deoxygenated haemoglo- tant tool for diagnosing neurodegenerative
bin (see Section 2.2). Finally, deoxygenated disorders, tumours and migraines (Telischak
haemoglobin shows a paramagnetic property et al. 2015).
that disturbs the local magnetic field and
decreases the MR signal (Thulborn et al. 1982).
1.2.6 General Considerations of the
A higher MR signal reflects the effect of an
Limitations of Neuroimaging Methods
increased proportion of oxygenated vs. deoxy-
genated haemoglobin, i.e. the BOLD effect, Though most of the neuroimaging methods
coupled with increased neural activity. In a have been developed for decades, their appli-
task-based fMRI study, researchers can infer cation has some limitations. One of the most
that a mental function is associated with a spe- critical considerations for all imaging methods
cific brain region by identifying changes in the is the spatial and temporal resolution of imag-
BOLD signal in the brain region. Therefore, ing. For both structural and functional neuro-
the discovery of the BOLD effect is essential imaging methods, a poor spatial resolution
for brain mapping, i.e. to map the location of renders it hard to localize the precise position
brain activation associated with functions of a specific brain region. The problem of low
(Jenkinson and Chappell 2018). spatial resolution is significant in PET, which
investigates the brain by a voxel sized between
1.2.5.2 Perfusion MRI – Arterial Spinning 5 and 10 mm3 (Gazzaniga et al. 2019).
Labelling Therefore, it provides spatial information at
A major limitation of the BOLD fMRI (also see the scale of gross anatomy. However, by using
Section 2.1) is that the BOLD signal should be different radioactive neurochemical agents,
interpreted in a relative sense, as the difference PET can detect the part of the brain which spe-
of brain signals between different conditions. cifically engages with the agents. The problem
The value from fMRI data per se cannot be of lower spatial resolution is also significant
directly referred to the actual neural activity. in magnetic resonance spectroscopy (MRS).
Because MRS signals are relatively weak, to ●● The BOLD fMRI detects the changes in the
increase the signals obtained in a voxel, a proportion of deoxygenated and oxygenated
larger voxel will be required for an MRS scan haemoglobin in the brain. This metabolic
(Gazzaniga et al. 2019). Due to the limitation event is indirectly associated with neural
of spatial resolution, neuroimaging methods activity.
provide less information about brain features
at the cellular level.
Temporal resolution is a critical factor for 1.3 How Does Neuroimaging
functional neuroimaging. For functional stud- Contribute to Clinical Practice?
ies, the fundamental question would be ‘do we
have a fine resolution to capture the mental 1.3.1 Introduction
functions we desire to see?’. Some mental pro-
cesses may last for minutes, such as the feeling How can neuroimaging contribute to dental
of a bad mood. In contrast, some mental pro- practice? Intuitively, it is hard to imagine that
cesses may arise transiently, such as shifting the knowledge of ‘brain activation’ would con-
one’s attention from one thing to another. tribute anything for dentists to complete a
Therefore, selecting a tool that is also fast Class II cavity restoration. However, the func-
enough to capture the different mental experi- tional perspective of the brain–stomatognathic
ences is very crucial. MRI is limited at its tem- connection (Figure 1.1) highlights the associa-
poral resolution due to the longer scanning tion between the brain, mental functions and
interval (i.e. a lower sampling rate, such as oral functions. Therefore, understanding the
two seconds for a scan) and the ‘sluggish’ brain is the key to understanding the individ-
hemodynamic response. In contrast to MRI, ual variation in oral functions and feeding and
EEG and MEG are more sensitive to a quick oral healthcare behaviour. In the following sec-
mental process, with a temporal resolution in tions, we elaborate this association by exam-
milliseconds (Gazzaniga et al. 2019). Further ples of dental neuroimaging studies. Firstly,
considerations of the pros and cons of MRI are we discuss the contribution of neuroimaging
outlined in Section 2.1. to oral neuroscience. Secondly, we discuss the
contribution of neuroimaging to the clinical
disciplines of dentistry.
1.2.7 Summary
●● The brain and mental functions, sometimes 1.3.2 Links Between Neuroimaging
metaphorized as a ‘black box’, can hardly be and Key Issues of Oral Neuroscience
examined directly at the chairside. Therefore,
a pivotal step to facilitate the investigation of As a new discipline of neuroscience, how does
the brain is to develop the technology for neuroimaging help investigate these key issues
quantifying brain structure and functions. of oral neuroscience? Table 1.2 shows that
●● Neuroimaging is a non-invasive approach neuroimaging research has been engaged with
that visualizes the CNS, especially the brain. all the key issues in oral neuroscience (Iwata
●● One of the major goals of using structural and Sessle 2019). For example, in terms of the
MRI is to investigate the morphology, includ- pathway of pain processing, neuroimaging
ing the size and shape, of the anatomical research extended our understanding of the
structure of the brain. circuitry from the spinal mechanism to brain
●● The functional MRI methods investigate mechanisms, showing cortical and subcortical
the brain signals associated with brain activation associated with acute and chronic
functions, including BOLD fMRI and per- orofacial pain (Brügger et al. 2012; Gustin
fusion MRI. et al. 2011) and complicated mechanisms of
pain modulation (Desouza et al. 2013; Gustin prosthodontic treatment. For patients with
et al. 2012; Moayedi et al. 2012; Younger tooth loss, replacing the missing teeth with a
et al. 2010). Importantly, because neuroimag- dental prosthesis will restore structural defi-
ing research is conducted in human subjects, cits. Moreover, patients should adapt to the
psychosocial factors, such as emotional and prosthesis and improve chewing function.
attentional factors related to pain, can be Recent neuroimaging findings have shed
investigated (Weissman-Fogel et al. 2011; light on the mechanisms of adaptation of
Abrahamsen et al. 2010; Youssef et al. 2014). dental prostheses (Table 1.3). For example,
Neuroimaging also helps to reveal the brain longitudinal research revealed that adapta-
mechanism of swallowing and chewing tion of a new denture was associated with not
(Lowell et al. 2012; Quintero et al. 2013). only the improvement of masticatory perfor-
Notably, by directly assessing the dental mance but also changes in brain activation
patients who received treatment, translation in the somatosensory cortex (Luraschi
between clinical practice (e.g. installation of et al. 2013). Consistently, tactile stimulation
denture prosthesis) and neuroscience can be on dental implants was associated with brain
achieved (Kimoto et al. 2011; Luraschi activation of the somatosensory areas
et al. 2013). Also, the perceptual processing of (Habre-Hallage et al. 2012). The findings
oral functions, which relies on self-reports suggest that changes in sensory feedback
from human subjects, including tactile and play a key role in improving oral functions by
gustatory senses, can be studied using neuro- prostheses. Moreover, in partially edentulous
imaging (Nakamura et al. 2011; Trulsson patients, reduced occlusion was associated
et al. 2010). As an in vivo imaging method, with reduced activation of the prefrontal cor-
neuroimaging has become the crucial method tex, and such reduced activation can be mod-
for studying the perceptual and psychosocial ulated by installing a denture (Kamiya
aspects of oral functions, which are difficult to et al. 2016). In patients with maxillary dental
approach with animal models. implants, tactile stimuli induced brain acti-
vation not only in the somatosensory cortex
but also in the prefrontal cortex (Habre-
1.3.3 Links Between Neuroimaging
Hallage et al. 2012). Changes in somatosen-
and Clinical Disciplines of Dentistry
sory and prefrontal activation were also
Clinically, neuroimaging research may help identified in another longitudinal fMRI study
dental professions to understand the associa- that dentated patients were chewing a piece
tion between mental functions and dental of gum when wearing a plate to cover their
treatment. For example, dentists may need to palate. The change in brain activation was
know the sensory processing of a patient who associated with the recovery of masticatory
is chewing with an implant-supported den- performance, which was initially impaired
ture. If a dental implant alters the sensory (by the palatal plate) and later restored
feedback from occlusion, one should expect to (Inamochi et al. 2017). These novel findings
identify changes in cortical activation at the suggest that beyond sensory processing, the
sensory area when subjects are chewing. attentional and cognitive processing related
Neuroimaging would be a suitable tool for the to wearing a denture, as evidenced by the
study related to the treatment outcomes of changes in the prefrontal cortex, may play a
dental practice. vital role in the adaptation of prosthodontic
treatment. Thus, the neuroimaging findings
1.3.3.1 Prosthodontics provide new clues for prosthodontic treat-
Restoration of both structural deficits and ment by highlighting the patient’s adaptation
functional impairment is key to successful to dental devices.
perceptive, the brain and the stomatognathic functional element related to oral functions,
system are both essential for maintaining even though the brain is not part of the stoma-
oral health. tognathic system. Both cortical and subcortical
However, the brain–stomatognathic connec- regions are closely associated with oral func-
tion stated above is descriptive and does not tions (Figure 1.2).
fully explain the underlying mechanisms. The
critical and yet unanswered question is ‘how do 1.4.2.2 Definition of the Behavioural Scope
the brain and the stomatognathic system work Another core element to be defined is our
together to maintain our oral health?’. The behaviour that the brain–stomatognathic con-
question is difficult to answer because, before nection relates to. This book focuses on human
the advent of neuroimaging, researchers have eating and feeding behaviour, which is also the
had few tools to directly observe the brain primary target for dental treatment. We define
mechanisms associated with oral functions in feeding behaviour as ‘behavioral responses or
human subjects. In the following sections, we sequences associated with eating including
outline several theoretical frameworks for the modes of feeding, rhythmic patterns of eating
brain–stomatognathic connection. The core and time intervals’ (MeSH 1969). However, we
concepts of the brain–stomatognathic connec- will need to revisit the definitions for several
tion are defined. Subsequently, three theoreti- considerations. Firstly, our primary concern is
cal frameworks on the brain–stomatognathic the feeding behaviour of healthy adults. Eating
connection are discussed. Finally, experimental disorders, such as anorexia and bulimia, are
design to test these theoretical frameworks are beyond the scope of our discussion. Secondly,
discussed. the stomatognathic system is critical to speech
(including the production of phonemes). The
issues related to speech science and language
1.4.2 Core Elements of the Brain–
production are not discussed in this book.
Stomatognathic Connection
Before discussing the brain–stomatognathic
1.4.3 Theoretical Frameworks of the
connection, we need to define the core ele-
Brain–Stomatognathic Connection
ments of the connection. Particularly, we will
clarify the functional element, i.e. the brain There has been literature reporting the brain–
and the stomatognathic system, and human stomatognathic connection for more than
feeding behaviour. 100 years. However, most studies focused on
the pathological mechanisms underlying dis-
1.4.2.1 Definition of the Functional Element eases, such as the infectious routes between
Based on the definition from Medical Subject the brain and the head-and-neck regions. At
Headings (MeSH) of the National Library of present, very few frameworks have been pro-
Medicine, USA, the stomatognathic system is posed for the functional perspective of the
defined as ‘the mouth, teeth, jaws, pharynx brain–stomatognathic connection. By synthe-
and related structures as they relate to mastica- sizing recent clinical and experimental find-
tion, deglutition and speech’ (MeSH 1986). By ings, in this section, we try to provide a ‘big
this definition, either intraoral structure (e.g. picture’ regarding the brain–stomatognathic
teeth and the tongue) or extraoral structure connection (Figure 1.3).
(e.g. the masseter) is part of the stomatog-
nathic system since all the structure contrib- 1.4.3.1 The Oral-to-Behaviour Framework
utes to maintaining normal oral functions. The most intuitive framework consists of the
Notably, from the functional perspective, our stomatognathic system as the only functional
brain should also be considered as part of the apparatus for feeding behaviour. According to
Figure 1.3 Theoretical frameworks of the association between the brain, oral functions and behaviour.
(a) The oral-to-behaviour (OB) framework, (b) the oral-brain-behaviour (OBB) framework and (c) the
brain–stomatognathic axis (BSA).
the oral-to-behaviour (OB) framework, a a surgeon who masters the surgical work in the
sound stomatognathic apparatus directly links oral cavity.
to good eating and feeding behaviour
(Figure 1.3a). Based on this framework, the 1.4.3.2 Challenges from the OB Framework
structural and functional parameters of the However, the OB framework may not always
stomatognathic system determine how well provide a good prediction of the outcome of
one can eat, such that the more teeth and the dental treatment. For example, temporoman-
greater biting force one has, the better mastica- dibular disorders (TMD) are associated with
tion and swallowing one will achieve. Notably, various deficits of teeth, the temporomandib-
the OB framework consists of a bi-directional ular joint and muscles. According to the OB
relationship: while improving the stomatog- framework, a primary step of treating TMD
nathic system will help better eating, a poor would be to fix the structural deficits, such as
eating experience will motivate individuals to adjusting patients’ occlusion by reshaping
fix the deficits of the stomatognathic system. cusp morphology. However, cumulating evi-
Empirically, the simple logic that ‘if you have dence suggested that the relationship between
got something wrong with eating, fix your occlusal adjustment and the improvement of
teeth first’ goes well most of the time. It echoes patients’ symptoms is controversial (Xie
the traditional view that dentists are trained as et al. 2013). In contrast, more evidence from
the sensorimotor control of limb movement and the integrity of sensorimotor control of
has revealed that human action is maintained oral functions both contribute to good eating
by the corresponding motor program, shaped ability. However, the framework simplifies the
by learning and adaptation via a complex association between the brain and oral health.
mechanism of the brain (Wolpert and In addition to sensorimotor control (which has
Flanagan 2016). Notably, the brain would also been widely investigated via animal research),
play a major role in the stomatognathic func- mastication and swallowing are also associated
tions since most of these functions related to with cognitive, affective and motivational pro-
feeding – either mastication or swallowing or cessing of the brain (Figure 1.3c). For example,
pain, are highly associated with the integra- as shown in Chapter 5, the tactile (e.g. ‘chewy’)
tion between sensory feedback and motor and gustatory (e.g. ‘yummy’) experience from
commands, both mandated by the brain. chewing is associated with an increased
Therefore, to strengthen the original OB hedonic value and reward processing of food.
framework, the link between the stomatog- Therefore, the brain–stomatognathic axis
nathic system and behaviour needs to be (BSA) framework highlights multiple associa-
revisited. tions between brain functions and feeding
behaviour. Most importantly, the BSA frame-
1.4.3.3 The Oral-Brain-Behaviour Framework work highlights that all the functions partici-
According to the oral-brain-behaviour (OBB) pate in the adaptation of oral conditions. When
framework, to maintain good eating ability, dysfunction occurs (either due to structural
one needs (i) to restore structural deficits of deficits, aging or brain impairment), individu-
the oral cavity/teeth and (ii) to maintain the als also learn how to adapt to this new condi-
sensorimotor control of oral functions. tion. For example, when having a meal, the
Traditionally, point (ii) is relatively ignored patients with a new denture may keep on
in dental treatment because dentists assume detecting if the denture is well fitted and judg-
that sensorimotor processing works well. ing if the food bolus is good to swallow. All the
However, as shown in Chapter 7, the sensori- cases suggest that feeding behaviour is not a
motor processing of the brain may alter as simple translation of oral sensory and motor
age increases or in patients with brain functions. It is crucially associated with the
impairment (e.g. neurodegeneration or attentional, cognitive, motivational and emo-
stroke). Therefore, oral dysfunctions and dif- tional processing related to eating.
ficulty in feeding may be associated with In the BSA framework, the term ‘axis’
deficits in brain functions (Figure 1.3b). emphasizes a bi-directional and dynamic rela-
According to the OBB framework, for elderly tionship between the brain and the stomatog-
or special needs patients, both fixing struc- nathic system (Lin 2018). In gastroenterology,
tural deficits and maintaining brain func- the concept of ‘gut-brain axis’ (GBA) has been
tions are critical to improving patients’ proposed and widely distributed for many
oral health. years. The GBA consists of ‘bidirectional com-
munication between the central and the
enteric nervous system, linking emotional and
1.4.4 The Brain–Stomatognathic Axis
cognitive centres of the brain with peripheral
While the OBB model emphasizes that the intestinal functions’ (Carabotti et al. 2015). In
brain is critical to the stomatognathic func- parallel, the BSA emphasizes that the brain
tions, it does not directly account for the plays a more comprehensive role in sensori-
individual differences in feeding behaviour. motor and affective–cognitive processing on
The OBB framework suggests that a good the stomatognathic functions and feeding
stomatognathic condition (e.g. fully dentated) behaviour.
1.4.5 How Can Neuroimaging Research apparatus is impaired (e.g. losing teeth), indi-
Help Studying the Brain– viduals can maintain feeding behaviour. A criti-
Stomatognathic Connection? cal step to test this general hypothesis is to focus
on the individuals who show structural deficits
One of the greatest advantages of neuroimag-
(e.g. with a higher number of missing teeth) but
ing is to explore the brain mechanisms associ-
maintain a good oral function. In this group, the
ated with feeding behaviour directly on
degree of sensorimotor adaptation and learning
human subjects. For example, researchers can
is assessed using specific tasks (see Chapter 8).
measure the signals related to brain activities
The association between task performance and
associated with chewing and swallowing with
brain activation, particularly in the regions
different imaging approaches. The following
associated with cognitive, affective and motiva-
sections focus on research design for investi-
tional processing, is assessed using neuroimag-
gating the brain–stomatognathic connection.
ing. Notably, the superiority of neuroimaging is
1.4.5.1 Investigation of the OBB Framework that it can assess brain activation associated
The OBB framework highlights sensorimotor with complicated processing of learning and
control of oral functions. The brain mecha- adaptation in human subjects – which may be
nisms associated with oral functions can be challenging to perform on animal subjects.
investigated using a task-based study, in which
brain activities are recorded concurrently when 1.4.6 Summary
an individual is performing an oral function.
The association between sensory or motor pro- ●● The OB framework consists of the stomatog-
cessing and oral functions can be modulated by nathic system as the only functional appara-
different experimental conditions. For exam- tus for eating and feeding behaviour.
ple, when chewing harder food, one should According to the OB framework, a sound
expect stronger sensory feedback from the peri- stomatognathic apparatus means good feed-
odontal tissue than soft food. In this case, ing behaviour.
changes in brain activation, as shown by func- ●● The OBB framework highlights the role of
tional MRI, would be identified at the soma- the exchange of sensorimotor information
tosensory cortex, and this activity is known to between the brain and the stomatognathic
reflect the intensity of sensory inputs (Onozuka system in oral functions.
et al. 2002; Takahashi et al. 2007). It is notewor- ●● In the BSA framework, the brain is not just a
thy that the interpretation of an association passive translator for the sensorimotor infor-
between task parameters (e.g. the hardness of mation but also a ‘moderator’ that actively
food) and brain features (e.g. activation of the engaged with one’s feeding behaviour.
motor cortex) may be complicated. For exam- ●● Neuroimaging research on the BSA frame-
ple, when chewing a harder piece of food, one work focuses on identifying individual differ-
may pay more attention to the texture of the ences in oral functions. The BSA emphasizes
food. Therefore, the brain regions with activa- that the brain plays a more comprehensive
tion may be associated with attention and cog- role in sensorimotor and affective–cognitive
nitive control as well as sensory processing processing on the stomatognathic functions
(Onozuka et al. 2002, Takahashi et al. 2007). and feeding behaviour.
References
Abrahamsen, R., Dietz, M., Lodahl, S. et al. (2010). Gustin, S.M., Peck, C.C., Cheney, L.B. et al.
Effect of hypnotic pain modulation on brain (2012). Pain and plasticity: is chronic pain
activity in patients with temporomandibular always associated with somatosensory cortex
disorder pain. Pain 151: 825–833. activity and reorganization? J. Neurosci. 32:
Anderson, T. (1790). Pathological observations 14874–14884.
on the brain. Lond. Med. J. 11: 182–190. Habre-Hallage, P., Dricot, L., Jacobs, R. et al.
Avivi-Arber, L., Lee, J., Sood, C. et al. (2015). (2012). Brain plasticity and cortical correlates
Long-term neuroplasticity of the face primary of osseoperception revealed by punctate
motor cortex and adjacent somatosensory mechanical stimulation of osseointegrated
cortex induced by tooth loss can be reversed oral implants during fMRI. Eur. J. Oral
following dental implant replacement in rats. Implantol. 5: 175–190.
J. Comp. Neurol. 523: 2372–2389. Hayes, G.B. (1889). Reflex neurosis in relation
Avivi-Arber and Sessle (2018). Jaw sensorimotor to dental pathology. Am. J. Dent. Sci. 23:
control in healthy adults and effects of ageing. 289–298.
J. Oral Rehabil. 45: 50–80. Horinuki, E., Shinoda, M., Shimizu, N. et al.
Bandettini, P.A. (2012). Twenty years of (2015). Orthodontic force facilitates cortical
functional MRI: the science and the stories. responses to periodontal stimulation. J. Dent.
NeuroImage 62: 575–588. Res. 94: 1158–1166.
Brügger, M., Lutz, K., Brönnimann, B. et al. Inamochi, Y., Fueki, K., Usui, N. et al. (2017).
(2012). Tracing toothache intensity in the Adaptive change in chewing-related brain
brain. J. Dent. Res. 91: 156–160. activity while wearing a palatal plate: an
Carabotti, M., Scirocco, A., Maselli, M.A., and functional magnetic resonance imaging study.
Severi, C. (2015). The gut-brain axis: J. Oral Rehabil. 44: 770–778.
interactions between enteric microbiota, Iwata, K. and Sessle, B.J. (2019). The evolution of
central and enteric nervous systems. Ann. neuroscience as a research field relevant to
Gastroenterol. 28: 203–209. dentistry. J. Dent. Res. 98: 1407–1417.
Desouza, D.D., Moayedi, M., Chen, D.Q. et al. Jenkinson, M. and Chappell, M. (2018).
(2013). Sensorimotor and pain modulation Introduction to Neuoimaging Analysis. Oxford
brain abnormalities in trigeminal neuralgia: a University Press.
paroxysmal, sensory-triggered neuropathic Jones, D.K., Knösche, T.R., Turner, R. (2013).
pain. PLoS One 8: e66340. White matter integrity, fiber count, and other
Gazzaniga, M.S., Ivry, R.B., and Mangun, fallacies: the do’s and don’ts of diffusion
G.R. (2019). Cognitive Neuroscience: The MRI. Neuroimage 73: 239–254.
Biology of the Mind. W. W. Norton & Kamer, A.R., Pirraglia, E., Tsui, W. et al. (2015).
Company. Periodontal disease associates with higher
Gould, D.J., Clarkson, M.J., Hutchins, B., and brain amyloid load in normal elderly.
Lambert, H.W. (2014). How neuroscience is Neurobiol. Aging 36: 627–633.
taught to north American dental students: Kamiya, K., Narita, N., and Iwaki, S. (2016).
results of the basic science survey series. Improved prefrontal activity and chewing
J. Dent. Educ. 78: 437–444. performance as function of wearing denture
Gustin, S.M., Peck, C.C., Wilcox, S.L. et al. in partially edentulous elderly individuals:
(2011). Different pain, different brain: functional near-infrared spectroscopy study.
thalamic anatomy in neuropathic and PLoS One 11: e0158070.
non-neuropathic chronic pain syndromes. Kaneko, M., Horinuki, E., Shimizu, N., and
J. Neurosci. 31: 5956–5964. Kobayashi, M. (2017). Physiological profiles of
Dental Neuroimaging: The Role of the Brain in Oral Functions, First Edition. Chia-Shu Lin.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
Companion website: www.wiley.com/go/lin/dental-neuroimaging
Figure 2.1 The general concept of the blood-oxygen-level-dependent (BOLD) mechanism. (a) Transportation
of oxygenated haemoglobin during a resting condition, when neural activity is low. (b) Transportation of
oxygenated haemoglobin when neural activity increases. The neurons demand more energy by consuming
oxygen provided by oxygenated haemoglobin. Via a complex haemodynamic process (e.g. an increasing rate
and volume of cerebral flow), the amount of oxygenated haemoglobin increases (relatively to the amount of
deoxygenated haemoglobin), leading to an over-supply or compensation of the oxygen demand from neurons.
Section 2.1.4). In fMRI, brain mapping of men- research design. Another crucial advantage
tal functions is indirectly inferred from the of MRI is its approachability. The installation
local change of the BOLD signal, which reflects of MRI scanners has been popularizing in
the metabolic changes related to neural activity. many countries, summarized by the report
from the Organisation for Economic Co-
operation and Development (https://data.
2.1.3 Practical Advantages of the
oecd.org/healtheqt/magnetic-r esonance-
MRI Approach
imaging-mri-units.htm). It is not uncommon
In the preceding sections, we have outlined the to find an MRI unit in a major hospital in the
advantage of MRI from a technical perspective. urban area or the university. In sum, the versa-
Here, we highlight two more ‘practical’ reasons tility that an MRI scanner can perform differ-
for the popularity of MRI: the versatility and ent imaging methods and its approachability
the approachability of MRI. The versatility of in medical and research centres render MRI
MRI can be exemplified by the fact that one a practically good option for neuroimaging,
can use a single MRI scanner to perform dif- offering great flexibility for research design
ferent imaging methods for different pur- (Lin and Yeung 2020).
poses. For example, by setting different
parameters of imaging acquisition, one can
2.1.4 Methodological Considerations
perform the magnetic resonance spectro-
of the MRI Approach
scope (MRS) to assess the MR behaviour of a
specific neurotransmitter or the diffusion In the following section, we delve into several
magnetic resonance imaging (dMRI) to methodological issues of the MRI approach.
assess the orientation of diffusion of water These issues should be carefully evaluated for
molecules, which reflects the architecture of the research design and the interpretation of
the fibrous compartment in the brain. Also, neuroimaging findings.
one can magnetically label the blood and
quantify its flow rate in the brain using arte- 2.1.4.1 Temporal and Spatial
rial spinning labelling (ASL). In other words, Resolution of MRI
the same MRI scanner can provide different The temporal and spatial resolution of imaging
information about our brain. This versatility methods plays a key role in the clinical applica-
has greatly increased the flexibility in tion of neuroimaging. Simply speaking, a higher
temporal resolution means a greater number of metabolic change related to the oxygen level,
‘time frames’ to be acquired within a fixed which is indirectly associated with neural
period. The higher the temporal resolution is, activity (Figure 2.1). Nevertheless, the associa-
the shorter period the time frame is. Practically, tion between metabolic and neural changes is
a method with a higher temporal resolution can much complicated. The explanation of the
be used to ‘capture’ an event that occurs in a very BOLD effect is based on the concept that an
short period. A higher spatial resolution means increased or decreased CBF/CMRO2 is cou-
a greater number of voxels – i.e. the spatial units pled with neural activity (see Box 2.1). To con-
that compose a 3D image, an analogue to ‘pixels’ firm this association, researchers have used
for a 2D image – to be acquired within a fixed microelectrodes to investigate the electric field
volume (see Box 2.2 in the Companion Website). of the brain using animal models. The findings
The higher the spatial resolution is, the smaller reveal that the BOLD signals are not directly
the actual size of a voxel is. Practically, an imag- driven by the neural activity, such as the spik-
ing method with a higher spatial resolution can ing signals derived from the neuron. Findings
be used to identify the signal within a smaller from animal research reveal that the BOLD
area. The improvement of resolution is associ- signals were much associated with the low-
ated with a decreased signal-to-noise ratio (SNR) frequency local field potentials (LFPs) of the
for the signals acquired (Huettel et al. 2004). brain rather than the neuronal spiking rate.
Theoretically, one can increase the spatial reso- The LFP may be associated with the excitatory
lution of MRI into the scale of the sub- and inhibitory post-synaptic potentials (EPSPs/
millimetre. However, this would be done at the IPSPs) of multiple neurons (Ekstrom 2010).
cost of a decreased SNR, which means a greater Therefore, it is considered a peri-synaptic neu-
possibility to obtain more noises within a voxel. ral activity. Though the neuronal spiking cova-
The association between spatial resolution and ries with the LFP, reducing the spiking per se
the SNR further relates to the magnetic field of did not significantly affect the BOLD response
MRI. By increasing the strength of the magnetic (Logothetis and Pfeuffer 2004).
field, the resolution can be improved while However, the ‘LFP-coupling hypothesis’
keeping the SNR acceptable. This is one of the (Logothetis and Pfeuffer 2004) may be compli-
reasons an MRI scanner with a higher magnetic cated when one considers the variety of cellu-
field (e.g. 7 Tesla) has been gradually adopted in lar architectures of different brain regions.
neuroimaging research (Turner 2016). However, For example, the hippocampus consists of a
increasing the strength of the field may not be complex pattern of recurrent and excitatory/
an elixir to everything. High-field MRI may con- inhibitory circuitry, and therefore, the cou-
front technical challenges, such as the inhomo- pling mechanisms between the BOLD signals
geneity of the magnetic field, and practical and the neural activity in the hippocampus
challenges, such as the safety and comfort of may differ from the other brain regions
subjects (Karamat et al. 2016). In general, the (Ekstrom 2010). Moreover, from the point of
issues of temporal and spatial resolution of energy consumption, a majority of expendi-
imaging are associated with many physical fac- ture brain energy is to maintain the post-
tors, such as the change of SNR and magnetic synaptic metabolism of transmitters, such as
field. To balance between all these factors is a glutamine (Attwell and Iadecola 2002).
complicated and challenging task. Therefore, it is also possible that the BOLD
effect represents the haemodynamic responses
2.1.4.2 The Neurophysiological Nature of the associated with the signalling of neurotrans-
BOLD Signal mitters rather than the direct consumption of
As noted previously, fMRI utilizes the princi- energy (Attwell and Iadecola 2002). Due to the
ple that MR signals are sensitive to the complexity of the neurophysiological nature of
the BOLD signal, any inference of the micro- or time frames collected sequentially rather
scopic neural mechanisms drawn from the than a continuous stream of information. In
macroscopic neuroimaging evidence should be fMRI, a lower temporal resolution (or a larger
treated carefully. period of a time frame to be acquired) of
around two seconds would be enough for
capturing the long-lasting experience, such
2.1.5 The Limitations of Neuroimaging
as pain and emotion. The neuroimaging tools
Research Based on MRI
with higher temporal resolution, such as
In the following sections, we summarize sev- EEG and MEG, provide a higher temporal
eral limitations derived from the challenges sensitivity to detect the dynamic change of
stated above. Notably, some limitations are transient mental functions, such as atten-
common to all neuroimaging methods. These tional and perceptual processing, which may
limitations should be carefully evaluated occur in milli-seconds.
when neuroimaging data are interpreted or
translated for clinical applications. Limitations 2.1.5.3 Limitations of Detecting
associated with research design and data anal- the Neural Events
ysis are outlined in the other sections of the As shown in the preceding section, one should
chapter. be careful about making an inference of under-
lying neural events (e.g. the strength of neu-
2.1.5.1 Limitations in Spatial Resolution ronal spiking) of mental functions directly
For a 3-Tesla MRI scanner, one can acquire from the MRI results (e.g. the change of BOLD
structural and functional images with a spatial signals). For example, in a task-based fMRI
resolution at 0.75–1 and 2 mm, respectively. In study (see Section 2.2), the ‘activation’ or ‘deac-
other words, the actual size of a voxel in a struc- tivation’ of a brain region cannot be directly
tural brain image would be 1 mm × 1 mm × 1 mm translated as excitation or inhibition at the syn-
(i.e. 1 mm3). The resolution is generally better aptic level, and the direction of modulation,
than that of PET, magnetoencephalography e.g. a top-down or a bottom-up modulation,
(MEG), and electroencephalography (EEG). should not be asserted solely by neuroimaging
However, it is still at the scale of gross anatomy evidence (Logothetis 2008). Furthermore, it
rather than the microscopic scale. Therefore, should be noted that most of the task-related
one should be careful about relating the results BOLD signals do not reflect the quantitative
from most sMRI and fMRI studies to histologi- change in neural activity. For example, the
cal and cellular levels. Moreover, in fMRI, the BOLD signals may be associated with the
signals of an individual voxel are usually aver- regional variation in CBF and the morphology
aged by spatial smoothing, which may influ- of blood vessels, and the regionally specific
ence the ability of fMRI to identify a regional architecture of neural circuitry may influence
activation, particularly when the information is the resulting BOLD signals. In other words,
sparsely coded (Turner 2016). Though there the data from non-invasive neuroimaging
have been plenty of findings of the histological alone is not sufficient for fully elucidating the
and cellular architecture of the brain, these neural mechanisms of brain functions
findings may not be fully linked to the MRI (Logothetis 2008).
findings due to the coarse resolution of the lat-
ter (Turner 2016).
2.1.6 Summary
2.1.5.2 Limitations in Temporal Resolution ●● In fMRI, the brain mapping of mental func-
The ‘image’ acquired using MRI or PET tions is indirectly inferred from the local
should be taken as a set of discrete volumes change of the BOLD signal, which reflects
the metabolic change related to neural 2.2.2 A Case Study: Brain Mechanisms
activity. of Mastication
●● The issues of temporal and spatial resolu-
One of the most widely studied oral functions
tion of imaging are associated with many
in neuroimaging research is mastication. From
physical factors, such as the change of SNR
the scope of task-based fMRI, one aims to study
and magnetic field. To balance between all
the brain activation associated with mastica-
these factors is a complicated and challeng-
tion, which is based on a chewing task.
ing task.
Therefore, the critical element of a task-based
●● The association between the BOLD signal
study is the design of the task condition, which
and the change in neural activity has not
is closely associated with the mental functions
been fully elucidated. Therefore, any infer-
to be engaged. Different ‘chewing tasks’ varied
ence of the microscopic neural mechanisms
between the timing of chewing movement and
drawn from the macroscopic neuroimaging
the materials for chewing (Figure 2.2). In addi-
evidence should be treated carefully.
tion to the task condition, the ‘baseline condi-
●● The activation or deactivation of a brain
tion’ also needs to be defined so that the desired
region cannot be directly translated as exci-
mental functions can be contrasted between
tation or inhibition at the synaptic level or
the two conditions. The contrasted result is fur-
the direction of modulation. The data from
ther analyzed for each individual subject, and
non-invasive neuroimaging alone is not suf-
the results from all subjects are synthesized at
ficient for fully elucidating the neural mech-
the group level by taking individual differences
anisms of brain functions.
(e.g. age or clinical factors) into consideration
(Figure 2.2) so that brain activation associated
with mastication can be identified.
2.2 Research of Task-based
Functional Activation 2.2.2.1 What Is the ‘Task’ for Task-based
Functional Neuroimaging?
2.2.1 Introduction
It is always crucial to elucidate the details of the
The section gives an overview of the task- task design because a slight difference between
based functional neuroimaging research. the two tasks may result in different mental
Functional neuroimaging is the approach to functions, and thus very different patterns of
identify the brain signals associated with brain activation. For example, if a task requires
mental functions. Here, the term ‘task-based’ subjects to chew a piece of gum for five minutes
highlights the experimental condition from continuously, the subjects may gradually feel
which mental functions derive. For example, bored and uncomfortable, and it becomes possi-
fMRI can be performed to investigate brain ble that the acquired BOLD signals reflect not
activation associated with pain and visual only chewing movement but also the change in
processing, based on a pain-stimulation and subjects’ emotional status. Therefore, in most of
a visual judgment task, respectively. In the the studies, a period of chewing is often followed
following sections, we first clarify the associ- by a resting period so that subjects may not
ation between a task and the baseline condi- develop fatigue or distress. However, researchers
tion. Secondly, we highlight the importance need to further remind subjects to resume chew-
of the assessment of inter-subject variability. ing after the resting period, usually by an audi-
Finally, the key points about how to interpret tory or a visual cue. The use of sensory cues may
the statistical findings of a task-based study introduce an additional auditory or visual expe-
are outlined. rience, which is further associated with brain
activation related to auditory or visual process- Such a baseline brain activity of sustained vis-
ing. The examples here highlight the importance ual attention is not specifically associated with
of a clear definition of a task condition in neuro- chewing. Therefore, a condition for control-
imaging research. The key issue for interpreting ling the baseline activity is usually necessary.
the neuroimaging results is to clarify what sub- As shown in Figure 2.2, most of the chewing
jects actually do during a task and what mental tasks were contrasted to a control task in
functions are engaged within the task. which subjects were instructed to rest (i.e. not
chewing). The contrast between chewing vs.
2.2.2.2 What Is the Baseline Condition Used not chewing may reflect different conditions
for Control? in jaw movement. However, the two condi-
It is intuitive to think that the BOLD signals tions also differ in the degree of sensory pro-
collected when a subject is performing some cessing because subjects may not have any
task must be task-related. This is partially cor- teeth contacted during the baseline (resting)
rect because not all the signals acquired dur- condition. Therefore, brain activation may be
ing the task condition are task-specific. In fact, associated with both sensory and motor pro-
most of the signals can be irrelevant to the cessing during chewing. If subjects are asked
task because there are some undergoing men- to occlude their teeth during the baseline con-
tal activities that are not specifically linked to dition, the contrast may be more specifically
the task. For example, during a chewing task, related to motor control because the sensory
subjects may continuously pay attention to a factor is better balanced between the task and
visual cue that guides the task procedure. the baseline condition. In other words, the
definition of the baseline condition plays a key are not compatible with an MRI scanner, in
role in interpreting neuroimaging results which the use of ferromagnetic materials is
because it reflects the factors being controlled restricted for safety reasons. In general, these
when the mental function of interest is practical issues should be carefully considered
contrasted. when experimental tasks are designed.
Figure 2.3 Methodological considerations of a functional magnetic resonance imaging study. (a) Subjects
may show great inter-individual variability in their general conditions, such as sex, age and general
physical/psychological conditions. (b) Subjects may show great inter-individual variability in their personal
trait and performance (e.g. pain ratings) related to an experimental task. (c) Subjects differ in brain
morphology. When individual brains are compared, the individual images are normalized to a template
image, using linear transformation (i.e. translation, rotation, resizing and shearing) and nonlinear
transformation approaches.
T1-weighted structural images from healthy within a group. It is commonly used for inves-
young adults (Evans et al. 2012). It is notewor- tigating the association between different
thy that this template brain does not represent groups (e.g. patients vs. controls). Finally, a
a brain of a real person. Instead, the template variety of methods of statistical thresholding
reflects an ‘on-average brain’ estimated by sta- will be adopted for controlling Type I and Type
tistical methods. The step to register each indi- II errors of imaging results.
vidual brain image to the template is spatial
normalization. After normalization, the indi- 2.2.4.1 Statistical Analysis at
vidual brain images are positioned in the same the Individual Level
spatial format (i.e. the format of the template The individual-level analysis aims to investi-
image), and then they can be compared. gate the association between the BOLD signals
Methodologically, brain images can be spa- and experiment conditions. Exactly speaking,
tially transformed based on linear and nonlin- the analysis focuses on the association between
ear approaches. In linear transformation, two the time-varying BOLD signals (i.e. time series)
images are rotated and translated in the 3D and the progression of a task. As shown in
space so that they can be overlapped. The origi- Figure 2.4a, an fMRI study of chewing may
nal images are translated in the x-, y- and z- consist of four chewing blocks interleaved
axes and rotated over the x-, y- and z-axes (i.e. with four resting blocks. If the BOLD signal of
the rigid transformation) (Figure 2.3c). a voxel is highly associated with the task, one
Additionally, the image will be resized and should expect to see a strong association
sheared to match the template. In nonlinear between the BOLD time series of the voxel and
transformation, the images will be locally the task progression. The association can be
warped to match with the template (Jenkinson quantified as the correlation between these
and Chappell 2018). Notably, the nonlinear two. In other words, the purpose of the first-
warping will lead to the deformation of the level analysis is to examine how the BOLD
original image at the voxel level. For each voxel, time series nicely fits the predicted BOLD
local geometric changes will be recorded. response, which is derived from the task tim-
Technical issues regarding the algorithms of ing (Jenkinson and Chappell 2018). To achieve
imaging analysis are beyond the scope of the this goal, the general linear model (GLM) is
book. Nevertheless, it is critical to recognize widely used to estimate the coefficient of the
that these procedures are crucial for analyzing fitting process. A greater coefficient of the
neuroimaging data (see Box 2.3 in the GLM suggests a greater effect size of the fitting
Companion Website). process, i.e. a stronger association between the
BOLD time series of the voxel and the task (e.g.
Voxel A in Figure 2.4a, compared to Voxel B).
2.2.4 Statistical Analysis of the
For a whole-brain analysis, the GLM analyses
Task-based Research
will be performed for all the voxels within the
The statistical analysis of task-based fMRI data whole brain for individual subjects. Therefore,
consists of multiple sequential steps. The steps by mapping the coefficients at the whole-brain
can be encapsulated into three ‘major pipe- scale, one can identify the voxels or clusters of
lines’ (Figure 2.4). The analysis of individual voxels that are associated with the task.
data (i.e. the first-level analysis) aims to inves-
tigate the association between the BOLD sig- 2.2.4.2 Statistical Analysis at the Group Level
nals and experiment conditions by each voxel, The analysis at the individual level identifies
separately for each subject. The analysis of the brain activation associated with the
group-based data (i.e. the second/higher-level experimental conditions for each subject.
analysis) aims to combine the individual data The results reflect the association between
Figure 2.4 Statistical analysis at the individual level and the group level. (a) The analysis at the individual
level focuses on the association between task progression and the blood‑oxygen-level-dependent (BOLD)
time series, as shown in the left panel. For each voxel, a strong association indicates that the BOLD signals
of the voxel can be predicted by the task condition, in contrast to the baseline condition (e.g. Voxel A), as
shown in the right panel. (b) The analysis at the group level focuses on the association between brain
features (e.g. brain activation of grey matter volume) and group factors. The association may reflect the
difference in brain features between patient and control groups (the left panel) or the correlation between
brain features and clinical factors (the right panel). (c) A typical image result consists of the statistical
values (e.g. the t-score) from multiple voxels (represented as the grid), which are visualized by a colour
scale, as shown in the left panel. The result can be thresholded according to intensity (i.e. the t-score). For
example, only the voxels with a t-score >6 are preserved after thresholding, as shown in the middle panel.
The result can be thresholded according to the size of a cluster of voxels. For example, only the clusters
with a size larger than 100 voxels will be preserved after thresholding, as shown in the right panel.
mental functions (which is contrasted by a individual results within a group. The group-
task and a baseline condition) and the BOLD based analysis provides a great advantage for
time series of a voxel. The second-level anal- investigating clinical questions. For exam-
ysis, in contrast, aims to combine the ple, does the pattern of brain activation
associated with chewing (derived from the map that shows the difference of chewing-
individual-level analysis) differ between associated activation between groups (in
TMD patients and healthy controls? To which voxel intensity denotes a t-score for
answer this question, the Student’s t-test can between-group comparison). The brain map
be performed by comparing the brain activa- may also present the strength of association
tion between a group of TMD patients and a between masticatory performance and
group of controls (Figure 2.4b). Notably, the chewing-associated activation (in which voxel
‘brain activation’ derived from the individual- intensity denotes a regression coefficient).
level analysis consists of the results from all However, as for all research, we need to evalu-
the voxels in the brain. Therefore, in the case ate the Type I and Type II errors related to our
of a between-group comparison, the t-tests results, which are associated with the risk of
will be performed for each individual voxel. finding false-positive and false-negative
The result of the t-tests (e.g. the t-score) will results, respectively. For example, should we
be assigned to each voxel. Therefore, the believe that a voxel with a very low t-score
group-level analysis generates a new brain really reflects a genuine between-group dif-
image, in which each voxel value represents ference? A practical way to control for these
the difference of brain activation (quantified errors is to set a threshold for the imaging
by the t-score) between two groups. results. For example, a threshold of
The same approach can be used to investi- t-score = 2 means that all the voxels with a
gate the brain activation associated with indi- t-score > 2 will be considered ‘significant’ or
vidual differences. For example, do people ‘important’ (Figure 2.4c). In other words, we
with a better chewing function have higher may set up a threshold to the intensity for
activation in the motor cortex? To answer this each voxel, and only the voxels ‘survive’ under
question, regression analyses can be per- such a threshold will be taken for further
formed by investigating the association interpretation (Figure 2.4c).
between brain activation and task perfor- In addition to the approach of intensity-
mance (e.g. chewing performance) or indi- based thresholding, the approach of cluster-
vidual factors (e.g. number of missing teeth). based thresholding aims to evaluate if the size
The association can be quantified by the beta of a cluster of voxels is large enough to be
coefficient of the regression model. The called significant or important. This is based
group-level analysis can be performed at the on the assumption that, biologically, neurons
whole-brain scale (e.g. for all the voxels work coherently for a mental function. For
within the brain) and specifically for a region example, it may not be very convincing if
of interest (ROI). In the latter (i.e. an ROI- painful stimulation is associated with the acti-
based analysis), brain activation is analyzed vation of just one single voxel in the sensory
specifically for the voxels within a pre- cortex (even though the ‘intensity’ of the acti-
defined ROI, which is usually selected based vation is very high). It would be doubtful that
on a priori knowledge of brain functions such activation may be a ‘false positive’ one
(Poldrack 2007). due to signal noise. In contrast, a cluster (i.e.
spatially contiguous activation of voxels) with
2.2.4.3 Statistical Thresholding a larger size may be a more plausible result.
for Imaging Results The larger cluster suggests that all the voxels
The results from the group-level analysis may in the brain region would participate in pain
already provide some useful information for processing (Figure 2.4c). Both thresholding
research and clinical purposes. In the case of approaches are widely used in neuroimaging
chewing research, we might obtain a brain research.
2.2.5 Further Considerations Regarding contrasting approach is valid given that the
Experimental Design and Analysis assumption of pure insertion is true (Amaro
Jr and Barker 2006). Under pure insertion, we
In the next sections, we focus on the design of
assume that ‘pain processing’ is the only men-
an fMRI experiment, which is critical for inter-
tal function that differs between ‘pain’ and
preting neuroimaging results.
‘rest’ conditions. However, it is possible that
2.2.5.1 Considerations About the ‘Subtraction’ something other than ‘pain’ is included in the
Design and Pure Insertion contrast (Figure 2.5a). For example, subjects
The task-based design is based on the assump- may pay greater attention during the pain-
tion that mental functions can be contrasted stimulating condition but not the resting
by comparing between a task condition and a condition. In this case, the observed brain
baseline condition. For example, the brain activation may represent not only pain pro-
activation related to pain can be identified by cessing but also increased attentional process-
contrasting the condition of ‘painful stimula- ing. The validity of the subtraction design
tion’ to the condition ‘resting’, which is usu- would greatly influence how we interpret the
ally expressed as a form of subtraction (e.g. behavioural meaning of observed brain
‘pain > resting’ or ‘pain – resting’). Such a activation.
Figure 2.5 Experimental design of functional neuroimaging research. (a) Under the assumption of pure
insertion, the difference of brain activation between two experimental conditions only reflects the mental
function contrasted by the conditions (e.g. perception of pain intensity). However, the contrast may be
associated with more than one mental function (e.g. perception of pain intensity and attention to noxious
stimuli). (b) A factorial design helps to delineate the association between two mental functions. For
example, the light grey area denotes the effect of increased pain on brain activation, and the dark grey area
denotes the effect of increased attention. (c) A conjunction design focuses on the pattern of brain activation
common to two experimental conditions (e.g. a clenching task and a chewing task). The activation may
reflect the brain mechanisms of a mental function common to both task conditions.
has been identified in the dentate gyrus, and Neuroimaging data are composed of multiple
this neurogenesis can be associated with angi- voxels, which represent the basic unit for spa-
ogenesis (Pereira et al. 2007). However, tial measurement. Therefore, the analysis of
changes in grey matter may also be associated the volumetric information of brain regions
with synaptogenesis and more dendritic can be performed on the voxel level. In the fol-
branching at the synapse of a neuron. In addi- lowing sections, we introduce VBM, i.e. a
tion, even with the same number of neurons, voxel-based approach to quantify morphologi-
changes in grey matter can be found when cal features of the brain.
there is substantial sprouting of the neural
axon, which reflects more connection between 2.3.3.1 What Does VBM Means?
a neural cell with the others. Furthermore, the In terms of neuroimaging, ‘morphometry’ can
glial cells, including oligodendrocyte progeni- be considered an approach that measures the
tor cells (OPCs) and astrocytes, can divide morphological metrics of the brain, including
during adulthood. Changes in the brain fea- the size, shape and position of a part of the
tures of grey matter may reflect an increased brain or the whole brain (Ashburner and
number of the glial cells or morphological Friston 2000). Conventionally, this can be car-
changes of the cells (Zatorre et al. 2012). ried out by selecting an anatomical structure
of research interest (e.g. the thalamus) and
2.3.2.2 Biological Significance of Changes having its border delineated by an expert man-
in White Matter ually. The size, shape and position of the
There are several indices for quantifying selected structure can be assessed and com-
changes in white matter, including white mat- pared between groups. When it comes to VBM,
ter volume and fractional anisotropy (FA), an the measurement focuses on the voxel level
index mainly assessed using diffusion tensor rather than the anatomical structure as a
imaging (DTI). An increased white matter vol- whole (Ashburner and Friston 2000;
ume may be associated with changes in the Ashburner and Friston 2001). The brain
glial cells and angiogenesis (Zatorre images from two individuals are ‘matched’ at
et al. 2012). In contrast, FA may reflect the the voxel level, so the difference can be
degree of the fibrous architecture, which has a revealed. Such a process of ‘matching’ is
profound effect on limiting the diffusion of achieved by registering the individual brain
water molecules. Therefore, an increased FA images to a template image. It should be noted
(i.e. the molecules move more ‘directionally’) that when two brains are matched at the voxel
can be found in a region with more fibrous level, some morphological differences, such as
structure. the shape of individual brains, cannot be
revealed because the shapes of individual
brains have already been registered to the
2.3.3 Voxel-based Morphometry
standard image, via multiple steps of transfor-
One of the benefits of sMRI is to quantify the mation (Figure 2.3c). The VBM focuses on the
size of structural features. Changes in the size intensity of each voxel, which may disclose
of a brain region may provide an important important clues about the types of brain tis-
clue for the clinical diagnosis, such as atrophy sues, while discounting the individual differ-
of the hippocampus in patients with ence in the global shape of the brain
Alzheimer’s disease. However, it is not easy to (Ashburner and Friston 2000, Ashburner and
compare the size of a brain region between two Friston 2001). For simplicity, VBM can be rec-
individuals because individual differences in ognized as ‘a voxel-based comparison of the
size are always associated with other morpho- local concentration of grey matter between
logical features, such as shape and position. two groups of subjects’ (Ashburner and
Friston 2000). In other words, it is a method template image. Conventionally, VBM analy-
sensitive to the local composition of brain tis- sis uses a customized template, i.e. an ‘aver-
sues, ‘while discounting positional and other aged brain’ estimated by the brain images
large-scale volumetric differences in gross from all the subjects included in the analysis.
anatomy’ (Ashburner and Friston 2001). Individual images are registered to the tem-
plate via linear and nonlinear transforma-
2.3.3.2 Steps of VBM Analysis tion (see Section 2.2). Finally, the intensity
Practically, VBM is frequently performed on of images is usually ‘modulated’ for estimat-
the T1-weighted MRI data, which shows a ing the volumetric information. During nor-
good contrast between WM (with higher/ malization, the local geometric features have
brighter intensity), cerebrospinal fluid (CSF), been ‘morphed’ so that some areas may be
with lower/darker intensity), and GM (with contracted and some areas can be expanded.
moderate intensity). Conventionally, the VBM Notably, after the local deformation, the
approach consists of the following steps: (i) tis- intensity value of each voxel will need fur-
sue classification, (ii) spatial normalization ther adjustment to compensate for the
and (iii) intensity modulation. change of volumetric expansion/contrac-
The first key step is tissue classification. tion. This adjustment is performed by ‘mod-
The original T1-weighted image will be ulating’ the voxel intensity, which integrates
matched to tissue probability images. These the voxel intensity (based on tissue segmen-
images contain the information regarding tation) and the information of local deforma-
the probabilistic information that a given tis- tion (Jenkinson and Chappell 2018).
sue type (WM, GM or CSF) is found in the
voxel (Ashburner and Friston 2005). The 2.3.3.3 Interpretation of the VBM Results
original T1-weighted image will be seg- As noted in the preceding section, in the GM
mented according to the probabilistic image derived from segmentation, the voxel
images. The segmentation generates three value should be realized in a probability
separate images that contain the GM, WM sense, i.e. how possible this voxel can be clas-
and CSF area, respectively (i.e. GM/WM/ sified as part of grey matter. Therefore, this
CSF-segmentation images). Notably, in these value does not quantify the ‘actual size’ of
segmented images, the area of a tissue type is grey matter tissue. The VBM results consider
not defined by a binary value, which repre- both the results of tissue classification (e.g.
sents the clearly-cut border of the segmented the amount of GM) and the volume change
part of the tissue. Instead, the voxel intensity (i.e. spatial normalization). Therefore, the
of the segmented images ranges from 0 to 1, results can be considered ‘the proportion of
which represents the proportion of a tissue voxel that is grey matter, having adjusted for
type within a voxel. Therefore, a voxel may the confounding effects of warping the
have an intensity of 0.8 in the GM- brains’, or from the perspective of volumetric
segmentation image and 0.2 in the WM- change, as ‘the proportion of volume change
segmentation image, suggesting it more attributable to grey matter’ (Ashburner and
likely to be classified as GM compared to Friston 2001). Some interpretations of VBM
WM. The second key step is spatial normali- results can be over-simplified and mislead-
zation. Back to the original concept of mor- ing. For example, when a VBM analysis
phometry, if one needs to compare at the revealed a smaller grey matter volume of the
voxel level, an essential step is to match the hippocampus in patients with dementia,
individual brain images with each other. In compared to healthy controls, the results
VBM, this is achieved by spatial normaliza- should not be interpreted as a smaller hip-
tion, which registers the individual brain to a pocampal size because it ignores the fact that
the volume change is adjusted from specific surface features, including sulcal depth (as an
tissue type (here, for the grey matter). index for the boundary between brain lobes)
and curvature (i.e. how curved or flat the part
of the surface is), in addition to cortical thick-
2.3.4 Surface-based Approach
ness. Thirdly, the analysis may also help to
The VBM approach analyzes the amount of a match images between individual subjects.
specific tissue type based on the concept that For example, across subjects, the structural
the brain can be dissected into multiple voxels, images can be normalized to a template by
and all the analyses are performed at the level matching the position of sulci, which may
of brain voxels. The surface-based approach, provide more accurate registration of brain
instead, is based on the concept that the brain images between different individuals.
is composed of layers of tissues. The impor-
tance of the layer structure can easily be recog- 2.3.4.2 Segmentation of Brain Regions
nized from the gross anatomy of the brain As noted earlier, segmenting different types of
since the brain cortex is folded into multiple brain tissues is a critical step for analyzing the
gyri and sulci. In the following sections, we structural features of the brain. This tissue seg-
introduce the basic concepts of the surface- mentation can be performed decently by
based approach for assessing brain VBM. However, segmenting different brain
morphometry. structures, i.e. to label a voxel as part of the
hippocampus or the amygdala, can be a real
2.3.4.1 Assessment of Cortical Thickness challenge because this classification cannot be
The cortex is mainly composed of grey matter, carried out merely by assessing the intensity of
which forms continuous layers folded into a voxel. To achieve this, researchers have
multiple gyral and sulcal regions. One of the adopted positional information from the brain,
primary morphological features of the layer using a similar concept as surface-based analy-
structure is its thickness. In the surface-based sis. In addition to assessing the information of
approach, such as the method based on voxel intensity, the relative spatial location of
FreeSurfer (Fischl 2012), the thickness of the brain structures is considered in the algorithm
cortical layer (i.e. cortical thickness) is of segmentation (Fischl 2012). For example, a
defined as the distance between the grey/ voxel will show a greater probability of being
white and pial surfaces. The surface-based classified as the amygdala than the hippocam-
methods have been proved to be a useful way pus if its spatial location is in front of a voxel
of studying cortical thickness. For example, classified as part of the hippocampus (based
Fischl and Dale (2000) have investigated the on the a priori knowledge that the amygdala is
thickness of the cerebral cortex from 30 anatomically anterior to the hippocampus).
younger adults, showing an average ± stand- The surface-based approach, such as the
ard deviation thickness of 2.7 ± 0.3 mm in the method based on FreeSurfer, has been widely
gyral area and 2.2 ± 0.3 mm in the sulcal area. used for the automatic segmentation of corti-
In general, the cortical thickness is within the cal and subcortical regions and the estimation
range of 1– 4.5 mm, consistent with the find- of their volumes.
ings from post-mortem research (Fischl and
Dale 2000). The thickness-based analysis has 2.3.4.3 Steps of Surface-based Approach
provided useful information for our under- The general step of the surface-based analysis
standing of the brain. Firstly, the analysis is similar to VBM analysis, in which segmenta-
considers the realistic shape of the brain cor- tion and spatial normalization play a key role.
tex, which is full of folding areas. Secondly, An obvious difference between VBM and the
the analysis is also used for assessing other surface-based approach is the method for
image registration. In VBM, the individual method for assessing the features of neural
brain images are registered to a template image, fibres (Jones et al. 2013). dMRI assesses the
which is positioned in a 3D space composed of physical properties of the diffusion of water
orthogonal axes in three directions. In the molecules in brain tissues. During MRI,
surface-based approach, the registration is per- increased signals are obtained with all the mol-
formed in a ‘surface space’, which can be visu- ecules which show a similar tendency in spin-
alized as a inflated spheric area that represents ning (i.e. phasing). However, the signals from
the whole surface area of the cortex. The ‘sur- phasing may be attenuated with the molecules
face space’ is different from the ‘voxel space’ of spin in different ways (i.e. dephasing). The
VBM analyses. In VBM, brain voxels are organ- dephasing effect is more prominent in a
ized in a fixed grid with three dimensions. In compartment where the molecules can freely
contrast, the surface space is organized as a diffuse in all directions, compared to a com-
mesh composed of multiple small triangles partment that is restricted in diffusion.
(i.e. faces) with intersected points (i.e. verti- Therefore, dMRI assesses the degree of attenu-
ces). Notably, while the number of voxels is ation of signals, which reflects the diffusion of
fixed in VBM (based on its spatial resolution), water molecules, and diffusion is associated
the number of vertices of the mesh is not fixed with the presence of axon bundles in a com-
and needs to be estimated to fit the individual partment. In this compartment, water mole-
brain (Jenkinson and Chappell 2018). The dif- cules can preferentially diffuse along the
ferent ways of organization of brain images direction where axons are oriented (Jenkinson
have a profound influence on how data are and Chappell 2018; Jones et al. 2013).
analyzed. Therefore, dMRI has been widely used for
Before registration, the original brain image assessing the local microstructural properties
needs to be transformed into this surface area of molecular diffusion and the estimation of
by mapping the x-, y- and z-coordinates from anatomical connectivity (Jenkinson and
the ‘brain space’ to the surface space. This is Chappell 2018).
performed by an algorithm that deforms the
folded brain (with a lot of sulci) into an inflated 2.3.5.1 Assessment of Microstructure
sphere, much like blowing air to a crumpled of White Matter
plastic bag to inflate it like a balloon. By this dMRI assesses the features of molecular diffu-
deformation, the locational information of the sion from multiple directions and the attenua-
sulci and gyrus can be transformed into a tion of magnetic resonance signal associated
smoother surface for analysis (Jenkinson and with the molecules spinning. Therefore, when
Chappell 2018). The template to be registered signal attenuation is more pronounced in one
is also represented in the surface space, and direction than directions, it signifies that in
therefore, within the same surface space, indi- this particular direction, molecules can diffuse
vidual brain images are normalized to the tem- more freely compared to the other directions.
plate. Notably, because the 3D gyral and sulcal In DTI, axial diffusivity (AD) is the diffusivity
regions are ‘flattened’ into a surface form, assessed along the long axis. Diffusivity is also
many morphological features, including the assessed along the direction perpendicular to
surface area of cortical regions, can be ana- the axial direction, i.e. the radial diffusivity
lyzed via this approach. (RD). The mean diffusivity (MD) is the average
of diffusion over all directions. The FA repre-
sents the degree that the estimated pattern of
2.3.5 Diffusion MRI
diffusion deviates from an isotropic form. The
dMRI is able to ‘map the fibre architecture of higher the FA value is, the more likely that
tissue’ and, to the present, the only in vivo diffusion is restricted in a single direction
(i.e. anisotropic) rather than spread evenly in approach focuses on tissue-specific features on
all directions. It is noteworthy that the local the voxel level. Even when the spatial resolu-
properties only reflect the molecular behav- tion is very high, one cannot guarantee that
iour of diffusion, they cannot be overdrawn as only one type of brain tissue exists in a voxel.
an index of white matter integrity (which The partial volume (PV) effect, i.e. the signals
should be preserved in a pathological condi- of a voxel is a ‘weighted average’ of the signal
tion), and the changes in the local properties from different types of brain tissues, should be
cannot be directly translated to the changes in considered. Such a weight reflects the fraction
structural connectivity (Jones et al. 2013). of the volume associated with each tissue
(Jenkinson and Chappell 2018). Likewise,
2.3.5.2 Tractography: Estimation dMRI does not provide a direct assessment of
of Anatomical Connectivity the size and density of axons or the degree of
In addition, dMRI is also a powerful tool for myelination in a voxel (Jenkinson and
researchers to explore the global architecture of Chappell 2018). The PV effect should also be
the brain by answering questions such as ‘which considered for dMRI because a voxel size is
gray matter regions are inter-connected by relatively large, and multiple fibres of different
white matter fibers?’ (Jones et al. 2013). To directions may be included in a voxel. To clar-
investigate the structural connectivity between ify the interaction between multiple fibres
brain regions, one would adopt the approach within a voxel is still a major challenge in
tractography, an analysis ‘to determine anatom- dMRI (Jenkinson and Chappell 2018).
ical connectivity by tracing the paths of axonal VBM is a method based on the intensity of
fiber bundles through the white matter’ the T1-weighted image. The spatial distribu-
(Jenkinson and Chappell 2018). It should be tion of the intensity values is affected not only
noted that tractography is the method primarily by the tissue types but also by other imaging
for qualitatively visualizing the connection factors such as relaxation time (which is fur-
between grey matter regions based on the local ther associated with tissue myelination) and
properties of molecular diffusion. One needs to field inhomogeneity (Zatorre et al. 2012). The
be careful when using the results of tractogra- same consideration applies to dMRI. The
phy for quantitative use, such as to identify the dMRI signal is associated with the microscale
‘number of fibres existing’ between brain architecture that hinders diffusion. However,
regions. As shown in Section 2.4, the link diffusion per se can be influenced by other fac-
between the streamlines estimated by tractogra- tors, such as the mobility of molecules, which
phy and the real ‘axonal fibres’ should not be further depends on the viscosity and tempera-
simplified because the streamlines are esti- ture factor (Jones et al. 2013).
mated based on the properties of local diffusion, Finally, a general rule of the analysis of
and the estimation per se is influenced by many sMRI data is that the results from the current
factors, such as the density of fibres in a voxel analysis are based on those from earlier analy-
and the length and shapes of fibres, which can- ses. That means the effect of any methodologi-
not be easily resolved by the dMRI data (Jones cal variations (or even an error) from an earlier
et al. 2013). stage of analyses will contribute to later analy-
ses. For example, the methods of tissue seg-
mentation at the beginning of VBM and
2.3.6 Further Considerations of
surface-based analyses will affect the final
Structural MRI
results of the between-group comparison.
It is noteworthy that, again, spatial resolution Therefore, the methodological details of imag-
plays a dominant role in interpreting all the ing processing need to be documented and
structural results. For example, the VBM explicitly stated (Ridgway et al. 2008).
(i.e. focusing on the clusters or ‘blobs’ of brain quantifying structural and functional connec-
activation related to a task) to ‘connectology’ tivity are introduced in the following sections.
(Smith 2012). In the following sections, we
provide a brief introduction to the methods of
2.4.3 Functional Connectivity
investigating brain connectivity.
Functional connectivity can be broadly
defined as ‘the temporal correlations between
2.4.2 Segregation and Integration
spatially remote neurophysiological events’
of Brain Functions
(Friston 1994). Based on the definition, two
As a network where mental experience emerges, aspects are critical for the assessment of func-
the brain is governed by some basic principles, tional connectivity. Firstly, the temporal cor-
including functional segregation and functional relations derive from the association between
integration (Tononi et al. 1994; Friston 2011; functional signals, such as the BOLD time
Sporns 2013). The concept of functional segre- series acquired via fMRI. Therefore, to calcu-
gation refers to ‘neuronal processing carried out late the correlation, one needs to first define
among functionally related regions arranged the duration of the functional signals. For
within modules’ (Sporns 2013). This concept example, in resting-state fMRI research, the
highlights that some brain units may work correlation is calculated based on the BOLD
closely as a community or a function-specific time series during the whole resting period.
module for achieving specific mental functions Secondly, the functional signals derived from
(Figure 2.6b). In contrast, the concept of func- two spatially very close regions (e.g. two
tional integration highlights how different parts neighbouring voxels) will have a strong corre-
of the brain interact with each other for achiev- lation because they are likely to be the same
ing mental functions. In other words, in func- anatomical structure. Therefore, researchers
tional integration, one is mainly interested in are interested in the correlation of signals
‘the efficiency of global communication’ and from geometrically remote brain regions. The
‘the ability of the network to integrate distrib- analysis of functional connectivity helps to
uted information’ (Sporns 2013) (Figure 2.6c). elucidate the association between the remote
Notably, the two concepts are not exclusive to brain regions of a known neural circuitry.
each other. The brain is balanced between the
status of highly segregated (i.e. each part of the 2.4.3.1 Intrinsic Brain Functions
brain works completely independently) and As noted earlier, functional connectivity can be
the status of highly integrated (i.e. each part of used to quantify the functional signals collected
the brain works completely dependently). The during a resting state, in which subjects are not
coexistence of segregation and integration is engaged in a task. Why is it important to explore
crucial to the complexity of the brain network the connectivity during a resting rather than a
(Tononi et al. 1994). At the core of our under- task condition? To explain this, one needs to
standing of functional segregation and integra- consider the intrinsic brain functions, which
tion is the concept of ‘brain connectivity’, which are less related to the sensory and motor func-
can be generally considered as the pattern of tions induced by a specific task. In other words,
links between brain regions. It may refer to the some brain functions are associated with the
structural connectivity (i.e. ‘the existence of ‘baseline status’ (i.e. the non-task-specific
white matter tracts physically interconnecting conditions) of our mind. These functions
brain regions’) (Uddin 2013) or the functional include information processing for ‘interpret-
connectivity (i.e. ‘the temporal correlations ing, responding to and predicting environmen-
between spatially remote neurophysiological tal demands’ (Raichle 2010). Experimentally,
events’) (Friston 1994). The methods for this baseline status is set by asking subjects to
relax and rest during fMRI scanning. The Therefore, the activity is usually assessed in a rest-
importance of the baseline status can be dem- ing state when subjects are not actively engaged
onstrated by the fact that the brain, with only with a specific sensory or motor task. The fluctua-
2% of our body weight, consumes around 20% tion of this BOLD activity may represent the
of the total energy during resting (Raichle 2010). stimulus-independent mental activity, such as
Notably, even when the brain is actively self-referential processing or memory, which are
engaged with a mental operation (e.g. perform- considered part of intrinsic brain function (Fox
ing a language task), the additional energy con- and Raichle 2007; Raichle 2010). It may also repre-
sumption is often less than 5% (Raichle 2010). sent a preparatory status for coming tasks by cou-
The findings of the high baseline consumption pling the brain regions that usually work together
of energy suggest that even when we are not for a specific function (Fox and Raichle 2007,
engaged with a mental function overtly, the Raichle 2010).
brain always maintains intrinsic and ongoing
activities. 2.4.3.3 Research Approaches
of Functional Connectivity
2.4.3.2 Spontaneous BOLD Activity and the Cumulating evidence has revealed that even
Resting-state fMRI during the resting state when individuals are
In terms of the connectivity of intrinsic brain not actively engaged with external stimuli,
functions, researchers focus on the temporal cor- the human brain still works actively, as dem-
relation of the spontaneous BOLD activity, a low- onstrated by the connectivity of the sponta-
frequency signal (0.01–0.1 Hz) collected during a neous BOLD activity. The study design of
resting state. The spontaneous BOLD activity can intrinsic functional connectivity is illustrated
be defined as ‘the activity that is not attributable to in Figure 2.7. Firstly, the spontaneous BOLD
specific inputs or outputs’ (Fox and Raichle 2007). activity is acquired during a resting state,
which relates to the intrinsic brain func- pattern of brain regions that form strong func-
tions of a non-task-specific ‘baseline’ status tional connectivity with the region of our inter-
(Figure 2.7a). The brain is segmented into mul- est (i.e. the seed).
tiple brain regions according to a brain atlas
(Figure 2.7b) (see Box 2.4 in the Companion
2.4.4 Structural Connectivity
Website). For an individual brain region,
BOLD time series is extracted from each of the Taking telecommunication as an analogue,
voxels within the brain region, and a mean functional connectivity of the human brain
time series of a region is calculated by averag- can be exemplified as the ‘communication’
ing the BOLD time series of all the voxels between brain regions. More communication
(Figure 2.7c). Several additional steps are is represented by a higher coherence of BOLD
included here, such as extracting the low- time series. In contrast, the study of structural
frequency component of the time series. Next, connectivity focuses on how lines are wired. In
based on the definition of functional connec- the human brain, these ‘lines’ reside in white
tivity, the strength of association between the matter, which is predominantly composed of
brain regions is quantified by correlation neural fibres. Therefore, the primary approach
coefficients (Figure 2.7d). The correlation to study structural connectivity is to investigate
coefficient represents the strength of the con- the spatial distribution of white matter of the
nection between each pair of brain regions brain, which can be assessed by dMRI, as dis-
(Figure 2.7e). For example, a stronger correla- cussed in Section 2.3. Another less straightfor-
tion is identified between region 1 and region ward approach is to estimate the covariance of
2, compared to the correlation between region structural features to provide a picture of the
2 and region 4. The approach is helpful to elu- strength of association of structural features
cidate a hypothesis based on a priori knowl- between different brain regions.
edge of neural circuitry. For example, one can
compare the strength of functional connectiv- 2.4.4.1 Structural Connectivity Based
ity between two plausible pathways (e.g. on Tractography
between region 1 and region 2 or between dMRI provides information regarding the
region 2 and region 4). One can also use the local properties of molecular diffusion at
same approach for investigating individual dif- the voxel level (see Section 2.3). Based on the
ferences in behaviour, such as quantifying the dMRI data, tractography can be used to reveal
association between perceived pain and the the streamlines connecting brain regions.
connectivity of pain-related regions. Structural connectivity is quantified accord-
Another widely used approach is the seed- ing to the ‘streamlines’ estimated using the
based approach. The approach explores the algorithm of deterministic tractography or
brain regions that are highly correlated with a probabilistic tractography. In deterministic
‘seed’ region at the voxel level. For the seed- tractography, each voxel is assigned with a
based approach, the BOLD time series from a single direction (Figure 2.8a), and a continu-
pre-defined brain region (i.e. the ‘seed region’) ous curve (i.e., a streamline) is formed by
is collected, and correlation is calculated tracking the direction of voxels (Jenkinson
between the time series of the seed region and and Chappell 2018). In probabilistic tractog-
the time series from all the voxels outside the raphy, there exists an uncertainty of the
seed region (Figure 2.7f). This step will gener- direction within a voxel. Therefore, the direc-
ate a functional connectivity map, which tion of a voxel may vary, forming a probabil-
reveals the area with stronger connectivity ity distribution. As shown by a simplified
with the seed region. The seed-based approach case in Figure 2.8b, the distribution presents
is especially useful to explore the global a high probability for the leftward direction.
Figure 2.8 Analysis of structural connectivity. (a) In deterministic tractography, each voxel is assigned with
a single direction, which reflects the principal direction of diffusivity. A continuous streamline is formed by
tracking the direction of voxels. (b) Probabilistic tractography assumes that there exists an uncertainty of
the direction within each voxel. In the right panel, the probabilistic distribution of the directions is
estimated for each voxel. A higher probability of a ‘leftward’ direction can be identified. In the right panel,
the intensity of a voxel represents the frequency that a streamline passes that voxel. For example, more
streamlines pass the yellow voxel (here four out of seven streamlines) compared to the red voxel (here one
out of seven streamlines). (c) Structural covariance quantifies the strength of association of a brain feature
between different brain regions across subjects. For example, the cortical thickness of six brain regions is
assessed for eight subjects. The right panel reveals the association between regions 2 and 6, as quantified
by the correlation coefficient of the cortical thickness between the regions.
cortical thickness from brain regions. The network of the brain can be visualized as a
degree of covariation can be quantified by the graph composed of nodes (i.e. brain regions)
correlation coefficient between the regional and linked by edges (i.e. the connectivity
cortical thickness across all the subjects between regions). By analyzing the topologi-
(Figure 2.8c). The biological mechanisms cal features of the graph, researchers can
underlying structural covariance have not identify the structural or functional archi-
been fully elucidated. The pattern of inter- tecture of the whole brain.
regional association revealed by structural
covariance shows a substantial degree of 2.4.5.1 The Definition of a Node
overlap with the pattern from intrinsic func- In the graph-based network analysis, the defi-
tional connectivity (Kelly et al. 2012b) and the nition of nodes is the most fundamental issue.
structural connectivity based on diffusion A node can be conceived as the elemental unit
connections (Gong et al. 2012). The brain to define a brain region. The links between
regions covary in structural features may brain regions (e.g. the inter-regional correla-
share coordinated rates of developmental tion of the BOLD time series) can be quanti-
process (Alexander-Bloch et al. 2013). fied only when we have defined what a brain
region is. For example, the node can be a brain
structure with a clear anatomical border, such
2.4.5 Analysis of the Brain Connectivity
as the thalamus and the amygdala, and func-
Using a Network-based Approach
tional connectivity is quantified as the correla-
In the previous sections, we outlined several tion between average BOLD time series from
widely used methods to quantify functional all the voxels within each region. It can even
and structural connectivity. In general, the be a single voxel, and functional connectivity
approach is very useful if we want to ana- will be quantified between all the pair-wise
lyze the connection between two specific BOLD time series from all the voxels. The defi-
regions, which have already been known as nition of nodes is sometimes associated with
critical elements in a neural pathway. These the ‘resolution’ of the brain network. A net-
methods help us to quantify the strength of work may consist of around 100 nodes, each
the connectivity. However, when it comes to defined as a structurally or a functionally dis-
the global pattern of connections of brain tinct region. In contrast, a network may con-
regions, one needs to assess the property of sist of more than 10 000 nodes when a node is
a network, which consists of multiple con- defined at the voxel level. To elucidate how
nections. From the biological perspective, nodes are defined is essential to interpret the
mental functions are associated with the meaning of a network properly (Butts 2009).
myriad numbers of interactions between For example, not all the ‘brain regions’ are
multiple brain regions. If we aim to study structurally or functionally well defined as the
how efficiently different parts of the brain thalamus or the amygdala. The border of some
communicate with each other, the key is to regions, such as the dorsolateral prefrontal
understand how the individual connections cortex, may show great inter-subject variabil-
assemble into a whole network. The term ity. For structural connectivity, one may define
‘network’ refers to a real-world complex sys- the region based on some anatomical land-
tem, defined by a collection of nodes and marks (e.g. gyri and sulci). However, the same
links (Rubinov and Sporns 2010). definition may not reflect the same brain func-
Mathematically, a network can be consid- tions across subjects. For example, it may be
ered as a graph consisting of ‘vertices’ (i.e. invalid to claim that the same area of the dor-
nodes) and ‘edges’ (i.e. links) (Figure 2.9a). solateral prefrontal cortex is ‘the node for
Therefore, the structural or functional attention control’ to all subjects. Secondly,
Figure 2.9 Graph-based analysis of brain connectivity. (a) The pattern of functional and structural
connections between brain regions can be translated from the ‘brain space’ to the ‘network space’ with
applications of graph theory. In a graph, the nodes represent brain regions and the links represent the
functional and structural connectivity between regions, which can be quantified by the correlation coefficient
between blood-oxygen-level-dependent (BOLD) time series and the streamlines identified by tractography,
respectively. (b) In the network analysis, the global metrics quantify the degree of integration of a network.
For example, characteristic path length can be calculated by finding the shortest path length between a pair
of nodes, such as the path A–B–D (but not A–B–E–D) between the nodes A and D. (c) The local metrics
quantify the degree of segregation of a network. For example, the clustering coefficient is used to quantify
the fraction of the triangular architecture in the whole network (e.g. A–B–C and E–G–H), which represents a
pattern of clustered nodes. Notably, a small-world network offers a balance between the efficiency of global
and local communication. A highly regular network (i.e. the middle-right panel) and a highly random network
(i.e. the middle-left panel) may suffer from a lower global and local efficiency, respectively.
back to the problem of the resolution of a net- different sub-nuclei (e.g. the ventromedial or
work, we need to consider how a structure lateral nuclei) are associated with different
should be defined as a functionally distinct roles in sensorimotor processing. Should we
unit. For example, it seems reasonable to include the whole thalamus as a node? Or
define the thalamus and the amygdala as dif- should each of the sub-nuclei be investigated
ferent nodes because they are associated with as an individual node? This will lead to differ-
sensorimotor and cognitive–affective process- ent biological interpretations of a brain
ing, respectively. However, in the thalamus, network.
degree of between-group connectivity different groups. Thirdly, one issue that has
(Rubinov and Sporns 2010). Just like the small- been debated for years is the use of global sig-
world architecture, being modular is also a key nal regression, an analytical step to remove the
feature of the functional or the structural net- global mean BOLD signal, which may be unre-
work of the brain (Bassett and Bullmore 2017). lated to neural activity (Murphy et al. 2013).
However, the adoption of this step is associated
with an increased proportion of negative con-
2.4.7 Methodological Considerations
nectivity in the functional network (i.e. more
of Brain Connectivity Research
links with anti-correlation between the BOLD
As one of the neuroimaging fields with the time series), which may greatly influence how
fastest advancement, research on brain con- we interpret the results (Murphy and Fox 2017).
nectivity has been revisited for many methodo-
logical issues. As outlined in the following 2.4.7.2 Misinterpretation of Findings
sections, these methodological considerations of Structural Connectivity
are important to interpretations of the biologi- A common myth of interpreting the local prop-
cal significance of brain networks. erties of molecular diffusion is that ‘if there is
a greater FA in a voxel, there are more neural
2.4.7.1 Confounding Factors of Analyzing fibres’. Such a direct translation between the
Functional Connectivity imaging findings at the voxel level and the bio-
Several factors may confound the findings of logical features of neural fibres is not valid.
intrinsic functional connectivity. Firstly, the Even for the findings of tractography, we need
functional signals acquired during a resting to keep in mind that some biological features,
state are associated with not only our baseline such as the size, shape and length of neural
mental functions but also individual physio- fibres, cannot be directly assessed from trac-
logical processes (Kelly et al. 2012a; Murphy tography (Jones et al. 2013). Therefore, one
et al. 2013). It is estimated that 5–15% of the should be careful of generalizing the results of
variance in intrinsic BOLD activity is accounted structural connectivity into clinical conditions.
for by cardiac and respiratory processes (Kelly For example, a claim that ‘individuals show
et al. 2012a). Therefore, individual differences loss of white matter integrity’ merely based on
in these physiological processes, which are a change of FA or MD value should be revisited
especially pronounced in patients with sys- cautiously (Jones et al. 2013). Similarly, the
temic diseases, should be carefully evaluated ‘tracts’ demonstrated by tractography merely
when we interpret the results of functional reflect the pattern of streamlines passing
connectivity. Secondly, head motion has a sub- through different brain regions. The neuroana-
stantial impact on the results of functional tomical features of neural fibres need to be
connectivity analyses (Van Dijk et al. 2012). confirmed by other methods, such as tract-
The effect of head motion may vary in different tracing methods based on neurochemi-
brain regions. Increased head motion is associ- cal agents.
ated with stronger intrinsic functional connec-
tivity of motor network but weaker connectivity 2.4.7.3 Methodological Variations
of the default mode and frontoparietal net- of Graph-based Analyses
works (Van Dijk et al. 2012). Critically, the As shown previously, graph-based network
degree of head motion may differ between dif- analysis is a powerful tool for quantifying some
ferent groups of individuals, such as younger topological metrics, which help us to under-
and older subjects. Therefore, the age factor stand the pattern of integration/segregation of
should be carefully considered when one com- brain networks. However, we should keep in
pares the pattern of connectivity between mind that all the metrics are calculated based
Box 2.5 From Research to Practice – There Is Something More Important Than an ‘Image’
for Neuroimaging
What is the most important part of neuroim- great challenge in conducting and interpret-
aging research on oral functions? It is often ing neuroimaging research is how to prop-
ignored that the functional or structural fea- erly execute the functional task, including
tures acquired from neuroimaging need to assessment and stimulation protocols. For
be interpreted according to the correspond- example, there are different definitions and
ing clinical/behavioural variables – other- methods for assessing masticatory perfor-
wise, the brain features per se may not mance. Therefore, to interpret a result such
disclose too much information. For example, as the one reported by Lin et al. (2015), one
Lin et al. (2015) investigated the grey matter needs to clarify what masticatory perfor-
volume of 25 older subjects and found a mance really is and relate its functional sig-
positive correlation between regional grey nificance to the results of brain imaging.
matter volume and masticatory performance. Similar to the study by Henderson et al.
In an fMRI study of patients with neuropathic (2013), one needs to clarify what the stimu-
pain, Henderson et al. (2013) identified corti- lation protocol really does – for example, dif-
cal activity at the somatosensory area by ferent stimuli (e.g. brushing or pinpricking)
brushing patients’ lower lip. In these studies, at different regions (e.g. a keratinized or a
‘grey matter volume’ and ‘somatosensory non-keratinized area) may evoke different
activity’ are the variables quantified by neu- sensory-affective experience. In sum, neuro-
roimaging. However, their meanings need to imaging research does not merely consist of
be interpreted in accordance with the clini- brain images. It is critical to clarify how the
cal variables, i.e. masticatory performance clinical variables are assessed and functional
and the brushing paradigm, respectively. A tasks are performed.
and individual variations in behavioural/ functional and the structural connectivity does
clinical aspects. not mean that the presence of strong structural
connectivity is a necessary condition for the pres-
2.4.8.2 Are Functional and Structural ence of strong functional connectivity. For exam-
Connectivity Well Coupled? ple, functional connectivity can be identified in
In this chapter, we discuss structural and func- some distant brain regions where no direct struc-
tional connectivity in separate sections due to dif- tural connectivity (e.g. white matter tract) is
ferent approaches for analysis. This does not found. In contrast, even for a well-established
imply that the two types of connectivity are irrel- neural pathway with strong structural connectiv-
evant. On the contrary, recent studies have ity, there can exist a great variation in functional
revealed that some features of the structural net- connectivity between individuals.
work can predict the features of the function net-
work (Honey et al. 2009). The association
2.4.9 Summary
between the two may vary based on the anatomi-
cal distance between brain regions. In general, ●● The brain is considered a complex system in
for the brain regions close to each other, their which mental states are associated with the
functional connectivity and structural connectiv- interaction between its functional and struc-
ity are correlated with a higher degree, compared tural units, rather than an assembly of inde-
to those distant from each other (Honey pendent anatomical regions, each specialized
et al. 2009). Notably, the correlation between the for a single mental function.
●● Even during the resting state when individu- ●● From the perspective of a network, mental
als are not disturbed by external stimuli, the functions are associated with the myriad
human brain still works actively, as demon- numbers of interactions between multiple
strated by the inter-regional connectivity of brain regions. If one aims to ask the question
spontaneous BOLD activity. ‘how efficiently do different parts of the
●● The seed-based approach is widely used for brain communicate with each other’, one
investigating functional connectivity. To needs to view how the individual connec-
identify where in the brain has higher func- tions assemble into a whole network.
tional connectivity with the ‘seed’ region ●● Methodological issues including (i) the con-
(that researcher has put a special interest in), founding factors of analyzing functional
one checks for the brain regions which connectivity, (ii) the misinterpretation of
BOLD time series is highly correlated with findings of structural connectivity and (iii)
the time series from the seed region. methodological variations of graph-based
●● Using dMRI, one can estimate the probabil- analyses, should be considered when one
istic distribution of fibre orientation in a spa- interprets brain connectivity findings.
tial compartment by detecting water
diffusion. Furthermore, probabilistic trac-
tography can be used to estimate the poten- Further Readings
tial trajectories between brain regions, which
may be used for indexing the strength (in a Please see the Companion Website for
relative sense) of connection between differ- Suggested Readings and Box 2.2–2.4 (From
ent brain regions. Tools to Discovery).
References
Alexander-Bloch, A., Giedd, J.N., and Bullmore, Bassett, D.S. and Gazzaniga, M.S. (2011).
E. (2013). Imaging structural co-variance Understanding complexity in the human
between human brain regions. Nat. Rev. brain. Trends Cogn. Sci. 15: 200–209.
Neurosci. 14: 322–336. Butts, C.T. (2009). Revisiting the
Amaro, E. Jr. and Barker, G.J. (2006). Study foundations of network analysis. Science
design in fMRI: basic principles. Brain Cogn. 325: 414–416.
60: 220–232. Ekstrom, A. (2010). How and when the fMRI
Ashburner, J. and Friston, K.J. (2000). Voxel- BOLD signal relates to underlying neural
based morphometry--the methods. activity: the danger in dissociation. Brain Res.
NeuroImage 11: 805–821. Rev. 62: 233–244.
Ashburner, J. and Friston, K.J. (2001). Why Evans, A.C., Janke, A.L., Collins, D.L., and
voxel-based morphometry should be used. Baillet, S. (2012). Brain templates and atlases.
NeuroImage 14: 1238–1243. NeuroImage 62: 911–922.
Ashburner, J. and Friston, K.J. (2005). Unified Fairhurst, M., Wiech, K., Dunckley, P., and
segmentation. NeuroImage 26: 839–851. Tracey, I. (2007). Anticipatory brainstem
Attwell, D. and Iadecola, C. (2002). The neural activity predicts neural processing of pain in
basis of functional brain imaging signals. humans. Pain 128: 101–110.
Trends Neurosci. 25: 621–625. Fischl, B. (2012). FreeSurfer. NeuroImage 62:
Bassett, D.S. and Bullmore, E.T. (2017). 774–781.
Small-world brain networks revisited. Fischl, B. and Dale, A.M. (2000). Measuring the
Neuroscientist 23: 499–516. thickness of the human cerebral cortex from
magnetic resonance images. Proc. Natl. Acad. and other fallacies: the do’s and don’ts of
Sci. U. S. A. 97: 11050–11055. diffusion MRI. NeuroImage 73: 239–254.
Fornito, A., Zalesky, A., and Breakspear, Karamat, M.I., Darvish-Molla, S., and Santos-
M. (2013). Graph analysis of the human Diaz, A. (2016). Opportunities and challenges
connectome: promise, progress, and pitfalls. of 7 tesla magnetic resonance imaging: a
NeuroImage 80: 426–444. review. Crit. Rev. Biomed. Eng. 44: 73–89.
Fox, M.D. and Raichle, M.E. (2007). Spontaneous Kelly, C., Biswal, B.B., Craddock, R.C. et al.
fluctuations in brain activity observed with (2012a). Characterizing variation in the
functional magnetic resonance imaging. Nat. functional connectome: promise and pitfalls.
Rev. Neurosci. 8: 700–711. Trends Cogn. Sci. 16: 181–188.
Fox, P.T., Raichle, M.E., Mintun, M.A., and Kelly, C., Toro, R., Di Martino, A. et al. (2012b).
Dence, C. (1988). Nonoxidative glucose A convergent functional architecture of the
consumption during focal physiologic neural insula emerges across imaging modalities.
activity. Science 241: 462–464. NeuroImage 61: 1129–1142.
Friston, K. (1994). Functional and effective Lemaitre, H., Goldman, A.L., Sambataro, F. et al.
connectivity in neuroimaging: a synthesis. (2012). Normal age-related brain
Hum. Brain Mapp. 2: 56–78. morphometric changes: nonuniformity across
Friston, K.J. (2011). Functional and effective cortical thickness, surface area and gray matter
connectivity: a review. Brain Connect. 1: 13–36. volume? Neurobiol. Aging 33 (617): e1–e9.
Genon, S., Reid, A., Langner, R. et al. (2018). Lieberman, M.D. and Cunningham, W.A. (2009).
How to characterize the function of a brain Type I and type II error concerns in fMRI
region. Trends Cogn. Sci. 22: 350–364. research: re-balancing the scale. Soc. Cogn.
Genovese, C.R., Lazar, N.A., and Nichols, Affect. Neurosci. 4: 423–428.
T. (2002). Thresholding of statistical maps in Lin, C.S. and Yeung, A.W.K. (2020). What do we
functional neuroimaging using the false learn from brain imaging?-a primer for the
discovery rate. NeuroImage 15: 870–878. dentists who want to know more about the
Gong, G., He, Y., Chen, Z.J., and Evans, association between the brain and human
A.C. (2012). Convergence and divergence of stomatognathic functions. J. Oral Rehabil. 47:
thickness correlations with diffusion 659–671.
connections across the human cerebral Lin, C.S., Wu, S.Y., Wu, C.Y., and Ko, H.W. (2015).
cortex. NeuroImage 59: 1239–1248. Gray matter volume and resting-state
Henderson, L.A., Peck, C.C., Petersen, E.T. et al. functional connectivity of the motor cortex-
(2013). Chronic pain: lost inhibition? cerebellum network reflect the individual
J. Neurosci. 33: 7574–7582. variation in masticatory performance in healthy
Honey, C.J., Sporns, O., Cammoun, L. et al. elderly people. Front. Aging Neurosci. 7: 247.
(2009). Predicting human resting-state Logothetis, N.K. (2008). What we can do and
functional connectivity from structural what we cannot do with fMRI. Nature 453:
connectivity. Proc. Natl. Acad. Sci. U. S. A. 106: 869–878.
2035–2040. Logothetis, N.K. and Pfeuffer, J. (2004). On the
Huettel, S.A., Song, A.W., and McCarthy, nature of the BOLD fMRI contrast
G. (2004). Functional Magnetic Resonance mechanism. Magn. Reson. Imaging 22:
Imaging. Sunderland, MA: Sinauer Associates. 1517–1531.
Jenkinson, M. and Chappell, M. (2018). Malonek, D. and Grinvald, A. (1996).
Introduction to Neuoimaging Analysis. Oxford Interactions between electrical activity and
University Press. cortical microcirculation revealed by imaging
Jones, D.K., Knösche, T.R., and Turner, spectroscopy: implications for functional
R. (2013). White matter integrity, fiber count, brain mapping. Science 272: 551–554.
Murphy, K. and Fox, M.D. (2017). Towards a Rubinov, M. and Sporns, O. (2010). Complex
consensus regarding global signal regression network measures of brain connectivity: uses
for resting state functional connectivity and interpretations. NeuroImage 52:
MRI. NeuroImage 154: 169–173. 1059–1069.
Murphy, K., Birn, R.M., and Bandettini, Ruigrok, A.N., Salimi-Khorshidi, G., Lai,
P.A. (2013). Resting-state fMRI confounds and M.C. et al. (2014). A meta-analysis of sex
cleanup. NeuroImage 80: 349–359. differences in human brain structure.
Ogawa, S., Lee, T.M., Kay, A.R., and Tank, Neurosci. Biobehav. Rev. 39: 34–50.
D.W. (1990a). Brain magnetic resonance Smith, S.M. (2012). The future of FMRI
imaging with contrast dependent on blood connectivity. NeuroImage 62: 1257–1266.
oxygenation. Proc. Natl. Acad. Sci. U. S. A. 87: Sporns, O. (2013). Network attributes for
9868–9872. segregation and integration in the human
Ogawa, S., Lee, T.M., Nayak, A.S., and Glynn, brain. Curr. Opin. Neurobiol. 23: 162–171.
P. (1990b). Oxygenation-sensitive contrast in Thulborn, K.R., Waterton, J.C., Matthews, P.M.,
magnetic resonance image of rodent brain at and Radda, G.K. (1982). Oxygenation
high magnetic fields. Magn. Reson. Med. dependence of the transverse relaxation time
14: 68–78. of water protons in whole blood at high field.
Pereira, A.C., Huddleston, D.E., Brickman, Biochim. Biophys. Acta 714: 265–270.
A.M. et al. (2007). An in vivo correlate of Tononi, G., Sporns, O., and Edelman,
exercise-induced neurogenesis in the adult G.M. (1994). A measure for brain complexity:
dentate gyrus. Proc. Natl. Acad. Sci. relating functional segregation and integration
U. S. A. 104: 5638–5643. in the nervous system. Proc. Natl. Acad. Sci.
Poldrack, R.A. (2007). Region of interest analysis U. S. A. 91: 5033–5037.
for fMRI. Soc. Cogn. Affect. Neurosci. 2: 67–70. Turner, R. (2016). Uses, misuses, new uses
Poldrack, R.A., Fletcher, P.C., Henson, R.N. et al. and fundamental limitations of magnetic
(2008). Guidelines for reporting an fMRI resonance imaging in cognitive science.
study. NeuroImage 40: 409–414. Philos. Trans. R. Soc. Lond. Ser. B Biol. Sci. 371.
Price, C.J. and Friston, K.J. (1997). Cognitive Uddin, L.Q. (2013). Complex relationships
conjunction: a new approach to brain activation between structural and functional brain
experiments. NeuroImage 5: 261–270. connectivity. Trends Cogn. Sci. 17: 600–602.
Raichle, M.E. (2009). A brief history of human Van Dijk, K.R., Sabuncu, M.R., and Buckner,
brain mapping. Trends Neurosci. 32: 118–126. R.L. (2012). The influence of head motion on
Raichle, M.E. (2010). Two views of brain intrinsic functional connectivity
function. Trends Cogn. Sci. 14: 180–190. MRI. NeuroImage 59: 431–438.
Ridgway, G.R., Henley, S.M., Rohrer, J.D. et al. Watts, D.J. and Strogatz, S.H. (1998). Collective
(2008). Ten simple rules for reporting dynamics of ’small-world’ networks. Nature
voxel-based morphometry studies. 393: 440–442.
NeuroImage 40: 1429–1435. Zatorre, R.J., Fields, R.D., and Johansen-Berg,
Roy, C.S. and Sherrington, C.S. (1890). On the H. (2012). Plasticity in gray and white:
regulation of the blood-supply of the brain. neuroimaging changes in brain structure
J. Physiol. 11: 85–158. during learning. Nat. Neurosci. 15: 528–536.
Dental Neuroimaging: The Role of the Brain in Oral Functions, First Edition. Chia-Shu Lin.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
Companion website: www.wiley.com/go/lin/dental-neuroimaging
Figure 3.1 Methods of the assessment of oral cutting ability. (a) The sieving method quantifies the
proportion of the chewed food (e.g. peanuts) with different particle sizes, using multiple sieves with
different pore sizes (e.g. from the diameter of 355–3500 μm). The total weight of food particles that pass
through a sieve is plotted against the pore size of the sieve. A smaller median particle size (e.g. the grey
curve) represents better performance in cutting. Source: Chia-Shu Lin. (b) A test gummy jelly is customized
with a standardized size and shape. The chewed fragments are collected and photographed. Colour and
morphological features (e.g. the area and perimeter) of each fragment, which reflect individual cutting
ability, are assessed by analyzing the image. Source: Salazar et al. (2020). Reproduced with permission of
Elsevier.
single sieve is used (e.g. the US Standard No. 5 example, a stack of sieves may consist of a
sieve with a pore size of 4000 μm) (Cusson sieve with a larger pore size (4000 μm) and a
et al. 2015), masticatory performance (i.e. the sieve with a smaller pore size (2000 μm). If
cutting ability) is indexed by the percentage of 60% of the particles pass through the 4000 μm
the weight of the chewing food that passes sieve and 30% of them pass through the
through the sieve. A higher percentage of food 2000 μm sieve, the median particle size
particles through the sieve means better cut- would be somewhere between 2000 and
ting performance (Figure 3.1a). For the meas- 4000 μm. If a single sieve is used instead of
urement with multiple sieves, masticatory multiple sieves, the diameter of the single
performance can be indexed by the median sieve should be close to the median particle
particle size, which denotes the size (in the size of chewed food. Otherwise, the single
diameter of sieve pore) of a theoretical sieve sieve method would be less reliable com-
through which 50% (in weight) of the particles pared to the multiple sieve method (Van Der
can pass (Woda et al. 2010) (Figure 3.1a). For Bilt and Fontijn-Tekamp 2004).
3.1.2.2 Assessment of the Mixing Ability: of the colour features, such as the standard
The Two-colour Chewing-gum Test deviation of colours, the change in colour hue
The concept of assessing the mixing ability is and the pattern that colour portions are clus-
straightforward: if different portions of food tered (Lo et al. 2020; Halazonetis et al. 2013;
are well mixed, then mixed particles will form Schimmel et al. 2007). One of the widely used
a more homogenous group compared to the methods is to analyze the variance of the hue
original portions. The two-colour chewing- of the image (Halazonetis et al. 2013). The
gum test (TCGT) aims to quantify the degree of smaller standard deviation of hue (SDHue)
homogeneity of mixing by assessing the colour represents an increased homogeneity of colour
of different food portions. According to mixing, i.e. a better mixing ability. For exam-
Schimmel et al. (2007), this can be performed ple, the average value of SDHue decreased
by assessing the colour of a bolus mixed from from 0.144 to 0.023 when the chewing cycles
two pieces of chewing gums of different col- increased (i.e. a better masticatory effect) from
ours (Figure 3.2). The method offers a conveni- 20 to 50 cycles (Halazonetis et al. 2013).
ent and time-saving way for assessing
masticatory performance because chewing 3.1.2.3 Assessment of the Mixing Ability:
gum can easily be obtained, and at the chair- The Colour-changeable Chewing-gum Test
side, the degree of mixing (i.e. the homogene- The TCGT directly quantifies the change in
ity of colour features) can be judged by visual colour features induced by chewing, while the
inspection with good reliability (Silva colour-changeable chewing-gum test (CCGT)
et al. 2018). More precisely, the chewed bolus quantifies the degree of mixing indirectly via
can be digitalized by photographing or scan- the chemical reaction of the colour-changeable
ning, and the image of the bolus can be pro- gum (Hama et al. 2014; Hayakawa et al. 1998).
ceeded by imaging analysis. There have been The colour-changeable gum consists of two
various methods developed for quantifying the critical components: a pH-sensitivity dye that
degree of mixing according to different aspects develops red colour in alkaline conditions and
Figure 3.2 Methods of the assessment of oral mixing ability. (a) In the two-colour chewing gum test, the
degree of mixing food can be assessed by the colour hue of chewing gum with different colours. For
example, if a piece of red and a piece of yellow gums are well mixed, the resulting bolus would in orange
homogenously. The hue of the bolus can be quantified by imaging analysis. A smaller standard deviation of
hue represents a greater homogeneity of colour mixing, i.e. a better mixing ability. (b) The degree of mixing
is assessed according to the pattern of spatial clusters. A piece of juice chew with red and white portions
was chewed by a subject for 20 strokes and collected, as shown in the left panel. The degree of clustering is
assessed based on the analysis of variogram, which reflects how fine the clusters of different colours are. A
pattern with finer clusters (e.g. the case in the lower-right panel) reflects better mixing ability. Source: Lo
et al. (2020). Reproduced with permission of John Wiley and Sons.
citric acid that lowers the pH condition (i.e. higher chewing rate was also associated with
less alkaline). Before chewing, the pH value of a better mixing ability (Kikutani et al. 2009).
the gum is lowered by citric acid, and therefore Secondly, the number of chewing just before
the dye will not develop into a reddish colour. swallowing is a critical index for assessing
During chewing, the citric acid will be released behavioural adaptation in feeding (Peyron
and washed out. It increases the pH condition et al. 2004a). With a similar chewing rate,
(i.e. more alkaline) and leads to the develop- older people would chew longer than younger
ment of red colour (Tarkowska et al. 2017). people so that the total number of chewing
The mixing performance is then quantified by cycles (before swallowing) is greater in older
the degree of colour change before vs. after people. To increase the number of chewing
chewing. cycles would be an adaptative behaviour for
improving the feeding experience (Peyron
et al. 2004a).
3.1.3 Functional Assessments Related
to Masticatory Performance
3.1.3.2 Maximal Biting Force
The number of existing teeth, particularly The biting force has been conceived as a major
the teeth for posterior contact, is associated factor related to masticatory performance. The
with masticatory performance (Ikebe MBF can be evaluated either by clenching
et al. 2010, 2012). In addition, many sensori- with the whole dental arch (i.e. ‘full arch’)
motor factors of oral functions are associated (Ikebe et al. 2012) or clenching on a single pair
with masticatory functions, including the of molars (i.e. ‘single pair’) (Ogura et al. 2012;
number of chewing cycles, the maximal bit- Palinkas et al. 2010). Pressure films and
ing force (MBF) and the ability of oral stere- dynamometers are commonly used for the
ognosis. In the following sections, we briefly full-arch and single-pair assessment, respec-
outline the factors related to masticatory per- tively. Notably, the MBF is closely associated
formance, especially focusing on the factors with personal factors, including age and gen-
associated with sensorimotor control of the der. For example, the right-molar single-pair
stomatognathic system. MBF reduced from 339 to 324 N in male sub-
jects and from 221 to 203 N in female subjects,
3.1.3.1 The Number of Chewing Cycles in the younger (21–40 y/o) vs. the older (41–60
The number of chewing cycles can be y/o) subjects (Palinkas et al. 2010). In general,
assessed from two aspects. Firstly, the num- a lower occlusal force is associated with a
ber of chewing cycles within a fixed period is lower cutting ability (Ikebe et al. 2012).
defined as the chewing rate. The central pat- However, it should be noted that both factors
tern generator of the brainstem plays a piv- are closely related to the number of functional
otal role in rhythmic jaw movement (Moore teeth. In a study with large-sampled elderly
et al. 2014). Therefore, the central nervous people (aged ≧ 60 years), the mean full-arch
system (CNS) mechanisms are critical for MBF reduced from 530 N (for Eichner class A)
maintaining a constant rate of chewing. to 220 N (for Eichner class C), and the cutting
Additionally, the chewing rate of chewing a ability reduced from 2659 mm2 (Eichner class
piece of gum is also associated with the num- A) to 1316 mm2 (Eichner class C) (Ikebe
ber of functional teeth. In a study with 268 et al. 2012). Additionally, using magnetic res-
healthy older people (aged between 65 and onance imaging (MRI), researchers have
88 years), the mean chewing rate was 203.7 identified a great individual difference in the
(cycle/3 minutes) and 170.3, respectively, for size of the masseter, the major jaw-closing
older people with Eichner class A~B3 and muscles associated with mastication. The
Eichner class B4~C (Kikutani et al. 2009). A masseter muscle volume (MMV) showed a
weak-to-moderate and statistically signifi- was associated with better masticatory per-
cant correlation with the mixing ability (Lin formance (Ikebe et al. 2007b). In healthy
et al. 2017a). Biting force is also associated adults, the ability to perceive the size of
with the processing of sensory feedback intraoral objects was associated with the cut-
(Trulsson et al. 2012). Differences in the MBF ting ability (Engelen et al. 2004). The role of
can be found between the individuals with perceptual processing in oral stereognosis is
natural teeth and those who with dental also evidenced by an earlier functional mag-
implants (Trulsson et al. 2012). Because peri- netic resonance imaging (fMRI) study, which
odontal mechanoreceptors are removed dur- revealed that oral stereognosis was associ-
ing extraction, normal sensory feedback (via ated with functional brain features (Fujii
the periodontium) is compromised in an et al. 2011). The findings suggest that the
implant-supported prosthesis. Therefore, in individual difference in oral stereognosis
the patients, changes in the MBF may be may be associated with not only peripheral
associated with altered sensory processing. It but also central factors.
is also noteworthy that during eating, the bit-
ing force may interact with the types of food.
3.1.4 Functional Assessments
For example, the reduction of the maximal
for Swallowing Performance
contraction of the masticatory muscles was
noted in older people, compared to younger In contrast to the assessment of masticatory
people, only when they chewed a harder food performance, there have been more assess-
(Galo et al. 2007). ments of swallowing performance being devel-
oped. As summarized in the following sections,
3.1.3.3 Oral Stereognosis many of the tests are used for assessing the risk
Oral stereognosis refers to the ability to rec- of dysphagia.
ognize and discriminate an intraoral object
(Jacobs et al. 1998). Clinically, assessments 3.1.4.1 Water Swallowing Test
of oral stereognosis can be conducted by ask- The water swallowing test (WST) is one of the
ing subjects to evaluate the size or shape of a most widely used chairside methods for
standardized object intraorally. For example, assessing the ability to swallow. It is a simple
in one study, subjects needed to evaluate the task that requires the subject to swallow 3 ml
size of steel spheres with the diameter vary- of water, and the assessor identifies whether
ing between 4 and 9 mm (Engelen et al. 2004). any choke occurs during swallowing (Brodsky
In the other studies, they were asked to dis- et al. 2016). The test is mainly used for screen-
criminate the size (12 × 12 × 3 mm3 vs. ing potential patients with functional dyspha-
8 × 8 × 2 mm3) and the shape (circles, ellip- gia. A recent meta-analysis based on 22
ses, semicircles, squares, rectangles and tri- studies revealed that the WST offers sufficient
angles) of testing objects (Hirano et al. 2004; screening for aspiration (Brodsky et al. 2016).
Kawagishi et al. 2009; Ikebe et al. 2007b). Notably, the sensitivity and specificity for
The performance can be quantified accord- screening aspiration may be associated with
ing to the accuracy of evaluation and the the methodological variations of the WST,
response duration (Hirano et al. 2004). Older including the volume to sip and the way of
dentate subjects showed a worse perfor- sipping (i.e. consecutive vs. single sips)
mance in the assessment of oral stereognosis (Brodsky et al. 2016). Another recent meta-
compared to younger subjects (Ikebe analysis on patients with stroke also reported
et al. 2007a). In older edentulous patients that WST is a useful screening tool for evalu-
wearing a complete denture, a better perfor- ating the aspiration of stroke patients (Chen
mance of the oral stereognosis assessment et al. 2016).
3.1.4.2 Repetitive Saliva Swallowing Test swallowing actual foods, which would vary
The repetitive saliva swallowing test (RSST) is a greatly due to different eating and diet habits
simple chairside test developed and popularized across individuals. To this purpose, self-report
in Japan (Oguchi et al. 2000a,b). The test tools, such as questionnaires, are designed to
requires a subject to perform voluntary swal- assess the difficulty when one is chewing or
lows repeatedly in 30 seconds. The score of the swallowing normal meals. A combination of
RSST is indexed by the number of swallows that both the patient-based assessment (for individ-
the subject can complete during 30 seconds, ual eating experience) and laboratory-based
which can be identified using the palpation of assessment (for standardized performance)
laryngeal elevation (Oguchi et al. 2000a,b). The should be considered for a full-scale evaluation
subjects are instructed to swallow as many of oral functions (Woda et al. 2011). The self-
times as possible (Lin et al. 2019; Persson report assessments for the qualitative experi-
et al. 2019). During the procedure, the subject ence of mastication and swallowing will be
cannot drink or eat anything but swallows his/ outlined in the following sections.
her own saliva. The RSST score revealed a sig-
nificant age-related difference, reducing from 3.1.5.1 Masticatory Experience
7.4 ± 1.7 (mean ± standard deviation) in younger The design of self-report assessment for masti-
adults to 5.9 ± 2.3 in older adults (mean age catory experience can be simple and straight-
68.1 years) (Oguchi et al. 2000b). For its clinical forward, such as a single yes/no question that
validity, safety and simplicity, the RSST is sug- assesses if subjects are satisfied with their
gested for evaluating functional dysphagia chewing activity or not (Miura et al. 1998). The
(Oguchi et al. 2000a). However, it should be assessment may be a more complicated ques-
noted that the RSST score also reflects the abil- tionnaire that assesses different aspects of
ity of voluntary swallowing, which is regulated the eating experience. The experience to be
by the CNS (Ertekin 2011). Therefore, the test assessed may include one’s sensory, motor and
score may reflect not only the function of the emotional experience regarding mastication
oropharyngeal system but also the CNS mecha- and behavioural changes associated with the
nisms of swallowing. As shown in a recent difficulty of mastication. For example, Tsuga
study, the RSST score was associated with indi- et al. (1998) used a customized questionnaire
vidual variability in structural brain features to assess the masticatory ability of 160 elderly
(Lin et al. 2019). The individual difference in participants (aged ≧ 80 years). The question-
the RSST score may be associated with both naire assessed both sensory experiences, such
peripheral and central factors. as pain during chewing and dryness of mouth,
and behavioural changes related to mastica-
tion, such as taking a long time to finish a meal
3.1.5 Self-report Assessments
and avoid eating with others (Tsuga et al. 1998).
of Eating Experience
The result showed that a higher score of masti-
So far, the masticatory and swallowing assess- catory experience (i.e. more problems during
ments discussed here focus on assessing one’s chewing) was significantly correlated with a
ability to masticate and swallow. However, these lower number of teeth and a lower MBF (Tsuga
functional assessments do not provide informa- et al. 1998). Notably, individual eating experi-
tion about the qualitative experience of mastica- ence may vary greatly due to individual differ-
tion and swallowing when individuals are ences in diet habits. There exists a great
eating. Unlike the functional assessments that regional/cultural difference in diet habits.
quantify oral performance based on a standard- Therefore, research findings cross-referred
ized testing condition, these assessments focus between different countries or regions should
on individual experience about chewing and be carefully interpreted. To better evaluate the
association between masticatory experience ‘sometimes I eat here on the gum’ (Zelig
and diet, the types of food to eat are specified et al. 2019). These findings suggested that
in some assessments, and eating experience is older people require more attention and cog-
assessed for each type of food. For example, nitive control when eating in a challenging
the food questionnaire by Koshino et al. (2008) condition. On the contrary, when adaptation
requires subjects to rate the eating experience failed, maladaptive behaviour such as avoid-
of 25 food items based on Japanese cuisine. ance of some types of food may occur (Zelig
Subjects’ difficulty of chewing the selected et al. 2019).
group of food is indexed as a masticatory score,
which showed a significant correlation with 3.1.5.2 Swallowing Experience
their cutting ability, as assessed using the sieve Several questionnaires have been developed
method. In a similar vein, Hsu et al. (2012) for screening older people with a risk of dys-
designed a questionnaire consisting of 23 food phagia. The Eating Assessment Tool (EAT-10)
items based on Taiwanese cuisine. Their find- is a self-rating questionnaire consisting of
ings suggested that lesser masticatory diffi- 10 items, which assesses individual experi-
culty of the food was associated with better ence during swallowing (Belafsky et al. 2008).
preservation of teeth, i.e. more than 20 existing As a tool for evaluating the risk of dysphagia,
natural teeth and at least eight functional teeth the questionnaire focuses on the individual
units (Hsu et al. 2012). Since diet habit is experience of swallowing difficulty, including
closely related to regional and cultural factors, the occurrence of pain and cough during swal-
the studies of masticatory experience should lowing. Moreover, it differentiates the swal-
adopt a culturally valid tool to better reflect the lowing of different types of food, such as
real eating experience of subjects. liquids, solids and pills, because different
It is noteworthy that at present most of the physical properties of the food may influence
questionnaires have focused on the experience one’s swallowing experience. The EAT-10
of the difficulty of chewing. In other words, showed good internal consistency and test–re-
the questionnaires are mainly used for evalu- test reliability and a good clinical validity in
ating the disturbances (e.g. pain) associated discriminating patients with and without dys-
with chewing. A limitation of these assess- phagia (Belafsky et al. 2008). An EAT-10 score
ments of ‘masticatory difficulty’ is to neglect of 3 or higher can be abnormal in swallowing
the fact that individuals will develop some function based on the normative data from
adaptive strategies to cope with these chal- 482 patients (Belafsky et al. 2008). Another
lenges during chewing (Bourdiol et al. 2020; assessment designed for assessing swallowing
Woda et al. 2010). For example, older people difficulty is the swallowing disturbance ques-
with tooth loss or a newly installed denture tionnaire (SDQ), which consists of 15 yes/no
may adapt to their oral condition via some questions about swallowing disturbances. The
behavioural strategies, such as chewing with total SDQ score was associated with the score
more cycles or chewing with a different side. from fibreoptic endoscopic evaluation of
Until now, the experience of adaptation of swallowing (FEES) (Manor et al. 2007). The
oral conditions has been rarely investigated Swallowing Quality of Life questionnaire
(Al-Sahan et al. 2020). A recent study reported (SWAL-QOL) is a longer questionnaire con-
that older people (aged ≧ 65 years) with tooth sisting of 44 items (McHorney et al. 2002). The
loss would adopt different strategies to cope questionnaire assesses the experience of swal-
with their conditions (Zelig et al. 2019). Some lowing difficulty from multiple dimensions,
adaptive chewing strategies were used for a including its effect on social interaction, sleep
better eating experience, such as ‘I’ve got to and fatigue, thus focusing on the various
make sure that I have chewed it real good’ or aspects of quality of life. The validity of the
(Liang et al. 2015). Together, the findings sug- the association between oral functions and the
gest that via assessment one can better clarify factors related to diet, including the selection
the association between clinical interventions and processing of food. For example, the sieve
and improvement of oral functions brought by method can be used for quantifying the parti-
the interventions. cle size distribution of food boluses (Peyron
The use of assessment also contributes to et al. 2004b). Based on a test of six natural
tracing the change of oral functions over a long foods, Peyron et al. (2004b) found that upon
period. For example, the longitudinal fMRI the completion of mastication, the sizes of the
study by Luraschi et al. (2013) revealed that the chewed particles were similar among the nuts
mixing ability and MBF significantly increased and among vegetables, and only little inter-
when the patients adapted to a new denture individual variability was found for the
during a three-month follow-up. In contrast, chewed particle size. The findings echoed the
some research did not show a significant importance of maintaining an optimal size of
change in the cutting ability over one year of food particles for swallowing. Upon swallow-
treatment (Aras et al. 2009). It should be noted ing, one needs to chew the food until it is bro-
that at present most studies of functional ken down into small particles with sufficient
assessments have focused on their validity, lubrication, and the detection of particle size
which can be demonstrated by a correlation with oral mucosa may play a key role (Prinz
with other clinical symptoms and signs (e.g. and Lucas 1995). With the results from a func-
the more the missing teeth, the lower the mas- tional assessment of mastication, one can
ticatory performance) (Woda et al. 2010; quantify the association between individual
Tarkowska et al. 2017). There exists less evi- masticatory performance and the properties
dence regarding the temporal stability, i.e. the of food.
test vs. re-test reliability, of the assessments
because most research was conducted with a
3.1.7 Summary
cross-sectional design. Good temporal stability
for either functional or questionnaire assess- ●● Masticatory performance, i.e. the degree of
ments would be essential for evaluating the comminution, can be assessed according to
change of oral functions over a long period. the size of the food being cut down (i.e. the
cutting ability) or the degree that food is
3.1.6.2 Evaluation of the Association Between mixed to form a bolus (i.e. the mixing
Oral and Systemic Factors ability).
Secondly, through the assessments, it becomes ●● The degree of food mixing can be experi-
feasible to investigate the association between mentally quantified by how different colour
oral functions and the systemic factors related portions are mixed during chewing, such as
to general physical/mental conditions. In chewing two pieces of gums of different
terms of physical conditions, a recent study colours.
with a large sample size (n = 5104) of elderly ●● A greater number of teeth, a stronger MBF
people (aged ≧ 65 years) showed that the sta- and a better ability of oral stereognosis may
tus of physical frailty, including non-frail, pre- be associated with better masticatory
frail and frail, was associated with the full-arch performance.
biting force (Watanabe et al. 2017). In terms of ●● The subjective experience of mastication
cognitive impairment, patients with cognitive and swallowing, as assessed using question-
impairment showed a worse masticatory mix- naires, may provide information about the
ing and cutting performance (Campos quality of individual swallowing. The ques-
et al. 2017; Kim et al. 2017). In addition, the tionnaires for assessing swallowing experi-
assessment of oral functions helps to clarify ence can be an important tool for screening
the risk of dysphagia. The questionnaires the quantitative sensory testing (QST) (Rolke
may reflect the real eating experience and et al. 2006), are more frequently conducted at
masticatory difficulty of patients. the chairside. The experimental design of
neuroimaging research combined with these
assessments is also discussed.
3.2 Assessment of Orofacial Pain
and Somatosensory Experience 3.2.2 Oral Assessment and Psychophysics
Psychophysics provides useful theoretical
3.2.1 Introduction
foundations for the assessment of the sen-
Using neuroimaging methods, researchers sory experience. The major purpose of psy-
have gradually disclosed the central mecha- chophysics is to establish a psychophysical
nisms underlying the complicated symptoms function that depicts the stimulus–response
of chronic pain. In most of the studies, the relationship. For example, a psychophysical
difference of brain features is reported to function can be established, based on empir-
reflect the difference between patients with ical evidence, to map the relationship
pain and healthy controls or the individual between the intensity of a physical stimulus
variability in pain-related symptoms (for a (e.g. the weight of an object) and the subjec-
detailed review, see Chapter 6). Notably, these tive experience induced by the stimulus (e.g.
neuroimaging findings are mostly based on the feeling of ‘heaviness’). Psychophysical
the results of clinical assessment, including approaches are useful to investigate the sen-
the assessment of pain, pain-related emo- sory threshold. An absolute threshold refers
tional and behavioural factors (e.g. avoidance to the amount of a stimulus that induces a
of pain), and sensory disturbance related to sensation. Experimentally, to identify the
oral functions (e.g. altered sensation in chronic absolute threshold of a stimulus, subjects
orofacial pain. In terms of oral neuroscience, a are asked to detect the presence of a sensa-
proper assessment of the clinical symptoms tion. Notably, because subjects’ responses
and signs related to pain and somatosensory may vary between different trials of detec-
disturbance is the foundation for neuroimag- tion, their responses are usually recorded
ing research on brain mechanisms. In the fol- many times, and a threshold is identified
lowing sections, we discuss some widely used when the sensation is detected half of the
methods for assessing orofacial pain and soma- time. For example, a threshold of thermal
tosensory experience. Especially, we focus on pain is identified if, at a certain tempera-
the assessments that are commonly used for ture, subjects would feel the heat stimulus
neuroimaging research. Just like the assess- to be painful in five trials and non-painful in
ments for oral functions, the assessments for another five trials (out of 10 trials). The con-
pain and somatosensory experience can be cat- cept of an absolute threshold is different
egorized into functional assessments and ques- from a differential threshold, which aims to
tionnaire assessments, which focus on the quantify individual ability to discriminate
psychophysical metrics of pain and sensation the intensity change of a physical stimulus.
(e.g. pain threshold) and subjective experience Experimentally, the differential threshold is
of pain and sensation (e.g. the unpleasantness identified when subjects can just detect the
of pain). As discussed in the following sec- presence of ‘difference’ of physical intensity
tions, some questionnaire assessments can be between two stimuli (denoted as Δ). Here,
conducted inside an MRI scanner as part of the the individuals with a higher sensitivity of
task condition in a task-based fMRI study. discrimination are those who can discrimi-
While some functional assessments, such as nate a smaller Δ.
Understanding the basic concepts and meth- may provide quantified results as discrete
ods of psychophysics is critical to the assess- scores, with information of an ordinal level of
ment of clinical symptoms and signs. For measurement, such as the descriptors of pain
example, in electrical pulpal testing, dentists as ‘very painful’ and ‘moderately painful.’ Some
are tuning the intensity of electrical stimuli scales may provide results as continuous scores,
while patients need to respond if the stimulus with information for interval or ratio measure-
is painful or not. In terms of psychophysics, the ment, such as the results from the VAS. In addi-
approach requires experimenters to alter the tion to pain, other subjective experiences can
stimulus intensity until it matches a fixed level be quantified using a similar design of a scale
of subjects’ percept (e.g. ‘feeling pain’). In the by changing the instruction and the anchoring
approach of cross-modality matching, a series descriptors of a scale. For example, a scale can
of stimuli intensity are presented, and subjects be used to assess the degree of anxiety or pain
need to match each stimulus proportionally to by defining the start point and end point as ‘not
their percept on another scale. A common anxious at all’ and ‘extremely anxious,’ respec-
application of cross-modality matching is the tively. In the following sections, we first discuss
visual analogue scale (VAS) for assessing pain, the scales for assessing pain. Due to its reliabil-
in which subjects need to proportionally map ity, validity and simplicity, scales are frequently
between their pain and the percept of length as used in functional neuroimaging studies for
denoted by the VAS. The use of psychophysical assessing subjective experience induced during
principles constitutes the basic elements of the task condition.
clinical pain scales and quantitative sensory
tests (see Box 3.1). 3.2.3.1 Numerical Rating Scale
A typical numerical rating scale (NRS) is pre-
sented with only integral numbers and two
3.2.3 Quantifying Pain Using
anchoring descriptors. No additional verbal
the Pain Scales
descriptor is included in an NRS. The NRS has
A scale is a very simple form of assessment to been considered a valid tool for assessing both
quantify a subjective experience. Some scales acute pain induced by cold (Ferreira-Valente
et al. 2011) and heat stimuli (Price et al. 1994). assess acute pain (Bijur et al. 2001). In contrast
It also shows good validity and reliability for to the NRS, the VAS is commonly considered
assessing chronic pain, including cancer pain as a ratio scale. Based on a test of thermal stim-
(Jensen 2003) and osteoarthritic knee pain ulation, Price et al. (1983) concluded that the
(Alghadir et al. 2018). The NRS is widely used score from a VAS is a valid tool as a ratio scale
in assessing chronic orofacial pain, such as the for assessing pain.
Graded Chronic Pain Scale (GCPS), also part of
the assessment tools of the Diagnostic Criteria 3.2.3.3 Verbal Descriptor Scale
for Temporomandibular Disorders (DC/TMD) A verbal descriptor scale (VDS) is commonly
(Schiffman et al. 2014). The scale consists of seen in the chairside assessment of pain. A
11 integers (0–10) with the anchoring descrip- typical VDS consists of a fixed number of
tors ‘0’ for ‘No pain’ and ‘10’ for ‘Pain as bad as descriptors that are systematically organized.
could be.’ Notably, it is under debate if the For example, in the SF-MPQ (Melzack 1987),
scores from an NRS should be considered an four descriptors, ‘none,’ ‘mild,’ ‘moderate’ and
ordinal or a ratio level of measurement (Price ‘severe,’ are used for assessing different pain
et al. 1994). Some researchers highlight the use experiences. The descriptors are also pre-
of an NRS only for comparing between differ- sented as adjectives or combinations of adjec-
ent scores, such as the condition rated as ‘8’ is tives and adverbs, such as ‘moderately painful’
more painful than that rated as ‘4.’ While other or ‘extremely painful.’ Subjects need to choose
researchers concluded that an NRS showed a the single descriptor best matching for their
good sensitivity as the VAS, its score can be experience. The score from the VDS is regarded
analyzed with parametric methods (Ferreira- as an ordinal rating. Notably, the use of VDS
Valente et al. 2011). should not be confused with the assessment of
the quality of pain. The assessment of the
3.2.3.2 Visual Analogue Scale quality of pain also relies on the use of
A typical VAS is exemplified by the short-form ‘descriptors,’ but these descriptors mainly
McGill Pain Questionnaire (SF-MPQ) describe the quality (i.e. how it feels like) of
(Melzack 1987). The scale consists of a hori- pain, such as ‘throbbing’ or ‘shooting,’ as
zontal line with two anchoring descriptors, ‘No exemplified by the SF-MPQ.
Pain’ at the left end and ‘Worst Possible Pain’ at
the right end. According to its instruction, sub- 3.2.3.4 Comparison Between the Pain Scales
jects need to indicate how bad the pain is on From the practical perspective, it is not
the line by noting that ‘at the left end of the uncommon that researchers would adopt dif-
line means no pain at all’ and ‘at the right end ferent scales for assessing pain in a study
means worst pain possible’ (Melzack 1987). since different scales have their pros and cons
Notably, a typical VAS is composed of only the for specific conditions. In general, the pri-
line and anchors, without additional numbers mary consideration of choosing a pain scale
or descriptors attached. Likewise, an NRS is is the subjects’ ability to give a proper
composed of only numbers. The VAS has been response and the clinical scenarios where
considered a valid tool for assessing both acute assessment is conducted. A scale used for
pain induced by cold (Ferreira-Valente unsuitable subjects or in the suboptimal situ-
et al. 2011) and heat stimuli (Price et al. 1994). ation will greatly compromise the validity of
It also shows good validity and reliability for the assessment. For example, the visuomotor
assessing chronic pain, including cancer pain demand of using a VAS would cause a major
(Jensen 2003) and osteoarthritic knee pain challenge to patients with sensorimotor dys-
(Alghadir et al. 2018). Researchers have also functions (e.g. the patients with visual defi-
found good reliability in using the VAS to cits or Parkinson’s disease) (Perez-Lloret
et al. 2016). In contrast, a VDS can be used the VAS. Therefore, one should keep in mind
via a telephone call and is still a reliable tool that when a direct comparison is made
for patients with cognitive impairment (Ware between pain derived from different scales,
et al. 2006). Combining the results from dif- the psychometric nature of the scales should
ferent scales for a clinical interpretation is be carefully considered.
also challenging because the scales aim to
assess different experiences of pain. In an 3.2.3.5 Pain Scales Based on Diagrams
earlier study, Tammaro et al. (1997) have In addition to an NRS, a VAS and a VDS, there
investigated the difference in the individual have been pain scales developed for special
representation of pain rated on a VDS and a age groups, such as the Faces Pain Scale-
VAS in dental students, patients receiving Revised (FPS-R) for children (Hicks
periodontal treatment and patients with den- et al. 2001) and the Revised Iowa Pain
tal phobia. Subjects were asked to rate the Thermometer (IPT-R) for older people (Ware
degree of pain described by five common et al. 2015). Both scales adopt a diagram, in
descriptors on a VAS. The results showed company with numbers and/or verbal descrip-
that, for example, subjects would rate ‘mild tors, to help patients indicate their pain expe-
pain’ averagely as 10 (out of a 100 mm VAS) rience. The FPS-R consists of a series of face
and ‘moderate pain’ as 40 (Tammaro pictures that reveal painful expression aligned
et al. 1997). Notably, the discrepancy between with the numbers 0–10, where 0 means ‘no
mild and moderate pain (represented by the pain’ and 10 means ‘very much pain.’ The
VAS score) is smaller than that between mod- original IPT consists of a diagram of a ther-
erate and intense pain, and such a nonlinear mometer, positioned vertically, aligned with
relationship varied between the three study 13 verbal descriptors from the bottom area
groups, who differed in their anxiety level (‘no pain’) to the top area (‘the most intense
(Tammaro et al. 1997). The inconsistency pain imaginable’) (Herr and Mobily 1993).
exists not only between a VDS and a VAS but In the following version IPT-R, the same
also between an NRS and a VAS. In another thermometer diagram is aligned with both
study, Bijur et al. (2003) have asked the numbers (0–10) and verbal descriptors,
patients of an emergency department to rate accordingly. The IPT-R has shown good relia-
their pain using the NRS and the VAS. It is bility and validity for assessing the pain of
not surprising that when patients suffered older adults with a diverse cognitive status
minor pain or extreme pain, the two scales (Ware et al. 2015).
gave very consistent results. However, when
the pain is moderate to strong, the discrep- 3.2.3.6 Limitations and Challenges of Using
ancy between the scales was pronounced. For Pain Scales
example, two subjects may both rate their Some of the common problems of using pain
pain as seven on a 0–10 NRS. However, one scales are listed as follows. Firstly, a combina-
would rate the same pain as 80 in a 100 mm tion of verbal descriptors on an NRS, such as a
VAS, and another may rate the pain as 60 in word ‘mild’ nearby ‘3’ and a word ‘moderate’
the VAS. In other words, the consistency nearby ‘5,’ may confuse subjects, especially
between a rating from an NRS and a rating when the choice of the descriptors and their
from a VAS may vary between individuals. As location on the scale are not well validated. As
shown in the previous study, though the shown in previous cases, a VAS consists of only
scores from both scales were significantly the line and the NRS consists of only the num-
correlated (r = 0.94) (Bijur et al. 2003), the bers, as shown in the SF-MPQ (Melzack 1987)
results revealed substantial individual varia- and the GCPS (Schiffman et al. 2014). Secondly,
tions in the consistency between the NRS and the assessment of pain scales without any
instruction (or with a wrong instruction) should 3.2.4.1 Composition of the QST
be avoided. This is particularly important for an The QST consists of seven tests that measure 13
assessment of a more complicated experience of sensory factors, including (i) thermal detection
pain. For example, one question of the GCPS thresholds of cold detection threshold (CDT),
requires subjects to indicate their ‘maximal warm detection threshold (WDT) and paradox-
pain’ during a period, and in another question, ical heat sensations (PHS), thermal sensory
subjects need to indicate their ‘average pain’ limen (TSL), and thermal pain thresholds,
during that period. If the instruction is not clear including cold pain threshold (CPT) and hot
or confusing, the results can be misleading. pain threshold (HPT); (ii) mechanical detec-
Thirdly, the influence of the ‘anchors’ should be tion threshold (MDT); (iii) mechanical pain
noted. While the minimal point of a pain scale threshold (MPT); (iv) stimulus–response func-
should refer to ‘no pain,’ the meaning that the tions, including mechanical pain sensitivity
maximal point would refer to is less clear-cut. A (MPS) and dynamic mechanical allodynia
study investigated 183 patients with chronic (ALL); (v) temporal pain summation, which
pain and found that the anchoring descriptor measures the wind-up ratio (WUR) of repeti-
‘the worst pain experienced/bearable’ was the tive pinprick stimuli; (vi) vibration detection
most preferable, followed by the descriptor ‘the threshold (VDT) and (vii) pressure pain thresh-
worst pain.’ Though the scores assessed by all old (PPT) (Rolke et al. 2006). The testing results
the scales are statistically significantly corre- may reflect the engagement between pain/
lated (Yokobe et al. 2014). Finally, pain assess- somatosensory experience and peripheral sen-
ment of non-verbal individuals, e.g. patients sory channels. For example, the touch and
with severe dementia, is still a challenging issue vibration detection thresholds (i.e. MDT and
(Lobbezoo et al. 2011). Special assessment VDT) may reflect the processing of mecha-
should be used to assess orofacial pain of patients nosensation innervated by Aβ fibre, and the
with dementia (Lobbezoo et al. 2017). MPT may reflect the processing of nociception
innervated by Aδ and C fibres (Zhou et al. 2018).
the sensory thresholds. For example, the CDT example, an NRS of pain can easily be modi-
revealed that the gingiva showed a lower sensi- fied for assessing the disability associated with
tivity to cold stimuli (12.3 °C from baseline pain, as shown in the GCPS. It should be noted
temperature) compared to facial skin (1.2 °C that assessing the pain-related experience,
from baseline temperature). The gingival area rather than pain intensity per se, is an essential
also showed a higher MDT (17.2 mN) com- part of the assessment of orofacial pain.
pared to the tongue (0.3 mN) (Pigg et al. 2010). Because pain is defined as a multi-facet experi-
The difference may reflect the structural varia- ence, which includes sensory and emotional
tion between these sites, such as the degree of experiences (IASP 2020), only assessing
keratinization (Zhou et al. 2018). patients’ pain intensity may not provide a full
picture of their suffering. In the following sec-
3.2.4.3 Practical Issues of Using the QST tions, we focus on the emotional and behav-
A major challenge of adopting QST at the ioural experiences that are critical for assessing
chairside is the demand for an extra device. orofacial pain.
For example, the tests of thermal sensations
need to be conducted using a computer- 3.2.5.1 Assessment of Emotional
controlled device that can generate the gradi- Experiences of Pain
ent of thermal stimuli with a fixed speed Until now, very few studies have reported the
(Rolke et al. 2006). In addition, most of the emotional experience regarding pain in the
tests require the use of standardized instru- field of oral neuroscience (except for fear and
ments, such as pinprick mechanical stimula- anxiety, see Chapter 6). Even though pain is
tors for MPT and von Frey hairs for MDT. All defined as both sensory and emotional experi-
the instruments need to be well calibrated. ences (IASP 2020), most of the pain scales
Even though the original QST is not very pop- used in clinical studies do not differentiate
ular for ordinary oral investigation, the psy- between the two aspects. The usage of scales
chophysical mechanisms behind the QST are to evaluate the affective dimensions of
commonly seen in daily dental practice. As patients’ pain, as one of the major roles of
shown in the preceding section, the use of an pain assessment (Gracely and Dubner 1981),
electrical pulp tester (EPT) can be considered is largely ignored. Though the separate assess-
an assessment of pain based on an approach ment of pain intensity and unpleasantness is
similar to the QST. During the EPT, patients relatively less seen in the dental field, it is not
need to respond to the electrical stimuli of uncommon for dentists to assess the fear and
different intensities on their teeth by report- anxiety of dental patients, which are usually
ing ‘yes’ or ‘no’ to indicate if they perceived associated with their pain. These subjective
the stimuli or not. Based on the same psycho- feelings can be quantified using the same
physical principles, some simplified methods methods as the assessment of pain, such as
for sensory testing have been proposed for the same VAS with different anchors and
specific clinical purposes, such as the assess- instructions. The real challenge would be
ment of sensory disturbance after mandibu- how to interpret the results. The emotional
lar implant surgery (Misch and Resnik 2010). experiences may be more likely to vary due to
different clinical settings. For example, the
fear and anxiety of pain are associated with
3.2.5 Assessment of Emotional
the general conditions of dental treatment
and Behavioural Experiences of Pain
and the perceived threat towards a specific
As noted earlier, the same psychophysical scenario (e.g. the moment to receive a needle
principles can be adopted for assessing both injection). Therefore, assessors need to clarify
pain and other subjective experience. For the experience that subjects should focus on,
phase (Figure 3.3a). This approach provides a experimental design is to ask patients to trace
simple framework based on psychophysics for their current pain using a VAS-like device,
researchers to map the relationship between which consists of a finger-span device. The
stimuli and response (i.e. pain and brain acti- device was synchronized with the scanning
vation). Furthermore, in a task-based fMRI protocol of fMRI, and its results are presented
study, pain can be modulated by different as a bar via visual feedback. Therefore, subjects
experimental factors. The effect of these can continuously rate their spontaneous pain
manipulations (e.g. strong pain vs. mild pain) by changing the size of the visual bar via
can be identified by the corresponding change manipulating the distance of the span between
in ratings. the thumb and index finger (Baliki et al. 2006;
The approach stated above is suitable for Apkarian et al. 2001) (Figure 3.3b). The design
short-lasting and trial-based stimulation, such can provide a continuous and dynamic record
as acute pain induced by electrical stimulation. of patients’ pain, and further imaging analyses
However, for patients with chronic pain, pain can identify the brain regions where activation
could be ‘always there,’ without any stimula- corresponds to this dynamic pain experience.
tion to trigger it. In addition, because pain is a Moreover, different ‘stages’ of spontaneous
long-lasting experience, it is hard to be arbi- pain can be extracted from the continuous
trarily dissected into different phases. record. For example, one can identify the
Therefore, to assess the brain activation associ- epochs of high spontaneous pain, defined as
ated with long-lasting and spontaneous pain, a the periods that pain ratings are higher than
different approach can be adopted. A suitable the mean of pain ratings. One can also identify
the epochs of rapidly increasing pain, defined features, such as the morphological features
as the periods that pain ratings show a positive of grey matter and white matter, are extracted,
rate of change for a sustained duration (Baliki and their correlation with pain ratings is ana-
et al. 2006). lyzed (Figure 3.3c). Again, the assessment of
pain can be performed somewhere outside
the scanner, i.e. separate from the scan of
3.2.6.2 Separate (‘Off-line’) Record of Brain structural images. The approach provides a
Features and Ratings simple framework for researchers to investi-
An obvious drawback of the design stated gate the association between individual vari-
above is that the inclusion of the rating phase ability of functional/structural brain features
may prolong the total time of scanning. For and the variability of clinical symptoms
example, the stimulation phase may last for (e.g. pain).
just three seconds (e.g. a quick electrical stim-
ulation), but the rating phase lasts for more
3.2.7 Summary
than six seconds. Another drawback is that
the inclusion of ratings would interrupt the ●● Proper assessment of the clinical symptoms
mental functions one desires to investigate. and signs related to pain and somatosensory
For example, for a study that requires subjects disturbance is the foundation for neuroim-
to keep a steady emotional experience, the aging research.
behaviour of rating may interrupt their ongo- ●● The major purpose of psychophysics is to
ing mood. In some neuroimaging studies, the establish a psychophysical function that
assessment of pain is separated from scan- depicts the relationship between physical
ning. In other words, the period of imaging stimuli and sensory experience.
only consists of the stimulation phase (and ●● The primary consideration of choosing a
the baseline phase for control), and pain is pain scale is the subjects’ ability to respond
assessed beyond the scanning period or out- and the clinical scenarios where assessment
side the scanner. For example, Gustin et al. is conducted. A scale used for unsuitable
(2012) studied the experience of neuropathic subjects or in a suboptimal situation will
pain in patients with trigeminal neuropathic greatly compromise the validity of the
disorders. When the patients were receiving assessment.
an fMRI scan, the experimenters triggered ●● To assess the pain-related experience, rather
somatosensory processing by directly brush- than pain intensity per se, is an essential part
ing the orofacial sites, such as the mechanical of the assessment of orofacial pain. Because
stimuli of the QST. During the stimulation pain is defined as a multi-facet experience,
phase, brushing stimuli were applied with a which includes sensory and emotional experi-
plastic brush with a speed of two strokes per ences, only assessing patients’ pain intensity
second on the lower lip and the fingers of may not provide a full picture of their
subjects (Gustin et al. 2012). In contrast, the suffering.
ongoing facial pain of the patients was ●● In terms of neuroimaging research on pain
assessed before the scan began. The pain rat- and somatosensory experience, the experi-
ing, which may represent individual differ- mental design can be categorized based on
ences in clinical symptoms, can be analyzed the way to explore the stimulus–response
for its correlation with the brain activation relationship, which includes (i) a concurrent
acquired (during the stimulation phase) recording of brain signals and pain/sensory
(Figure 3.3c). ratings during stimulation and (ii) separate
The same approach is common in research (‘off-line’) record of brain features and pain/
on structural brain features. Structural sensory ratings.
MMSE is a screening tool, its score should be symptoms also show high comorbidity with
carefully interpreted according to subjects’ age chronic pain, including the pain related to
and years of education. A study of the popula- TMD (Reiter et al. 2015). Assessment of these
tion norms in the United States revealed that, emotional experiences of dental patients may
regardless of age, the median MMSE score var- play a key role in pain and behavioural man-
ies from 29 for the subjects with at least nine agement of dental patients.
years of education to 22 for those with zero to
four years of education (Crum et al. 1993). The 3.3.3.1 Assessments Related to Anxiety
MMSE score may show little variation for The State-Trait Anxiety Inventory (STAI) is
subjects with a higher level of education. one of the most widely used assessments for
Therefore, it may not be sensitive to detecting assessing anxiety. The STAI consists of
cognitive impairment in this group of individ- 40 items, with 20 for SA and 20 for trait anx-
uals (Crum et al. 1993). iety (Spielberger et al. 1983). Here SA refers
to the anxiety one is feeling at the present
moment (Lin et al. 2017b). A heightened SA
3.3.2.2 The Montreal Cognitive Assessment
is usually associated with a specific spatial
Like the MMSE, the Montreal Cognitive
and temporal setting, such as ‘in a dental
Assessment (MoCA) is a tool developed for
clinic,’ and anxiety becomes diminished
screening the risk of cognitive impairment.
when individuals are disengaged with the
It consists of the items from several cogni-
settings. For example, individuals may feel
tive domains also included in the MMSE,
strong anxiety when visiting a dentist but
such as the recall of short-term memory,
not a pharmacist. Even in the dental clinic,
attention and language (Nasreddine
some individuals may feel a higher SA upon
et al. 2005). In addition, the MoCA consists
receiving an anaesthetic injection, while
of the clock-drawing task (CDT) for assess-
others may feel a higher SA upon receiving
ing visuospatial abilities. Deficits of visuos-
tooth drilling. In contrast, higher trait anxi-
patial functions may be an early sign of
ety (TA) is associated with a personal dispo-
dementia (Agrell and Dehlin 2012).
sition of feeling anxious. Such a disposition
Compared to the MMSE, the MoCA may be
or tendency may not necessarily be state-
a better tool for screening mild cognitive
specific. In the case of dental visiting,
impairment (MCI), evidenced by a recent
i ndividuals with a higher TA are prone to
meta-analysis (Ciesielska et al. 2016).
feel anxiety regardless of the stage of treat-
ment. They may feel strong anxiety when
they anticipate ‘something bad’ would hap-
3.3.3 Assessments for Patients’
pen, regardless of visiting a dentist or a
Emotional Experience
pharmacist.
Anxiety is the emotion related to the percep- In addition to the STAI, the seven-item
tion of a threat. It refers to a future-oriented Generalized Anxiety Disorder (GAD-7) ques-
emotional state associated with an antici- tionnaire is widely used for assessing the sever-
pated threat (Keogh and Asmundson 2004). ity of GAD (Spitzer et al. 2006). Instead of
Anxiety is closely associated with pain and rating the intensity of the anxiety-related
avoidance of dental treatment symptoms, subjects are asked to rate the fre-
(Armfield 2013). Depressive symptoms are quency that they have been bothered by the
the key elements of major depressive disor- symptoms described in each item. For exam-
ders, a mental disorder with high prevalence ple, subjects need to rate how often they expe-
in adults (Ferrari et al. 2013). Depressive rience excessive worry or easily feel annoyed.
The GAD-7 is also recommended as the diag- 3.3.4.1 The Fear of Pain Questionnaire-III
nostic instrument for Axis II of the DC/TMD The fear of pain questionnaire-III (FPQ-III)
(Schiffman et al. 2014). consists of 30 items that assess the fear of
pain in three domains: severe pain, minor
3.3.3.2 Assessments Related pain and medical pain. A higher score of the
to Depressive Symptoms FPQ-III is associated with greater avoidance
The Beck Depression Inventory (BDI-II) (Beck of pain (McNeil and Rainwater 3rd 1998).
et al. 1996) is the revision of the original BDI, Patients with chronic pain reported the great-
which has been widely used for assessing the est fear of pain in the domain of severe pain
symptoms related to depression since the (McNeil and Rainwater 3rd 1998). The FPQ-
1960s. The BDI-II is a self-report questionnaire III also revealed good reliability and validity
that consists of 21 items focusing on various in a non-clinical sample (Osman et al. 2002).
aspects of depressive symptoms, including Clinically, orofacial pain patients showed a
emotional experience (e.g. feeling sad) and greater fear of severe pain compared to con-
behavioural changes (e.g. loss of sleep and trol subjects. A short-form version of the
appetite). Subjects are asked to rate how FPQ-III that consists of 20 items is published.
intense the symptoms are, according to their In the FPQ-SF, four domains of fear of pain,
experience. including severe pain, minor pain, injection
Another common tool for assessing depres- and dental pain, are identified (Asmundson
sive symptoms is the nine-item Patient Health et al. 2008). The score of fear of pain was
Questionnaire (PHQ-9) (Kroenke et al. 2001). associated with dental fear (McNeil
The PHQ-9 was designed based on DSM-IV et al. 2001).
and used for quantifying the severity of depres-
sion. It also includes items assessing for emo-
tional experience (e.g. feeling hopeless) and 3.3.4.2 The Pain Catastrophizing Scale
behavioural changes (e.g. feeling difficulty in The pain catastrophizing scale (PCS) assesses
concentrating on things). Just like the GAD-7, the psychological construct of catastrophiz-
the PHQ-9 is also recommended as the diag- ing, which refers to ‘an exaggerated negative
nostic instrument for Axis II of the DC/TMD mental set brought to bear during painful
(Schiffman et al. 2014). experiences’ (Sullivan et al. 2001). Such a
mindset can be characterized by three domains
of catastrophic thoughts: rumination, magni-
3.3.4 Assessments for Pain-related
fication and helplessness (Sullivan et al. 1995).
Cognitive–Affective Experience
The PCS consists of 13 items focusing on these
As discussed in Section 3.2, a critical aspect of three domains (Sullivan et al. 1995). The PCS
pain assessment is to evaluate patients’ emo- assesses one’s frequency of generating cata-
tional experience related to pain. The cogni- strophic thoughts about pain. The assessment
tive–affective dimension of pain is relatively emphasizes more on the cognitive processing
ignored in research on orofacial pain. related to pain (e.g. feeling over-attentive to a
Assessment of the emotional experience, such threat and the inability to cope with it) than
as fear of pain, and the cognitive aspect related on the emotional experience of pain (i.e. fear/
to pain, such as catastrophic thoughts about anxiety). A higher PCS score is associated with
pain, are usually performed using more sophis- pain related to dental conditions, such as
ticated questionnaires, as outlined in the fol- patients’ experience of dental scaling and the
lowing sections. clinical pain of TMD (Lin 2013).
3.3.6 Summary
Further Readings
●● To study the brain mechanisms related to the
cognitive–affective factors of the dental Please see the Companion Website for
practice, it is critical to quantify these Suggested Readings.
References
Agrell, B. and Dehlin, O. (2012). The clock- the study of clinical pain states. Neurosci. Lett.
drawing test. 1998. Age Ageing 41 (Suppl. 3): 299: 57–60.
iii41–iii45. Aragón, F., Zea-Sevilla, M.A., Montero, J. et al.
Alghadir, A.H., Anwer, S., Iqbal, A., and Iqbal, (2018). Oral health in Alzheimer’s disease: a
Z.A. (2018). Test–retest reliability, validity, and multicenter case-control study. Clin. Oral
minimum detectable change of visual analog, Investig. 22: 3061–3070.
numerical rating, and verbal rating scales for Aras, K., Hasanreisoğlu, U., and Shinogaya,
measurement of osteoarthritic knee pain. T. (2009). Masticatory performance, maximum
J. Pain Res. 11: 851–856. occlusal force, and occlusal contact area in
Al-Sahan, M.M., Macentee, M.I., and Bryant, patients with bilaterally missing molars and
S.R. (2020). A metatheory explaining how distal extension removable partial dentures.
patients manage tooth loss. Gerodontology 37: Int. J. Prosthodont. 22: 204–209.
258–270. Armfield, J.M. (2013). Predicting dental
Apkarian, A.V., Krauss, B.R., Fredrickson, B.E., avoidance among dentally fearful Australian
and Szeverenyi, N.M. (2001). Imaging the pain adults. Eur. J. Oral Sci. 121: 240–246.
of low back pain: functional magnetic Asmundson, G.J., Bovell, C.V., Carleton, R.N.,
resonance imaging in combination with and McWilliams, L.A. (2008). The fear of pain
monitoring subjective pain perception allows questionnaire-short form (FPQ-SF): factorial
validity and psychometric properties. Pain of the diagnostic accuracy of the water
134: 51–58. swallow test for screening aspiration in stroke
Baliki, M.N., Chialvo, D.R., Geha, P.Y. et al. patients. J. Adv. Nurs. 72: 2575–2586.
(2006). Chronic pain and the emotional brain: Ciesielska, N., Sokołowski, R., Mazur, E. et al.
specific brain activity associated with (2016). Is the Montreal cognitive assessment
spontaneous fluctuations of intensity of (MoCA) test better suited than the mini-
chronic back pain. J. Neurosci. 26: mental state examination (MMSE) in mild
12165–12173. cognitive impairment (MCI) detection among
Baliki, M.N., Geha, P.Y., and Apkarian, people aged over 60? Meta-analysis. Psychiatr.
A.V. (2009). Parsing pain perception between Pol. 50: 1039–1052.
nociceptive representation and magnitude Crum, R.M., Anthony, J.C., Bassett, S.S., and
estimation. J. Neurophysiol. 101: 875–887. Folstein, M.F. (1993). Population-based norms
Beck, A.T., Steer, R.A., Ball, R., and Ranieri, for the mini-mental state examination by age
W. (1996). Comparison of Beck depression and educational level. JAMA 269: 2386–2391.
inventories -IA and -II in psychiatric Cusson, V., Caron, C., Gaudreau, P. et al. (2015).
outpatients. J. Pers. Assess. 67: 588–597. Assessing older Adults’ masticatory efficiency.
Belafsky, P.C., Mouadeb, D.A., Rees, C.J. et al. J. Am. Geriatr. Soc. 63: 1192–1196.
(2008). Validity and reliability of the eating Eliav, E., Gracely, R.H., Nahlieli, O., and
assessment tool (EAT-10). Ann. Otol. Rhinol. Benoliel, R. (2004). Quantitative sensory
Laryngol. 117: 919–924. testing in trigeminal nerve damage
Bijur, P.E., Silver, W., and Gallagher, E.J. (2001). assessment. J. Orofac. Pain 18: 339–344.
Reliability of the visual analog scale for Engelen, L., Van Der Bilt, A., and Bosman,
measurement of acute pain. Acad. Emerg. F. (2004). Relationship between oral
Med. 8: 1153–1157. sensitivity and masticatory performance.
Bijur, P.E., Latimer, C.T., and Gallagher, J. Dent. Res. 83: 388–392.
E.J. (2003). Validation of a verbally Ertekin, C. (2011). Voluntary versus spontaneous
administered numerical rating scale of acute swallowing in man. Dysphagia 26: 183–192.
pain for use in the emergency department. Ferrari, A.J., Somerville, A.J., Baxter, A.J. et al.
Acad. Emerg. Med. 10: 390–392. (2013). Global variation in the prevalence and
Bourdiol, P., Hennequin, M., Peyron, M.A., and incidence of major depressive disorder: a
Woda, A. (2020). Masticatory adaptation to systematic review of the epidemiological
occlusal changes. Front. Physiol. 11: 263. literature. Psychol. Med. 43: 471–481.
Brodsky, M.B., Suiter, D.M., González- Ferreira-Valente, M.A., Pais-Ribeiro, J.L., and
Fernández, M. et al. (2016). Screening Jensen, M.P. (2011). Validity of four pain
accuracy for aspiration using bedside water intensity rating scales. Pain 152: 2399–2404.
swallow tests: a systematic review and Ferro, K.J., Morgano, S.M., Driscoll, C.F. et al.
meta-analysis. Chest 150: 148–163. (2017). The glossary of prosthodontic terms.
Campos, C.H., Ribeiro, G.R., Costa, J.L., and J. Prosthet. Dent. 117: e1–e105.
Rodrigues Garcia, R.C. (2017). Correlation of Folstein, M.F., Folstein, S.E., and McHugh,
cognitive and masticatory function in P.R. (1975). ‘A mini-mental state’. A practical
Alzheimer’s disease. Clin. Oral Investig. 21: method for grading the cognitive state of
573–578. patients for the clinician. J. Psychiatr. Res. 12:
Chapman, C.R. (1978). Pain: the perception of 189–198.
noxious events. In: The Psychology of Pain (ed. Fujii, R., Takahashi, T., Toyomura, A. et al.
R.A. Sternbach). New York: Raven Press. (2011). Comparison of cerebral activation
Chen, P.C., Chuang, C.H., Leong, C.P. et al. involved in oral and manual stereognosis.
(2016). Systematic review and meta-analysis J. Clin. Neurosci. 18: 1520–1523.
Galo, R., Vitti, M., Mattos, M.D.G.C., and Regalo, IASP. (2020). IASP Terminology [Online].
S.C.H. (2007). Masticatory muscular Available: https://www.iasp-pain.org/
activation in elderly individuals during terminology?navItemNumber=576
chewing. Gerodontology 24: 244–248. [Accessed].
Geber, C., Klein, T., Azad, S. et al. (2011). Ikebe, K., Amemiya, M., Morii, K. et al. (2007a).
Test–retest and interobserver reliability of Comparison of oral stereognosis in relation to
quantitative sensory testing according to the age and the use of complete dentures. J. Oral
protocol of the German research Network on Rehabil. 34: 345–350.
neuropathic pain (DFNS): a multi-Centre Ikebe, K., Amemiya, M., Morii, K. et al. (2007b).
study. Pain 152: 548–556. Association between oral stereognostic ability
Gracely, R.H. and Dubner, R. (1981). Pain and masticatory performance in aged
assessment in humans --a reply to hall. Pain complete denture wearers. Int. J. Prosthodont.
11: 109–120. 20: 245–250.
Gustin, S.M., Peck, C.C., Cheney, L.B. et al. Ikebe, K., Morii, K., Matsuda, K., and Nokubi,
(2012). Pain and plasticity: is chronic pain T. (2007c). Discrepancy between satisfaction
always associated with somatosensory cortex with mastication, food acceptability, and
activity and reorganization? J. Neurosci. 32: masticatory performance in older adults. Int.
14874–14884. J. Prosthodont. 20: 161–167.
Halazonetis, D.J., Schimmel, M., Antonarakis, Ikebe, K., Matsuda, K., Murai, S. et al. (2010).
G.S., and Christou, P. (2013). Novel software for Validation of the Eichner index in relation to
quantitative evaluation and graphical occlusal force and masticatory performance.
representation of masticatory efficiency. J. Oral Int. J. Prosthodont. 23: 521–524.
Rehabil. 40: 329–335. Ikebe, K., Matsuda, K., Kagawa, R. et al. (2012).
Hama, Y., Kanazawa, M., Minakuchi, S. et al. Masticatory performance in older subjects
(2014). Properties of a color-changeable with varying degrees of tooth loss. J. Dent.
chewing gum used to evaluate masticatory 40: 71–76.
performance. J. Prosthodont. Res. 58: Jacobs, R., Bou Serhal, C., and van Steenberghe,
102–106. D. (1998). Oral stereognosis: a review of the
Hayakawa, I., Watanabe, I., Hirano, S. et al. (1998). literature. Clin. Oral Investig. 2: 3–10.
A simple method for evaluating masticatory Jensen, M.P. (2003). The validity and reliability
performance using a color-changeable chewing of pain measures in adults with cancer. J. Pain
gum. Int. J. Prosthodont. 11: 173–176. 4: 2–21.
Herr, K.A. and Mobily, P.R. (1993). Comparison Kawagishi, S., Kou, F., Yoshino, K. et al. (2009).
of selected pain assessment tools for use with Decrease in stereognostic ability of the tongue
the elderly. Appl. Nurs. Res. 6: 39–46. with age. J. Oral Rehabil. 36: 872–879.
Hicks, C.L., Von Baeyer, C.L., Spafford, P.A. et al. Keogh, E. and Asmundson, G.J. (2004). Negative
(2001). The faces pain scale-revised: toward a affectivity, catastrophising and anxiety
common metric in pediatric pain sensitivity. In: Understanding and Treating
measurement. Pain 93: 173–183. Fear of Pain (eds. G.J. Asmundson, J. Vlaeyen
Hirano, K., Hirano, S., and Hayakawa, I. (2004). and G. Crombez). USA: Oxford
The role of oral sensorimotor function in University Press.
masticatory ability. J. Oral Rehabil. 31: Kikutani, T., Tamura, F., Nishiwaki, K. et al.
199–205. (2009). Oral motor function and masticatory
Hsu, K.J., Lee, H.E., Lan, S.J. et al. (2012). performance in the community-dwelling
Evaluation of a self-assessed screening test for elderly. Odontology 97: 38–42.
masticatory ability of Taiwanese older adults. Kim, E.K., Lee, S.K., Choi, Y.H. et al. (2017).
Gerodontology 29: e1113–e1120. Relationship between chewing ability and
cognitive impairment in the rural elderly. Lin, C.S., Wu, C.Y., Wang, D.H. et al. (2019).
Arch. Gerontol. Geriatr. 70: 209–213. Brain signatures associated with swallowing
Koshino, H., Hirai, T., Toyoshita, Y. et al. (2008). efficiency in older people. Exp. Gerontol.
Development of new food intake 115: 1–8.
questionnaire method for evaluating the Lo, K.C., Lin, H.H., and Lin, C.S. (2020). A novel
ability of mastication in complete denture method for assessing oral mixing ability based
wearers. Prosthodon. Res. Pract. 7: 12–18. on the spatial clusters quantified by
Kroenke, K., Spitzer, R.L., and Williams, variogram. J. Oral Rehabil. 47: 951–960.
J.B. (2001). The PHQ-9: validity of a brief Lobbezoo, F., Weijenberg, R.A., Scherder,
depression severity measure. J. Gen. Intern. E.J. (2011). Topical review: orofacial pain in
Med. 16: 606–613. dementia patients. A diagnostic challenge.
Kugimiya, Y., Ueda, T., Watanabe, Y. et al. J. Orofac. Pain. 25: 6–14.
(2019). Relationship between mild cognitive Lobbezoo, F., Delwel, S., Weijenberg, R.A.F. et al.
decline and oral motor functions in (2017). Orofacial Pain and Mastication in
metropolitan community-dwelling older Dementia. Curr. Alzheimer. Res. 14: 506–511.
Japanese: the Takashimadaira study. Arch. Luraschi, J., Korgaonkar, M.S., Whittle, T. et al.
Gerontol. Geriatr. 81: 53–58. (2013). Neuroplasticity in the adaptation to
Leow, L.P., Huckabee, M.L., Anderson, T., and prosthodontic treatment. J. Orofac. Pain 27:
Beckert, L. (2010). The impact of dysphagia on 206–216.
quality of life in ageing and Parkinson’s Manor, Y., Giladi, N., Cohen, A. et al. (2007).
disease as measured by the swallowing quality Validation of a swallowing disturbance
of life (SWAL-QOL) questionnaire. Dysphagia questionnaire for detecting dysphagia in
25: 216–220. patients with Parkinson’s disease. Mov. Disord.
Liang, S., Zhang, Q., Witter, D.J. et al. (2015). 22: 1917–1921.
Effects of removable dental prostheses on McCracken, L.M. (1997). ‘Attention’ to pain in
masticatory performance of subjects with persons with chronic pain: a behavioral
shortened dental arches: a systematic review. approach. Behav. Res. Ther.: 271–284.
J. Dent. 43: 1185–1194. McHorney, C.A., Robbins, J., Lomax, K. et al.
Lin, C.S. (2013). Pain catastrophizing in dental (2002). The SWAL-QOL and SWAL-CARE
patients: implications for treatment outcomes tool for oropharyngeal dysphagia in
management. J. Am. Dent. Assoc. 144: 1244–1251. adults: III. Documentation of reliability and
Lin, C.S. (2019). Functional adaptation of validity. Dysphagia 17: 97–114.
Oromotor functions and aging: a focused review McNeil, D.W. and Rainwater, A.J. 3rd (1998).
of the evidence from brain neuroimaging Development of the fear of pain
research. Front. Aging Neurosci. 11: 354. questionnaire--III. J. Behav. Med. 21: 389–410.
Lin, C.S., Niddam, D.M., Hsu, M.L., and Hsieh, McNeil, D.W., Au, A.R., Zvolensky, M.J. et al.
J.C. (2013). Pain catastrophizing is associated (2001). Fear of pain in orofacial pain patients.
with dental pain in a stressful context. J. Dent. Pain 89: 245–252.
Res. 92: 130–135. Melzack, R. (1987). The short-form McGill pain
Lin, C.S., Wu, C.Y., Wu, S.Y. et al. (2017a). questionnaire. Pain 30: 191–197.
Age-and sex-related differences in masseter Misch, C.E. and Resnik, R. (2010). Mandibular
size and its role in oral functions. J. Am. Dent. nerve neurosensory impairment after dental
Assoc. 148: 644–653. implant surgery: management and protocol.
Lin, C.S., Wu, S.Y., and Yi, C.A. (2017b). Implant. Dent. 19: 378–386.
Association between anxiety and pain in Miura, H., Araki, Y., Hirai, T. et al. (1998).
dental treatment: a systematic review and Evaluation of chewing activity in the elderly
meta-analysis. J. Dent. Res. 96: 153–162. person. J. Oral Rehabil. 25: 190–193.
Moayedi, M. and Weissman-Fogel, I. (2009). Is bite force and masticatory muscles thickness.
the insula the ‘how much’ intensity coder? Arch. Oral Biol. 55: 797–802.
J. Neurophysiol. 102: 1345–1347. Perez-Lloret, S., Ciampi De Andrade, D., Lyons,
Moore, R., Brødsgaard, I., and Abrahamsen, K.E. et al. (2016). Rating scales for pain in
R. (2002). A 3-year comparison of dental Parkinson’s disease: critique and
anxiety treatment outcomes: hypnosis, group recommendations. Mov. Disord. Clin. Pract.
therapy and individual desensitization vs. no 3: 527–537.
specialist treatment. Eur. J. Oral Sci. 110: Persson, E., Wårdh, I., and Östberg, P. (2019).
287–295. Repetitive saliva swallowing test: norms,
Moore, J.D., Kleinfeld, D., and Wang, F. (2014). clinical relevance and the impact of saliva
How the brainstem controls orofacial secretion. Dysphagia 34: 271–278.
behaviors comprised of rhythmic actions. Peyron, M.A., Blanc, O., Lund, J.P., and Woda,
Trends Neurosci. 37: 370–380. A. (2004a). Influence of age on adaptability of
Müller, F., Duvernay, E., Loup, A. et al. (2013). human mastication. J. Neurophysiol. 92:
Implant-supported mandibular overdentures 773–779.
in very old adults: a randomized controlled Peyron, M.A., Mishellany, A., and woda,
trial. J. Dent. Res. 92: 154S–160S. A. (2004b). Particle size distribution of food
Nasreddine, Z.S., Phillips, N.A., Bedirian, boluses after mastication of six natural foods.
V. et al. (2005). The Montreal cognitive J. Dent. Res. 83: 578–582.
assessment, MoCA: a brief screening tool for Pfau, D.B., Rolke, R., Nickel, R. et al. (2009).
mild cognitive impairment. J. Am. Geriatr. Soc. Somatosensory profiles in subgroups of
53: 695–699. patients with myogenic temporomandibular
Nokubi, T., Nokubi, F., Yoshimuta, Y. et al. disorders and fibromyalgia syndrome. Pain
(2010). Measuring masticatory performance 147: 72–83.
using a new device and beta-carotene in Pigg, M., Baad-Hansen, L., Svensson, P. et al.
test gummy jelly. J. Oral Rehabil. 37: (2010). Reliability of intraoral quantitative
820–826. sensory testing (QST). Pain 148: 220–226.
Oguchi, K., Saitoh, E., Baba, M. et al. (2000a). Poluha, R.L., De La Torre Canales, G.,
The repetitive saliva swallowing test (RSST) as Bonjardim, L.R., and Conti, P.C.R. (2020).
a screening test of functional dysphagia (2) Somatosensory and psychosocial profile of
validity of RSST. Jpn. J. Rehabil. Med. 37: patients with painful temporomandibular
383–388. joint clicking. J. Oral Rehabil. 11:
Oguchi, K., Saitoh, E., Mizuno, M. et al. (2000b). 1346–1357.
The repetitive saliva swallowing test (RSST) as Porporatti, A.L., Costa, Y.M., Stuginski-Barbosa,
a screening test of functional dysphagia (1) J. et al. (2015). Quantitative methods for
Normal values of RSST. Jpn. J. Rehabil. Med. somatosensory evaluation in atypical
37: 375–382. odontalgia. Braz. Oral Res. 29, 1: –7.
Ogura, R., Kato, H., Okada, D. et al. (2012). The Price, D.D., McGrath, P.A., Rafii, A., and
relationship between bite force and oral Buckingham, B. (1983). The validation of
sensation during biting in molars. Aust. Dent. visual analogue scales as ratio scale measures
J. 57: 292–299. for chronic and experimental pain. Pain
Osman, A., Breitenstein, J.L., Barrios, F.X. et al. 17: 45–56.
(2002). The fear of pain questionnaire-III: Price, D.D., Bush, F.M., Long, S., and Harkins,
further reliability and validity with nonclinical S.W. (1994). A comparison of pain
samples. J. Behav. Med. 25: 155–173. measurement characteristics of mechanical
Palinkas, M., Nassar, M.S., Cecílio, F.A. et al. visual analogue and simple numerical rating
(2010). Age and gender influence on maximal scales. Pain 56: 217–226.
Prinz, J.F. and Lucas, P.W. (1995). Swallow generalized anxiety disorder: the GAD-7.
thresholds in human mastication. Arch. Oral Arch. Intern. Med. 166: 1092–1097.
Biol. 40: 401–403. Sullivan, M.J.L., Bishop, S.R., and Pivik,
Raja, S.N., Carr, D.B., Cohen, M. et al. (2020). The J. (1995). The pain catastrophizing scale:
revised International Association for the Study development and validation. Psychol. Assess.
of Pain definition of pain: concepts, challenges, 7: 524–532.
and compromises. Pain 161: 1976–1982. Sullivan, M.J., Thorn, B., Haythornthwaite,
Reiter, S., Emodi-Perlman, A., Goldsmith, J.A. et al. (2001). Theoretical perspectives on
C. et al. (2015). Comorbidity between the relation between catastrophizing and pain.
depression and anxiety in patients with Clin. J. Pain 17: 52–64.
temporomandibular disorders according to Tammaro, S., Berggren, U., and Bergenholtz,
the research diagnostic criteria for G. (1997). Representation of verbal pain
temporomandibular disorders. J. Oral Facial descriptors on a visual analogue scale by
Pain Headache 29: 135–143. dental patients and dental students. Eur.
Rolke, R., Baron, R., Maier, C. et al. (2006). J. Oral Sci. 105: 207–212.
Quantitative sensory testing in the German Taneja, P., Olausson, H., Trulsson, M. et al.
research Network on neuropathic pain (2019). Assessment of experimental orofacial
(DFNS): standardized protocol and reference pain, pleasantness and unpleasantness via
values. Pain 123: 231–243. standardized psychophysical testing. Eur.
Salazar, S., Hori, K., Uehara, F. et al. (2020). J. Pain 23: 1297–1308.
Masticatory performance analysis using Tarkowska, A., Katzer, L., and Ahlers,
photographic image of gummy jelly. M.O. (2017). Assessment of masticatory
J. Prosthodont. Res. 64: 48–54. performance by means of a color-changeable
Schiffman, E., Ohrbach, R., Truelove, E. et al. chewing gum. J. Prosthodont. Res. 61: 9–19.
(2014). Diagnostic criteria for Tombaugh, T.N. and McIntyre, N.J. (1992).
temporomandibular disorders (DC/TMD) The mini-mental state examination: a
for clinical and research applications: comprehensive review. J. Am. Geriatr. Soc.
recommendations of the International RDC/ 40: 922–935.
TMD Consortium Network and Orofacial Pain Trulsson, M., Van Der Bilt, A., Carlsson,
Special Interest Group. J. Oral Facial Pain G.E. et al. (2012). From brain to bridge:
Headache 28: 6–27. masticatory function and dental implants.
Schimmel, M., Christou, P., Herrmann, F., and J. Oral Rehabil. 39: 858–877.
Müller, F. (2007). A two-colour chewing gum Tsuga, K., Carlsson, G.E., Osterberg, T., and
test for masticatory efficiency: development of Karlsson, S. (1998). Self-assessed masticatory
different assessment methods. J. Oral Rehabil. ability in relation to maximal bite force and
34: 671–678. dental state in 80-year-old subjects. J. Oral
Silva, L.C., Nogueira, T.E., Rios, L.F. et al. (2018). Rehabil. 25: 117–124.
Reliability of a two-colour chewing gum test Tu, T.T.H., Miura, A., Shinohara, Y. et al. (2019).
to assess masticatory performance in complete Pharmacotherapeutic outcomes in atypical
denture wearers. J. Oral Rehabil. 45: 301–307. odontalgia: determinants of pain relief. J. Pain
Spielberger, C.D., Gorsuch, R.L., Lushene, Res. 12: 831–839.
R. et al. (1983). Manual for the State-Trait Van Damme, S., Crombez, G., Eccleston, C., and
Anxiety Inventory. Palo Alto, CA: Consulting Roelofs, J. (2004). The role of hypervigilance
Psychologists Press. in the experience of pain. In: Understanding
Spitzer, R.L., Kroenke, K., Williams, J.B., and and Treating Fear of Pain (eds.
Lowe, B. (2006). A brief measure for assessing G.J.G. Asmundson, J.W.S. Vlaeyen and
Part II
Dental Neuroimaging: The Role of the Brain in Oral Functions, First Edition. Chia-Shu Lin.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
Companion website: www.wiley.com/go/lin/dental-neuroimaging
Notes: ALE: activation likelihood estimation; EEG: electroencephalography; fMRI: functional magnetic resonance
imaging; MEG: magnetoencephalography; MRI: magnetic resonance imaging; MRS: magnetic resonance spectroscopy;
PET: positron emission tomography; TMS: transcranial magnetic stimulation.
studies (Table 4.1) of a variety of movements brain activation of the M1 during chewing and
and acts. (see Box 4.1 in the Companion clenching (Lin 2018) and drawing and writing
Website) The key brain regions identified for (Yuan and Brown 2015) (Table 4.1). Recent
the control system are discussed as follows. neuroimaging studies have disclosed a diverse
role of the M1 in motor control (Table 4.1).
4.1.3.1 Primary Motor Cortex (M1) Notably, the M1 activation is also associated
Traditionally, the M1 has been considered as with the condition when sensory processing
the major component of cortical motor areas, plays a dominant role, such as the task of pas-
which consists of the M1, the posterior pari- sive music listening (Gordon et al. 2018) and
etal cortex, the primary somatosensory cortex patients’ experience of chronic pain (Chang
(S1), the supplementary motor area (SMA) et al. 2018) (Table 4.1). When comparing
and the premotor cortex (PMC) (Figure 4.1). motor execution and motor imagery (i.e. men-
The M1 directly controls the motor system via tally rehearsing a movement without actually
the corticospinal pathway with the somato- executing it), researchers found that the M1
topic organization that corresponds to our played a dominant role during execution but
body. Therefore, the M1 is considered a critical not imagery. In contrast, the SMA and the
element for movement execution. As pre- PMC played a dominant role during motor
dicted, the M1 activation is associated with the imagery or observation (Filgueiras et al. 2018;
execution of a movement. This is evidenced by Hardwick et al. 2018; Winlove et al. 2018).
functional magnetic resonance imaging However, this distinction is not clear-cut. For
(fMRI) studies, which consistently showed the example, during motor execution, the brain
Figure 4.1 An overview of brain regions associated with motor control. The figure only displays the
relative position and size of the brain regions, not depicting the anatomical details.
activation of the SMA, the PMC and the r otation. Furthermore, Hirose et al. (2018)
M1 was noted (Yuan and Brown 2015; reported the brain activation of a finger-
Lin 2018); during the preparatory phase, the tapping task, in which the subjects were free
brain activation of the M1, the SMA and the to choose which finger would tap. They found
PMC can collectively predict the subsequent that before tapping (i.e. during the prepara-
movement (Hirose et al. 2018). The findings tory stage), the brain activation of the M1, the
suggest that the M1 forms a complicated net- SMA and the PMC (the PMd subregion, see
work of sensorimotor processing with other below) collectively predicted the choice of
components of the motor system. which finger to tap (i.e. during the execution
The interaction between the M1 and other stage). The findings revealed that the M1 may
components of the motor system is further participate in planning a movement. In sum,
demonstrated by two recent studies using the recent findings suggest that the M1 plays
multivariate pattern analysis (MVPA). Park a predominant role in motor execution but is
et al. (2015) reported that when subjects were not limited to execution per se. The M1 may
performing a motor execution or imagery also work closely with the SMA/the PMC
task (for hand grasping and hand rotation), during the preparatory stage.
brain activation at the M1, the PMC and the
SMA were consistently found in all experi- 4.1.3.2 Premotor Cortex
mental tasks. The MVPA revealed that the The human PMC resides in the dorsolateral
activation of the M1 predicted the executed area of the brain, just anterior to the M1
movement with the best accuracy, far beyond (Figure 4.1). The PMC is further differentiated
the SMA/the PMC. However, during the in the more dorsal part, the PMd, and the more
imagery task, the three regions showed a sim- ventral part, the PMv. Both animal and human
ilar prediction accuracy to imagined move- neuroimaging findings suggest that the rostral
ment in hand grasping, and the SMA showed and caudal parts of the PMd are different
the highest prediction accuracy to hand structurally and functionally (Picard and
Strick 2001). Anatomically, the caudal PMd et al. 2015) (Table 4.1). Re-organization of the
has projections to the M1 and the spinal cord, PMC is also observed in post-stroke patients,
with a close association with the motor area. In which can be associated with the recovery of
contrast, the rostral PMd does not show projec- motor functions in the patients (Kantak
tion to the M1. Instead, it is functional and et al. 2012).
structurally more associated with the PFC
(Picard and Strick 2001). 4.1.3.3 Supplementary Motor Area
Functionally, the PMC activation has been The human SMA resides in the dorsal medial
identified in various stages of motor control, wall of the brain, just anterior to the M1
not confined to the period of movement execu- (Figure 4.1). The SMA can be further differen-
tion. For example, the PMC activation was tiated into the pre-SMA and the SMA proper.
identified during dual tasking (i.e. performing The pre-SMA resides posterior to the
two tasks concurrently) and task switching PFC. While the SMA directly connects to the
(i.e. performing two tasks concurrently alter- M1 and the spinal cord, the pre-SMA connects
natingly) (Worringer et al. 2019) and during with the PFC (Picard and Strick 2001). The
visual imagery (Filgueiras et al. 2018; Winlove pre-SMA is only sparsely connected to the cor-
et al. 2018). Its activation was identified when ticospinal system (Nachev et al. 2008). The
one was imagining doing a movement or SMA also holds the somatotopic organization,
observing the other’s movement (Hardwick i.e. the movement of a specific body part will
et al. 2018). The PMC activation is associated correspond to a specific region of the SMA
with the processing of using a tool (Ishibashi (Nachev et al. 2008). The functional heteroge-
et al. 2016) and encoding short sequential ele- neity between the SMA proper and the pre-
ments for a movement (Diedrichsen and SMA can be identified in neuroimaging
Kornysheva 2015). Notably, the PMC activa- research: activation of the pre-SMA is more
tion is also identified during a sensory task associated with the cognitive processing of
(Gordon et al. 2018), suggesting its role in inte- motor action, which is also associated with
grating sensory and motor information. In an the PFC.
earlier study of 12 young adults, the research- Both the pre-SMA and the SMA show some
ers asked subjects to attend to moving objects similar functions in motor control. For exam-
with different focuses: to focus on the moving ple, according to the results from imaging
object per se, to focus on the spatial feature of meta-analyses, the activation of the SMA has
the movement and to focus on the temporal been identified for executing an action, e.g.
feature of the movement (Schubotz and Von chewing and drawing/writing (Yuan and
Cramon 2001). The results showed that the Brown 2015; Lin 2018), together with the M1
activation of the left superior ventrolateral and the PMC. Moreover, just like the PMC, the
PMC was stronger when focusing on the SMA activation is consistently found during
object, and the activation of the dorsolateral planning and control of a movement (Teghil
PMC was greater when focusing on the spatial et al. 2019; Filgueiras et al. 2018). While the
features (Schubotz and Von Cramon 2001). PMC plays a key role in a movement cued
The findings suggest that the PMC plays a key by an external signal, the SMA is associated
role in the processing of the sensory feature with self-initiated movements (Rosenzweig
related to movement. Finally, meta-analytic et al. 2002). The SMA is associated with cogni-
findings of patients with stroke have revealed tive control of movement. Here ‘cognitive con-
that a good motor recovery is associated with trol’ refers to flexibly alter the ongoing action,
the balance of interhemispheric activation of such as stopping the current action or change
the PMC as well as the sensorimotor cortex, from one action to another. For example, ani-
suggesting its role in motor learning (Tang mal research has revealed that the pre-SMA
neurons responded to change of a different from the striatum to the internal segment of the
motor programme (Matsuzaka and Tanji globus pallidus (GPi) and the substantia nigra
1996), and human neuroimaging research also (SNr). The indirect pathway consists of the pro-
reported that pre-SMA activation was associ- jection from the striatum to the external seg-
ated with changing or stopping a current ment of the globus pallidus (GPe). The GPe then
motor task (Nachev et al. 2008). Abnormal sends projections to the GPi and the SNr via the
sensorimotor activation of the SMA may subthalamic nucleus (STN). The BG has an
reflect motor deficits. For example, patients extensive network with the cerebral cortex and
with dystonia, i.e. an abnormal pattern of the cerebellum. It forms the corticostriatal cir-
muscle contraction, showed decreased activa- cuitry: the projections from the striatum to the
tion in the SMA and somatosensory cortex cortex (via thalamus) and the projection from
(Løkkegaard et al. 2016). All the findings high- the cortex to the striatum. Moreover, the STN
light the role of the SMA and the PMC in the neurons send projections to both the motor and
sensorimotor transformation for behavioural non-motor (Crus II) regions of the cerebellum
adaptation. (Bostan and Strick 2018). During motor control,
both the BG and the cerebellum receive and
4.1.3.4 Basal Ganglia send information to the M1 for executing a
The basal ganglia (BG) includes both a direct movement. Moreover, in terms of motor learn-
and an indirect pathway of control (Figure 4.2a). ing, the BG is widely considered a critical com-
The direct pathway consists of the projection ponent of reward-based learning, i.e. the
Figure 4.2 Sensorimotor control of the basal ganglia and the cerebellum. (a) The basal ganglia consist of
a direct and an indirect pathway of motor control. In both pathways, the striatum is activated by the cortex,
which forms a loop of control with the thalamus (the grey arrow). In the direct pathway (the solid black
arrow), the activation of the striatum inhibits the activity of the internal segment of the globus pallidus
(GPi) and the substantia nigra (SNr), which further inhibits thalamic functions. Therefore, cortical activation
is associated with an increased thalamic activity via the direct pathway. In the indirect pathway (the black
dashed line), the activation of the striatum inhibits the activity of the external segment of the globus
pallidus (GPe), which further inhibits the activity of the subthalamic nucleus (STN). Notably, the activity of
the STN further activates GPi/SNr, which decreases thalamic activity. Therefore, cortical activation is
associated with a decrease in thalamic activity via the indirect pathway. (b) The cerebellum plays a key role
as an internal model of motor learning. A forward model predicts the sensory outcomes when motor
commands are executed. It adjusts sensorimotor processing via feedback of the predicted sensory outcomes.
An inverse model calculates the motor commands that would produce the sensory outcomes from desired
actions. According to Wolpert et al. (2001), both models are crucial to motor control.
learning process that optimizes performance via role in the forward and inverse models of
reward signals. The cerebellum, in contrast, motor control. It transforms the desired
plays a key role in error-based supervised learn- movement trajectory to the motor commands
ing (see Section 4.4). required for achieving the goal of movement
(Wolpert et al. 1998) (Figure 4.2b).
4.1.3.5 Cerebellum While the cerebellum is conceived as a cal-
The cerebellum has long been conceived as a culator that estimates the association between
critical component of the motor system. sensory feedback and motor command, it does
Activation of the cerebellum is consistently not mean the cerebellum only works for senso-
found during chewing and drawing/writing rimotor learning. On the contrary, cumulating
(Yuan and Brown 2015; Lin 2018) together neuroimaging evidence has suggested that the
with other brain regions such as the M1 and cerebellum plays a key role in cognitive pro-
the SMA. Activation of the cerebellum can cessing, evidenced by its intense connections
also be identified during motor imagery tasks with the associate cortex (Stoodley 2012). The
(Filgueiras et al. 2018) (Table 4.1). In addition cerebellum forms connections with multiple
to controlling a movement, the cerebellum cortical regions, including the prefrontal, tem-
plays multiple roles in planning and learning poral and parietal lobes. The analysis of intrin-
a movement. One of the key roles of the cere- sic functional connectivity (see Section 2.4)
bellum is to control movement by integrating revealed that cerebellar lobule VI is associated
sensory and motor information. The impor- with the salience network. Cerebellar crus I
tance of the cerebellum in motor learning can and crus II are associated with executive con-
be exemplified by its role in the development trol and the lobule IX is associated with the
of automaticity, which means the movement default mode network (Sokolov et al. 2017).
can be performed with ‘a decreased need for The intense connections between the cerebel-
attention and cognitive control’ (Cohen and lum and the networks of cognitive processing
Poldrack 2008). For example, individuals do may explain why cerebellar deficits are associ-
not need to pay much attention to their jaw ated with a diverse of cognitive or emotional
movement during eating. When one is repeat- dysfunctions in addition to motor dysfunctions
ing a movement, an internal model was estab- (Schmahmann et al. 2019).
lished for predicting the outcome of motor
output (Wolpert et al. 1998). The development
4.1.4 Brainstem
of automaticity of movement can be con-
ceived as a process of prediction and error The brainstem plays a key role in motor control
correction. In motor learning, individuals of rhythmic movements, such as swallowing,
established an internal model that links sen- chewing and breathing (Marder and
sory and motor information. A forward model Bucher 2001). The rhythmic pattern of oral
predicts the sensory outcomes when motor functions is produced and maintained by cen-
commands are executed. In contrast, an tral pattern generators (CPGs). A CPG refers to
inverse model calculates the motor com- ‘a small network of neurons, or even a single
mands that would produce the sensory out- neuron, whose activity can generate specific
comes from desired actions (Wolpert movements with correct timing and sequences
et al. 2001) (Figure 4.2b). The forward model in the absence of sensory feedback’ (Moore
adjusts sensorimotor processing via feedback et al. 2014). A distinct feature of CPGs is that
of predicted sensory outcomes. In contrast, they can generate and maintain rhythmic activ-
the inverse model generates motor commands ity independent of the input of timing informa-
in a feedforward direction (i.e. to achieve a tion from an extrinsic source (Marder and
movement). The cerebellum may play a key Bucher 2001). Notably, the CPGs do not work
standalone to maintain the pattern of mastica- The cerebellum also forms connections with
tion. The inputs from the brain and the periph- multiple cortical regions, including the pre-
eral sensory receptors, including periodontal frontal, temporal and parietal lobes, contrib-
mechanoreceptors and muscle spindles, also uting to cognitive and emotional processing.
modify the pattern of masticatory movement ●● The CPGs of the brainstem play a key role in
via the CPGs (Lund and Kolta 2006). Most of maintaining the rhythmic pattern of move-
our knowledge of CPGs has been based on ani- ments. However, how the CPGs cross-talk
mal research, and findings are summarized in with the other brain regions, especially for
previous reviews (Marder and Bucher 2001; the top-down control from the cortical motor
Moore et al. 2014). Until now, there is a lack of regions, has yet to be explored in neuroimag-
a practical neuroimaging method for investi- ing research.
gating the activity of human CPGs associated
with oral functions, and the cross-talk between
CPGs and other cortical/subcortical regions 4.2 Brain Mechanisms
has not been fully elucidated. The use of
of Human Mastication
ultrahigh-field magnetic resonance imaging
(MRI) may also improve the imaging results of
4.2.1 Introduction
the brainstem by enhancing the spatial resolu-
tion of the images (Sclocco et al. 2018). For the past 20 years, neuroimaging has made a
substantial contribution to our understanding
of brain mechanisms related to human masti-
4.1.5 Summary
cation. The neuroimaging findings are comple-
●● The relationship between a motor pro- mentary to the findings from earlier clinical
gramme and motor learning is dynamic. A and animal research, which have extensively
motor programme can be adjusted by receiv- investigated the neurophysiological mecha-
ing feedback according to the outcome of the nisms of mastication. In this chapter, we do not
activities. focus on the general neurophysiological mech-
●● Recent neuroimaging findings suggest that anisms of mastication. In contrast, we high-
the M1 plays a predominant role in motor light the novel evidence provided by
execution but is not limited to execution per neuroimaging. We primarily focus on the find-
se. The M1 works closely with the SMA/the ings and methodological issues of fMRI
PMC during the preparatory stage or merely research on chewing movement and the find-
by imagining a movement. ings of magnetoencephalography (MEG) and
●● In contrast to the M1, the SMA and the PMC, functional near-infrared spectroscopy (fNIRS).
especially their rostral subregions, are more Finally, a tentative model of the human masti-
relevant to the attention and cognitive pro- catory process is proposed, based on the novel
cessing of a movement. The PMC is critical findings from neuroimaging research.
to integrate sensory information (which is
usually an external guide of movements),
4.2.2 Earlier Findings from Animal
and the SMA is critical to self-initiate move-
Research on Chewing Movement
ment and maintaining learned movement
sequences. Before the advancement of neuroimaging,
●● The cerebellum plays a key role in tuning the the findings from animal research have
sensory and motor information for motor contributed to a general framework of the
control so that a motor command can be neurophysiological mechanisms of mastica-
continuously updated and the movement tion (Kawamura 1964). A critical advan-
can be controlled for accuracy and speed. tage of animal research is that, via direct
intervention on the experimental animals, These findings further confirmed the key role
one can obtain more valid conclusions of the CMA in the brain mechanisms of
regarding the cause–effect relationship mastication.
between structure and function. For exam-
ple, the role of the motor cortex in maintain- 4.2.2.2 The Sensorimotor Circuitry
ing oral functions can be demonstrated by of Mastication and Its Plasticity
the findings that in awake monkeys jaw and During mastication, the pressure and nocicep-
tongue muscle activities were disrupted tive receptors from the periodontium and the
when the face motor cortex was inactivated oral cavity, as well as the muscle spindles of the
(Yamamura et al. 2002). Moreover, using ani- masticatory muscles, are the major sources of
mal models, researchers can explore such a sensory feedback (Trulsson 2006). Animal
relationship at the single-neuron level (Lin research has provided much evidence regard-
and Sessle 1994; Lin et al. 1994). Several ing the integration of sensory and motor infor-
landmark findings have been demonstrated mation related to mastication (Lin et al. 1994;
via animal research, including identifying Lin and Sessle 1994). Moreover, new evidence
the role of the cortical masticatory area supports the notion that at the cortical level
(CMA) and the CPGs in jaw movement, and the sensorimotor circuitry may be sculpted in
the plasticity of the sensorimotor circuitry of accordance with long-term experience. For
mastication, as outlined in the following sec- example, in rats, the modification of the pat-
tions (for detailed reviews, see Avivi-Arber tern of occlusion (by trimming the incisor) was
and Sessle (2018) and Lund (1991)). associated with an altered brain representation
of the jaw and the tongue in the face region of
4.2.2.1 Identification of the Role of the the M1 (Avivi-Arber et al. 2015). In mice, tooth
Cortical Masticatory Area loss (by extraction) induced a change in brain
In rats and monkeys, the CMA has been identi- volume in widespread regions of the brain
fied as the cortical areas that generate rhyth- (Avivi-Arber et al. 2017). The animal findings
mic jaw movement by receiving repetitive demonstrate the association between altered
stimulation (Isogai et al. 2012). The causal role oral conditions and brain plasticity, which may
of the CMA in normal mastication has been relate to the rehabilitation of oral functions
identified by animal research. For example, (Sessle 2019).
stimulating the anteromedial half of the cere-
bral cortex using prolonged electrical pulses
4.2.3 Experimental Design
triggered rhythmical masticatory movements
of Neuroimaging of Mastication
in rabbits (Lund et al. 1984), and lesioning the
masticatory area induced deficits of normal In the following sections, we outline the neu-
mastication in cats (Hiraba and Sato 2005). In roimaging findings of human mastication,
addition, the CMA works with the putamen, mainly based on fMRI research. A typical task-
one of the core components of basal ganglia, to based fMRI study of mastication consists of a
form the corticostriatal pathway for jaw move- task condition, in which subjects perform a
ment (Masuda et al. 2005). The CMA that chewing task, and a baseline condition, in
induces chewing-like rhythmic jaw move- which subjects do not chew but just rest.
ments also shows a strong connection with the During the chewing task, subjects chew an
thalamus (Isogai et al. 2012). Finally, recent experimental material, which is usually a piece
evidence revealed that during development a of odourless chewing gum, for a fixed number
lack of occlusal loading is associated with defi- of strokes. In order to identify the brain activa-
cits of rhythmic jaw movements induced by tion associated with chewing, the sessions of
the stimulation of the CMA (Aung et al. 2020). the chewing task were contrasted with the
Figure 4.3 Experimental design for neuroimaging of the brain mechanisms of chewing. (a) The basic
concept of study design. The study consists of multiple blocks of a chewing task and a baseline (no-
chewing) block. (b) Variations of the study design. Different studies may differ in the number of blocks of
tasks and the definition of the baseline block (e.g. resting or clenching the teeth). The variations lead to a
different interpretation of imaging results.
baseline (resting) sessions, in which chewing comparing the brain activation between these
was not performed (Figure 4.3a). For example, segments, they were able to delineate the brain
in the study by Onozuka et al. (2002), subjects activation associated with a specific stage of
chewed a piece of gum for four cycles, 32 sec- mastication, such as the period when chewing
onds for each cycle, during an fMRI scan. The was just initiated and the period when chew-
blood‑oxygen-level-dependent (BOLD) signals ing was terminated.
from this task condition were contrasted to the As discussed in Section 2.1, all the neuroim-
signals from the baseline condition, i.e. four aging evidence needs to be evaluated for its
sessions of resting. Increased activation in the internal and external validity. In the neuroim-
task condition compared to the baseline condi- aging research on mastication, the lack of
tion is considered to be associated with masti- internal validity means that the observed brain
cation (Onozuka et al. 2002) (Figure 4.3b). activation may not truly reflect the masticatory
The experimental design by Onozuka et al. process. For example, in most neuroimaging
(2002) helps to clarify the brain activation studies, subjects chew a piece of odourless
associated with mastication by contrasting chewing gum. If the chewing gum has some
the chewing conditions to non-chewing condi- flavour, the brain activation may reflect the
tions. However, the design cannot disclose the gustation related to the taste of the gum rather
brain activation associated with a specific stage than mastication. Therefore, the factors (e.g.
during mastication. Mastication may consist of the taste of the gum) that may confound the
different stages, and sensorimotor processing desired function (e.g. mastication) should be
may vary between the stages. For example, well controlled in experimental design.
individuals may perceive stronger sensory Nevertheless, when the experimental condi-
feedback when they start to chew a hard mate- tion is well controlled for every aspect, the
rial and adjust to a rhythmic pattern of chew- result derived from the study may not provide
ing when they get used to the material. Finally, good external validity. In other words, the con-
they need to stop the rhythmic movement and clusion from the research cannot be general-
prepare for swallowing. In order to investigate ized in our daily life. For example, researchers
the brain mechanisms of different stages of may ask all the subjects to chew a piece of
mastication, Quintero et al. adopted a design odourless gum for 10 cycles, with the fre-
similar to the one used by Onozuka et al. quency of one second per cycle, so that the
(2002). In addition, they divided the chewing conditions of the task can be standardized
session evenly into five segments or ‘time win- across all subjects. However, the ‘mastication’
dows’ (Figure 4.3b) (Quintero et al. 2013a). By performed during the task is far different from
that we have performed in our daily life. In ‘simulated’ in a task condition during an imag-
other words, the task condition – which is ing scan.
highly constrained by many factors – does not
really reflect how individuals would really
4.2.4 Neuroimaging Findings of Brain
chew during eating. These issues of external
Mechanisms of Mastication
validity should be carefully explained when it
comes to a study of elderly and special needs In the next sections, we briefly review the
patients. In daily life, these patients may face recent finding of the brain mechanisms of
many difficulties of feeding (e.g. having the mastication from neuroimaging research
risk of being choked), which are hard to be (Table 4.2). In these sections, we primarily
Non-treatment-related research
Lin et al. (2020b) 31 patients of cognitive impairment (31 sMRI/VBM
healthy controls)
Lin et al. (2020a) 40 patients of cognitive impairment (30 sMRI/VBM
healthy controls)
Nagashima et al. (2020) 35 healthy adults fNIRS/mastication and foot stepping
Sörös et al. (2020) 17 healthy adults fMRI/frontal, horizontal and vertical
tongue movements
Feng et al. (2019) 10 healthy adults fMRI/occlusion of the left first
premolar, left second premolar and
right first premolar
Yoshizawa et al. (2019) 15 healthy adults fMRI, EMG/incisal and molar biting at
three step-wise force levels
Lin et al. (2017) 26 older and 26 younger healthy adults rs-fMRI/graph-based network analysis
Kawakami et al. (2017) 14 healthy adults fNIRS/gum chewing and a
computerized Stroop test
Choi et al. (2017) 8 healthy adults fMRI/gum chewing
Lotze et al. (2017) 15 healthy adults fMRI/chewing on a unilateral side
Viggiano et al. (2015) 18 healthy adults ASL-MRI/gum chewing on a
unilateral side
Jang et al. (2015) 9 healthy adults fMRI/hand movements w/wo gum
chewing
Sakamoto et al. (2015) 13 healthy adults EEG/a somatosensory Go/no-go task
w/wo gum chewing
Jiang et al. (2015) 16 healthy adults with a right or left fMRI/rhythmic chewing
chewing-side preference
Lin et al. (2015) 25 older adults sMRI, rs-fMRI/VBM and functional
connectivity
Treatment/disease-related research
Kanzaki et al. (2019) 25 patients with mandibular prognathism fNIRS, EMG/a calculation task and a
(17 subjects with normal occlusion) chewing task
(Continued)
Table 4.2 (Continued)
Narita et al. (2019a) 16 partially edentulous patients fNIRS, EMG/chewing w/wo wearing a
denture
Ozdiler et al. (2019) 16 patients of Class II malocclusion, fMRI/chewing and biting movements
8 wearing appliances full time and
8 wearing appliances at night while
sleeping (10 pre-treatment controls)
Narita et al. (2019b) 15 patients with occlusal dysesthesia (15 fNIRS/chewing
healthy controls)
Inamochi et al. (2017) 28 fully dentate adults wearing fMRI on pre-insertion (Day 0), Day 1,
experimental denture-base palatal plates Day 3 and Day 7 after insertion/
chewing
1
Fan et al. (2017) 12 possible sleep bruxism patients (12 H-MRS
healthy controls)
Kamiya et al. (2016) 12 elderly edentulous subjects (12 young fNIRS and EMG/chewing w/wo
healthy controls) wearing a denture
Perumal et al. (2016) 20 edentulous patients wearing complete EEG/before and after chewing and w/
dentures wo wearing a denture
Tramonti Fantozzi 8 patients with clicking sounds fMRI, EMG/a finger to thumb motor
et al. (2019) task with teeth in direct contact or
with a bite interposed between arches
Banu et al. (2016) 10 edentulous patients receiving an EEG on six different phases of
implant-supported overdenture assessment
Yılmaz (2015) 12 patients with bruxism (12 healthy fMRI/tooth clenching
controls)
Notes: ASL-MRI: arterial spin labelling magnetic resonance imaging; EEG: electroencephalography; EMG:
electromyography; fMRI: functional magnetic resonance imaging; fNIRS: functional near-infrared spectroscopy;
MRS: magnetic resonance spectroscopy; rs-fMRI: resting-state functional magnetic resonance imaging;
sMRI: structural magnetic resonance imaging; VBM: voxel-based morphometry.
focus on the findings from fMRI published in activation at the right M1. Chewing is addi-
recent years. Key findings from other imaging tionally associated with brain activation at the
modalities, such as MEG and fNIRS, are left M1, compared with ‘clenching’ (i.e. a con-
discussed. dition without rhythmic jaw movement).
Across the 13 studies of chewing movement,
4.2.4.1 The Sensorimotor Circuitry the loci of brain activation showed great inter-
of Mastication study variations. Activation at the M1 and the
A recent meta-analysis derived from the results PFC has been consistently identified in most of
of 20 fMRI studies (including 13 studies of the studies (n = 10 and n = 9, respectively).
mastication and 7 studies of clenching) identi- Activation at the S1, the SMA, the thalamus,
fied the brain regions most consistently associ- the insula and the cerebellum was also reported
ated with mastication (Lin 2018). These (n = 6). The findings revealed that the primary
regions include the bilateral M1, the right S1, sensorimotor area is critically associated with
the bilateral thalamus, the SMA and the cere- both chewing and clenching. Moreover, in
bellum (Lin 2018) (Figure 4.4). Both chewing addition to the M1 and S1, the secondary motor
and clenching are associated with consistent area (including the SMA and the PMC), the
M1
Thalamus
SMA
SMA
Y=0 M1
Y=0 Y = –18
M1
S1
S1
X = 58 M1
X = 58
Y = –66 Cerebellar
lobule VI
X = –42
Insula
Posterior Central operculum
X = –42 insula
Figure 4.4 Brain activation associated with chewing and clenching. Source: Lin (2018). Reproduced with
permission of John Wiley and Sons.
PFC and the cerebellum also contribute to the subjects conducted a mastication exercise,
brain mechanisms of mastication. The meta- regardless of the side of chewing. For jaw
analytical results generally confirmed the find- movement, an electroencephalography (EEG)
ings from animal research. Moreover, based on study revealed that during mouth opening and
the meta-analytical results, the loci of activa- closing, the movement-related cortical poten-
tion at the S1 and M1 are consistent with the tials were symmetrically distributed (Yoshida
somatotopic region of the orofacial area, i.e. et al. 2003). An fMRI study also showed bilat-
the lateral and ventral areas of the M1 and the eral activation during chewing (Lotze
S1. As a critical component of the descending et al. 2017). The bilateral activation at both
motor system, the thalamus showed reciprocal hemispheres was also identified in the meta-
connections with the CMA for rhythmical jaw analytic results (Lin 2018).
movements (Isogai et al. 2012). Consistently, In addition to fMRI, MEG and fNIRS have
the meta-analysis identified consistent brain also provided critical evidence regarding the
activation at the thalamus among the studies brain mechanisms of mastication. Using MEG
of chewing movement (Lin 2018). Hemispheric combined with a jaw-motion tracking device,
dominance was frequently found in the sen- researchers have recorded the brain activity
sory and motor cortices for limbs. In a recent associated with the jaw movement of mouth
arterial spin labelling magnetic resonance opening and phoneme production (i.e. to
imaging (ASL-MRI) study, Viggiano et al. speak a/pa/sound). The findings revealed the
(2015) found increased blood perfusion in role of the M1 in the jaw movement for both
the right trigeminal principal nucleus after chewing and speech (Memarian et al. 2012).
In addition to jaw movement, tongue move- focused on the session when subjects were
ment also plays a key role in mastication. One executing the chewing movement (Table 4.2),
study investigated the MEG and electromyo- it is not surprising that activation at the
graphic (EMG) signals during tongue protru- M1 was reported more frequently compared
sion and found a MEG–EMG coherence at the to the PMC or the SMA. The activation of the
bilateral hemispheres, regardless of the side of SMA has been identified in earlier research
tongue protrusion (Maezawa et al. 2014). An (Onozuka et al. 2002) and it is more frequently
earlier fNIRS study showed a consistent acti- reported compared to the PMC, according to
vation at the somatosensory cortex when sub- the meta-analytic results (Lin 2018). The
jects received painless vibrotactile stimuli, PMC activation is often associated with a
regardless of the position of the teeth (maxil- movement guided by an external cue, such as
lary or mandibular incisors, canines, first pre- visual feedback (Debaere et al. 2003).
molars and first molars). Moreover, the first Therefore, the lack of PMC activation may
molars showed stronger activation compared reflect the nature of the intraoral movement,
to the other teeth (Shimazaki et al. 2012). which is less dependent on the guidance from
Another study investigated the effect of differ- an external cue. A recent study on structural
ent intensities during clenching, which was features of older people revealed that grey
calibrated to 20%, 50% and 80% of the maxi- matter volume at the motor regions, includ-
mum clenching force. The results showed an ing the M1 and the PMC, was associated with
increased fNIRS activity in the primary sen- the individual masticatory performance (Lin
sory and motor cortices between the different et al. 2015). Since the structural features of
levels of intensity (Shibusawa et al. 2009). the brain reflect the intrinsic architecture of
Consistently, recent fMRI studies reveal an the brain, the findings suggested that struc-
extensive activation in the primary sensorimo- tural variability in the motor regions may be
tor area during occlusion of a single premolar associated with the individual difference in
(Feng et al. 2019). Stronger activation in the masticatory function.
primary sensorimotor area and the cerebel-
lum was associated with greater strength dur- 4.2.4.3 The Prefrontal Cortex
ing a molar biting task (Yoshizawa et al. 2019). One of the notable findings from the recent
Together, the results highlight the role of the neuroimaging studies is the activation of the
primary sensorimotor area in occlusion. PFC during chewing. As discussed earlier, pre-
frontal activation has been seen in many fMRI
4.2.4.2 The Secondary Motor Area studies of chewing movement (Lin 2018).
The secondary motor area consists of the However, there exist great inter-study variabil-
PMC and the SMA (Nachev et al. 2008; Picard ity in the loci of activation, which were widely
and Strick 2001). As shown in Section 4.1, distributed in the whole frontal area (Lin 2018).
both regions are critical to motor control, Activation at the PFC has been identified since
especially for planning and learning a move- the early days of neuroimaging research on
ment. The PMC was frequently reported in chewing (Onozuka et al. 2003). Onozuka et al.
the learning task when the learners are (2003) found higher activation in the PFC
required to develop a movement guided by an during a chewing task in the older subjects
external cue. In contrast, the SMA was found compared to the younger subjects. The PFC
when the learner has internalized the motor activation was not pronounced for younger
plan after repeated practice. In this condition, subjects (Onozuka et al. 2002; Quintero
the learner will develop a movement mainly et al. 2013a). Brain activation at the frontal and
guided by his/her memory (Debaere parietal lobes differed between tongue move-
et al. 2003). Because most of the fMRI studies ments in different directions (Sörös et al. 2020).
Activation of the frontal and parietal lobes was mastication. For example, in older subjects,
also associated with chewing-side preference mastication may improve subjects’ general-
(Jiang et al. 2015). When mastication was con- ized attention, which may be associated with
ducted at the same time as other ongoing tasks a greater PFC activity (Nagashima
(e.g. hand movement and a decision task), et al. 2020). Chewing was also associated
changes in brain activity were found (Jang with increased activity of the dorsolateral
et al. 2015; Sakamoto et al. 2015). The findings PFC when subjects were performing a Stroop
suggest that the PFC may not play a consistent test, which was engaged with attentional
role in the sensorimotor circuitry of mastica- processing (Kawakami et al. 2017). PFC
tion, such as the M1, the SMA and the cerebel- activity may also reflect the individual
lum. Instead, the PFC activation may signify difference in perceptual processing of chew-
the individual and contextual factors related to ing. For example, a recent fNIRS study
mastication, such as subjects’ age and the revealed that the patients with occlusal dys-
attention and cognitive processing during the esthesia showed a lower PFC activity during
chewing task. chewing compared to healthy controls.
The role of the PFC activation has also Moreover, in the patients, the PFC activity
been identified in neuroimaging studies of was associated with the severity of somatiza-
dental patients (Table 4.2). To study jaw tion, assessed by the somatization subscales
movement related to oral rehabilitation (e.g. of the Symptom Checklist-90-R (Narita
using a denture), fNIRS would be a suitable et al. 2019b). In a longitudinal fMRI investi-
tool because subjects can perform a chewing gation on healthy subjects wearing a palatal
task under less physical restriction. In con- plate, Inamochi et al. (2017) found decreased
trast, in fMRI studies, subjects’ movement activation at the superior frontal gyrus at the
can be constrained inside an MRI scanner. time when the subjects have worn the plate
Using fNIRS, Narita et al. (2019a) found an for seven days compared to the time right
increased prefrontal activation when eden- before the plate was inserted. The accumu-
tulous subjects chewed with a denture com- lating evidence further suggests that the PFC
pared to chewing without wearing a denture. may play a key role in the adaptation of oral
Kamiya et al. (2016) report that when the conditions (Lin 2019).
elderly edentulous subjects chewed with
their dentures, the prefrontal activation was 4.2.4.4 Cerebellum
not significantly different from that in the In addition to the M1/S1 and the SMA, the cer-
young participants. In another study, ebellum is another brain region commonly
Kanzaki et al. (2019) compared the brain identified in fMRI studies of chewing move-
blood flow during chewing between subjects ment (Lin 2018). However, because the cere-
of normal occlusion and mandibular bellum consists of structurally and functionally
prognathism. They found increased activity heterogeneous subregions (see Section 4.1), it
in the inferior frontal gyrus in both groups, would be critical to differentiate the subre-
but the increase was smaller in the patients gions that were associated with chewing move-
with mandibular prognathism (Kanzaki ment. According to the meta-analytic results,
et al. 2019). The findings suggest that the the bilateral lobule VI and crus I are consist-
PFC activity may reflect the individual ently identified in the fMRI studies (Lin 2018).
difference in age, the experience of wearing Notably, both the lobule VI and the cerebellar
a denture or the pattern of occlusion. crus are part of the posterior lobe of the cere-
Furthermore, the PFC activity may be bellum, which may play a key role in cognitive
associated with the inter-individual variabil- processing, as demonstrated by its connectivity
ity in attention and cognitive processing of with the associative cortices and its association
with cognitive impairment (Stoodley and mastication, focusing on the ‘masticatory pro-
Schmahmann 2010). cess,’ is proposed.
One hypothesis of the role of cerebellar
activation is the development and mainte- 4.2.5.1 Investigating Cause–Effect
nance of the ‘automaticity’ of chewing move- Mechanisms of Mastication
ment. In terms of motor control, automaticity Animal studies have identified the role of the
generally refers to ‘a decreased need for CMA and the CPGs in controlling rhythmic
attention and cognitive control’ associated jaw movements. Researchers found that the
with cognitive or motor tasks (Cohen and pattern of jaw movement can be modified via
Poldrack 2008). Moreover, developing and stimulating the CMA and the CPGs in animal
maintaining the automaticity of motor con- subjects. The findings support a causal role of
trol is associated with extensive practice, the CMA and the CPGs in maintaining jaw
which can be conceived as a process of refin- movements (Lund 1991; Nakamura and
ing the model of motor control by sensory Katakura 1995). Neuroimaging research pro-
and motor information (Schmahmann vides evidence from directly assessing human
et al. 2019; Cohen and Poldrack 2008). If the subjects. However, most neuroimaging evi-
cerebellum is critical to the automaticity of dence only accounts for the correlation, but
motor control, one may conceive that a func- not the cause–effect relationship, between the
tional connection exists between the cerebel- brain and behaviour (see Box 4.2 and Box 4.3
lum and the sensorimotor cortices. This is in the Companion Website). The cause–effect
supported by the findings from task-based mechanisms of the brain can be elucidated by
fMRI research. During chewing, there was a adopting new technology and experimental
positive correlation of the signals between design. Firstly, the causal effect can be better
the posterior lobe of the cerebellum and the clarified by directly manipulating the activity
M1 (Quintero et al. 2013b). In addition, find- of brain regions and observing how the rele-
ings from structural magnetic resonance vant behaviour is altered. This can be per-
imaging (sMRI) research revealed that older formed by using interventional approaches of
people, those who showed a higher mastica- neuroimaging, such as transcranial magnetic
tory performance, showed a stronger resting- stimulation (TMS) or transcranial direct cur-
state functional connectivity between the rent stimulation (TDCS). Both TMS and TDCS
PMC and the cerebellum (Lin et al. 2015). belong to the approaches of brain stimulation,
These findings suggest that the individual which alter brain activity in a non-invasive
variability in the architecture brain connec- way. At present, the use of interventional
tivity may be associated with their oral approaches has been rarely seen for modifying
functions. mastication (but more frequently for modify-
ing swallowing, see Section 4.3). Secondly, one
may consider improving the research design
4.2.5 Synthesis of the Neuroimaging
of neuroimaging studies. At present, most of
Evidence of Mastication
the studies have been cross-sectional research
So far, we have reviewed the new evidence (Table 4.2). Longitudinal research with a pro-
from neuroimaging that adds to our current spective design may better capture the effect
knowledge of the mechanisms of human mas- of different factors on mastication. For exam-
tication. In the following section, we first focus ple, the study by Inamochi et al. (2017) traced
on the major challenge of neuroimaging both brain activation and masticatory perfor-
research on mastication, i.e. a lack of cause– mance for seven days. They found that the
effect mechanisms about human mastication, brain activation at the S1 and the putamen
and an integrated framework of human presents a similar pattern as the improvement
of masticatory performance after installing a from the stimulation of the CMA or the CPGs,
denture (Inamochi et al. 2017). An earlier such as changes in the pattern of jaw move-
study also identified a temporal change of ment, can be directly measured. Physiological
brain activation after denture installation, but responses, such as changes in muscle activity
the activation returned to its original level in and the feedback signals from oral sensory
the long run (Luraschi et al. 2013). The under- receptors, can also be recorded in animal
standing of the cause–effect relationship plays models. Therefore, animal research has already
a pivotal role in clinical applications because it made a major contribution in helping us
provides a more precise mechanism for treat- understand many of the neural mechanisms
ment planning and predicting treatment between the CNS and the peripheral appara-
outcomes. tus, as shown in Figure 1.2. What is missed
here is the mechanisms of sensorimotor
4.2.5.2 An Integrative Model of the Human control in the higher cortex. How do individu-
Masticatory Process als establish a motor plan and perform an act
As shown in Figure 1.2, the neural circuitries via motor learning? What is the role of
between the central nervous system (CNS) and cognitive and affective processing in motor
the peripheral apparatus of the stomatognathic control? How the sensory and motor
system have been systematically explored for information is integrated to achieve a complex
many years. However, the association between movement? These questions may be hard to
mastication and the organization of the brain, investigate merely using animal models
especially the cortical regions, remains unclear. because they are related to complicated per-
It should be noted that for most animal ceptual, cognitive and emotional experience,
research, the behavioural responses derived which are difficult to assess in animal subjects.
●● The meta-analytical results generally con- a long history because swallowing dysfunc-
firmed the findings from animal research, tions have a detrimental effect on health
revealing the loci of activation at the S1 and (Martino et al. 2005). Clinical research has
M1 that fit human sensory and motor soma- consistently reported that in patients with
totopy, as well as consistent activation at the strokes brain lesions are associated with their
SMA, the thalamus and the cerebellum. symptoms of swallowing difficulty (Suntrup-
●● Activation of the PFC is not consistently Krueger et al. 2017; Wilmskoetter et al. 2019).
reported in fMRI studies of mastication, such Consistently, animal research has identified
as the M1, the SMA and the cerebellum. Its that the brainstem plays an important role in
activation may signify the individual and the rhythmic and reflexive movement of
contextual factors related to mastication, swallowing (Ertekin and Aydogdu 2003;
such as subjects’ age and the attention and Lang 2009; Miller 2008). In this section, we
cognitive processing during the chewing task. focus on the neuroimaging evidence of
●● The cerebellar activation during chewing human swallowing research. Furthermore,
may be associated with the development and because dysphagia is a great clinical chal-
maintenance of the ‘automaticity’ of mastica- lenge to health, we discuss the methodologi-
tion. Variability in the pattern of brain con- cal issues of fMRI research on swallowing and
nectivity of the cerebellum may be associated propose several critical directions for future
with the individual difference in mastication. research.
●● The integrative model suggests that beyond
the sensorimotor cortices, an extensive net-
4.3.2 Earlier Findings from Animal
work of cortical and subcortical regions is
and Clinical Research on Swallowing
associated with the automaticity of masti-
cation. Furthermore, the engagement with Swallowing is generally considered a move-
multiple cortical regions (including the ment consisting of both voluntary and involun-
PMC, the SMA and the PFC) is critical to tary processes (Ertekin and Aydogdu 2003).
the learning and planning of oral motor Swallowing is not merely a movement of oral/
functions, which is key to the adaptation of pharyngeal/oesophageal apparatus but a com-
oral functions. plicated process associated with sensorimotor
●● To describe the holistic experience of masti- and cognitive–affective processing (Lang 2009;
cation, we prefer the term ‘masticatory pro- Windel et al. 2015). A critical element for suc-
cess,’ which consists of not only jaw cessful swallowing is the sensory input from the
movement but also the sensorimotor inte- peripheral apparatus (e.g. the oropharyngeal
gration of oral apparatus and cognitive– mucosa) to the nucleus tractus solitarius (NTS)
affective experience related to mastication, in the brainstem, via the trigeminal, glos-
focusing on how individuals adapt to their sopharyngeal and vagus nerves (Nishino 2012).
oral conditions and improve their perfor- The roles of several neural substrates in the
mance in feeding. brainstem have been identified based on animal
research. The NTS, a critical component of the
CPG, may play a key role in the pharyngeal and
4.3 Brain Mechanisms oesophageal phases, which are mostly an invol-
untary process of swallowing. The ventromedial
of Human Swallowing
nucleus of the NTS is critical to the coupling
between the pharyngeal and the oesophageal
4.3.1 Introduction
phases (Lang 2009). Notably, the different
Compared to mastication, research on the phases can occur independently of the other
brain mechanisms of human swallowing has phases. For example, the pharyngeal phase may
z = 8 mm z = 20 mm z = 36 mm z = 46 mm x = 40 mm
6 2b
4 3 6
1
2a
3
R
L R
4 15
3
Activation
1 10
6 likehood
(×10–3)
5
y = 4 mm y = –8 mm y = –56 mm y = 0 mm
z = 16 mm z = 40 mm z = 48 mm x = 36 mm
1c 1b 3
2a 1a
3
15
Activation
10 likehood
L R (×10–3)
1a 5
1c 1b 1a
2b 1b
2a 2a
3
y = –8 mm y = 8 mm y = –2 mm x = 6 mm
Figure 4.6 Brain activation associated with water (the upper panel) and saliva (the lower panel)
swallowing. Source: Sörös et al. (2009). Reproduced with permission of John Wiley & Sons, Inc.
one’s own saliva, (ii) covert swallowing, i.e. to (Inamoto et al. 2015). Furthermore, using
imagine the feeling about swallowing one’s fNIRS, Kamarunas et al. (2018) found that
own saliva, (iii) oral-sensory stimulation, i.e. when healthy subjects performed reflexive
receiving air pulses to the back of the mouth saliva swallowing, cortical activation can be
and (iv) breath holding. The comparison recorded before the pharyngeal phase of swal-
between overt and covert swallowing, which lowing in the S1, confirming the role of sen-
differed in the execution of swallowing, sory feedback in swallowing. Using EEG,
revealed increased activation at the bilateral Cuellar et al. (2016) found an increased mu
S1, the right M1, the right insula and thalamus activity in swallowing compared to tongue tap-
and the bilateral cerebellum. The pattern of ping. The EEG index may signify the process-
activation is generally consistent with the ing of sensorimotor integration in the late
results of the neuroimaging meta-analysis stage of swallowing. Using neuroimaging, one
(Sörös et al. 2009), which revealed an activa- can further investigate the brain mechanisms
tion at the bilateral S1/M1 and the right insula. related to motor planning and learning of
Notably, the findings from fNIRS also reveal swallowing. As shown in Table 4.3, recent find-
increased brain activity associated with volun- ings revealed that to execute swallowing and to
tary swallowing at the bilateral S1/M1 imagine swallowing were associated with an
Sörös et al. (2020) 17 healthy adults fMRI/frontal, horizontal and vertical tongue movements
Jing et al. (2020) 29 healthy adults fMRI/observing and executing swallowing
Dietsch et al. (2019) 21 healthy adults fMRI, VFSS/high-intensity taste stimuli
Kober et al. (2019) 11 healthy adults Real-time fMRI/executing and imagining swallowing
movements
Lin et al. (2019) 40 healthy older adults sMRI/VBM
Kamarunas et al. (2018) 14 healthy volunteers fNIRS/spontaneous reflexive saliva swallow
Jestrović et al. (2018) 15 healthy adults EEG/swallowing water of different bolus volumes w/wo
distraction
Toogood et al. (2017) 7 healthy adults fMRI/a delayed response, go, no-go study of four tasks of
swallowing
Mulheren et al. (2016) 15 healthy adults fNIRS/swallowing and taste stimuli
Saker et al. (2016) 20 healthy adults fMRI in thirst and oversated conditions/swallowing water
and sugar solution
Jestrović et al. (2016) 55 healthy adults EEG/swallowing water, nectar-thick and honey-thick
liquid with neutral and chin-tuck positions
Cuellar et al. (2016) 25 healthy adults EEG/water swallowing and tongue tapping
Jestrović et al. (2015) 55 healthy adults EEG/swallowing in the neutral and chin-tuck head positions
Windel et al. (2015) 27 old healthy adults and fMRI, SCR/water swallowing
24 young healthy adults
Inamoto et al. (2015) 15 healthy adults fNIRS/volitional water swallowing
Notes: EEG: electroencephalography; fMRI: functional magnetic resonance imaging; fNIRS: functional near-
infrared spectroscopy; SCR: skin conductance responses; sMRI: structural magnetic resonance imaging;
VBM: voxel-based morphometry; VFSS: videofluoroscopic swallowing study.
overlapped network, including the S1/M1, the mid-cingulate cortex and the putamen) did
insula, and the cerebellum (Kober et al. 2019), not show a significant difference between the
which were consistently identified in earlier two conditions (Lowell et al. 2008).
studies (Sörös et al. 2009). Activation of the
S1/M1 was also found when subjects were 4.3.4.3 Brain Mechanisms Associated
watching an action video of swallowing (with- with Sensory Processing of Swallowing
out executing swallowing) (Jing et al. 2020). In addition to motor processing, cumulating
During the preparation of swallowing, neuroimaging evidence highlighted the
increased activation was identified in the importance of sensory processing related to
anterior cingulate cortex and the PMC swallowing. The earlier study by Lowell et al.
(Toogood et al. 2017). Lowell et al. (2008) (2008) revealed a great extent of overlap of
found that covert swallowing was associated brain activation at the S1 between the overt
with less activation (in volume) in multiple and oral-sensory conditions, suggesting the
sensory regions (e.g. the S1 and the thalamus) role of sensory inputs in volitional swallow-
compared to overt swallowing. In contrast, ing. As shown in Table 4.3, the brain mecha-
the motor-related regions (e.g. the SMA, the nisms of swallowing may be associated with
taste perception. For example, supertasters, though activation at the PFC is also found in
i.e. individuals perceiving a greater intensity these fMRI studies, its role in swallowing has
from taste stimuli, showed decreased activa- remained unclear. Notably, the PFC is a large
tion at the S1 during gustatory stimulation area that consists of various functionally het-
compared to non-supertasters (Dietsch erogeneous regions. In the WST, activation
et al. 2019). An fNIRS study reported that was consistently found in the middle frontal
brain activity was greater when subjects swal- gyrus, which is engaged in a variety of cogni-
lowed a sour bolus compared to water tive processing. In contrast, in the SST, activa-
(Mulheren et al. 2016). In addition to taste, tion was consistently found in the thalamus,
the viscosity of fluid for swallowing was asso- which is spatially and functionally connected
ciated with the brain network in multiple to the motor cortex (Sörös et al. 2009). Such
frequency bands, as assessed using EEG diverse and widespread activation may indi-
(Jestrović et al. 2016). Vibration stimulation cate a greater variation dependent on tasks or
over the larynx, compared with sham stimu- individual differences, which have not been
lation, was associated with increased sponta- fully elucidated. As shown in Table 4.3,
neous swallowing and increased brain changes in the alpha and beta frequency bands
activity in the motor cortex (Mulheren and may reflect the attentional and cognitive
Ludlow 2017). Together, the evidence sug- demands during swallowing (Jestrović
gests that brain mechanisms may underlie et al. 2018). The PFC may play a key role in
the association between swallowing and sen- swallowing inhibition when subjects drunk
sory feedback from various modalities. excessive water. After overdrinking, the PFC
activation when subjects prepared to swallow
4.3.4.4 The Role of Brain Regions Other Than was correlated with their ratings of swallow-
the Sensorimotor Cortex ing effort (Saker et al. 2016). The findings sug-
In addition to the sensorimotor cortices, the gest that the PFC activation may reflect the
association between swallowing and wide- individual difference in cognitive, affective
spread regions, including the right insula, the and motivational aspects of swallowing, a
bilateral cingulate cortex and the PFC, has hypothesis not yet be fully elucidated.
been identified in recent neuroimaging stud-
ies (Table 4.3). It should be noted that, how- 4.3.5 Prominent Questions
ever, the activation of these regions is not for Further Investigation
consistent across studies. Among these
regions, the right insula and the cingulate cor- One of the major purposes of neuroimaging
tex were identified in most of the swallowing research is to facilitate the translational appli-
tasks. The loci of the insula activation are in cation of neuroscientific findings to clinical
the mid-posterior division of the insula, which management. For the brain mechanisms of
shows a more pronounced connection with swallowing, there are still some issues under
the primary somatosensory cortex (S1) and debate, and neuroimaging may be a suitable
the secondary somatosensory cortex (S2) com- tool for further clarifying these issues, as out-
pared to the anterior insula (Wiech et al. 2014). lined in the following sections.
Researchers also identified the brain connec-
tivity between the insula and an extensive 4.3.5.1 The Cognitive–Affective Processing
brain region, including the cortical sensorimo- in Swallowing
tor area, the PFC and the subcortical regions Until now, most of the neuroimaging studies
when subjects performed a voluntary swal- of swallowing focus on the sensorimotor
lowing task (Lowell et al. 2012). In contrast, processing of swallowing, especially the
execution of swallowing. However, just like interaction and the underlying brain mecha-
mastication (see Section 4.2), swallowing con- nisms may contribute to the clinical manage-
sists of not just movements of the oral/phar- ment of patients with dysphagia, for whom
yngeal/oesophageal apparatus but also aspiration is a common challenge.
complicated cognitive–affective processing.
For example, during swallowing, older adults 4.3.5.3 Association Between the Perception
showed a greater skin conductance rate, an of Food and Swallowing
index for heightened emotional arousal com- Tactile and thermal perception may play a key
pared to younger adults (Windel et al. 2015). role in normal swallowing by triggering the
Patients with cognitive decline showed a dif- pharyngeal reflex of swallowing. For example,
ferent pattern of brain activation compared to food modification (for changing its physical
healthy controls (Humbert et al. 2010). There properties) is a common method for improving
is also cumulating evidence showing that swallowing for patients with dysphagia (see
learning plays an important role in motor Section 8.3). It is surprising that the associa-
control of swallowing (Humbert and tion between the perception of food and swal-
German 2013). One of the major reasons why lowing has been rarely investigated. The
cognitive and affective processing is usually relevant issues should be investigated under a
ignored in swallowing research is the lack of broader scope, in which perceptual processing
assessment tools. Most of the clinical assess- of different sensations, the integration of mul-
ments, such as a self-report assessment (e.g. tisensory information and cognitive–affective
10-Item Eating Assessment Tool, EAT-10) or a processing about food (e.g. the pleasantness
functional assessment (e.g. the repetitive derived from swallowing) should be
saliva swallowing test, RSST), assess swallow- considered.
ing function per se rather than the cognitive or
affective experience associated with swallow-
ing (see Section 3.1). Understanding the cog- 4.3.6 Summary
nitive and emotional experience related to ●● Neuroimaging research on swallowing can
swallowing is critical for clinical manage- be performed using an SST, in which sub-
ment, especially for elderly and special needs jects swallow their own saliva without exter-
patients. In these patients, the decline in cog- nal stimuli, and a WST, in which swallowing
nitive functions and swallowing difficulty are is triggered by an external bolus of water
common due to healthy aging or age-related delivered to subjects via a tube. The SST or
disorders. the WST is an overt swallowing task that
requires subjects to swallow.
4.3.5.2 The Interaction Between Swallowing ●● Results from meta-analyses revealed a con-
and Other Rhythmic Movements sistent activation at the regions of sensori-
As noted earlier, the brainstem coordinates the motor processing, including the M1, the S1
rhythmic patterns of different movements. For and the insula, which may be associated
example, an interruption of the coordination with swallowing movement, independent of
between swallowing and breathing may the task condition. In contrast, the activation
increase the risk of aspiration. At present, most of the parietal and the cingulate cortices may
neuroimaging studies have focused on the reflect a task-dependent condition of
brain mechanisms of swallowing per se. The swallowing.
interaction between swallowing and other ●● In addition to the sensorimotor cortices,
rhythmic movements, especially breathing, the association between swallowing and
has yet to be clarified. Understanding this widespread regions, including the right
insula, the bilateral cingulate cortex and term ‘sensorimotor learning’ highlights the
the PFC, has been identified in neuroimag- importance of sensory processing in this learn-
ing studies. The PFC activation may reflect ing process. The neural mechanisms of sensori-
the individual difference in cognitive, motor learning can be interpreted under
affective and motivational aspects of swal- different models, according to ‘the type of infor-
lowing, a hypothesis not yet be fully mation that the motor system uses as a learning
elucidated. signal’ (Wolpert et al. 2011). It should be noted
that different types of information do not mean
the variety of sensory modalities (e.g. visual or
4.4 Cognitive Processing auditory). The major concern of sensorimotor
learning is how the flow of information process-
and Motor Learning
ing is controlled. The major models of informa-
of Oromotor Movement tion processing are discussed as follows.
4.4.1 Introduction
4.4.2.1 Error-Based Learning
In the preceding chapters, we have a brief One of the critical components of the sensori-
review of recent neuroimaging findings of the motor system is the continuous feedback from
brain mechanisms of mastication and swal- sensory input, which provides information
lowing. In general, mastication and swallow- regarding the result when a movement is made.
ing are both consistently associated with brain The sensory input from actual perception is
activation at the sensorimotor cortices. compared with the desired or predicted out-
However, as revealed by the neuroimaging come. A prediction error is generated when
findings, the sensory and motor regions may there exists a mismatch between the predicted
not work alone. For example, brain regions of and perceived outcomes. Such an ‘error’ does
the associative cortex (e.g. the PFC), the not mean a deficit of the sensorimotor system. It
regions associated with sensorimotor adapta- represents the discrepancy between the out-
tion (e.g. the cerebellum) and the regions of come (i.e. the sensory input perceived from
attentional processing (e.g. the insula and the executing a movement) and the prediction (i.e.
cingulate cortex) play a key role in mastication the sensory input that the brain anticipates to
and swallowing. In the following sections, we receive). The error signal is crucial for the self-
further look into how the sensory and motor adjustment of a control system. In order to gen-
information is modulated and how cognitive erate a more precise movement, the system will
processing, including attention control and be tuned so that the error signals are minimized.
learning, plays a key role in mastication and As shown in the following sections, learning
swallowing. We summarize recent evidence of based on error signals is also crucial for fast sen-
the following issues: sensorimotor learning, sorimotor adaptation. The cerebellum plays a
sensorimotor integration and sensorimotor critical role in error-based learning (Bostan and
adaptation. Directions of future research on Strick 2018) (Figure 4.2), evidenced by the clini-
the sensorimotor processing of oral functions cal findings that patients with cerebellar deficits
are proposed. would show impaired performance in the adap-
tation tasks (Tseng et al. 2007).
gradually by minimizing the error. However, executing the movement. The sensory system
human action can be more complicated than also receives feedback from the environment
this. Sometimes we need to adjust the way to so that the brain would calculate if the motor
act according to the changing environmental command has been carried out precisely. In
conditions. In addition to error signals, more this section, we focus on two specific topics
information will be used for optimizing motor regarding the contribution of sensory pro-
control, including the reinforcement signal, cessing in motor control. Firstly, the role of
which adopts ‘information about the relative the M1 and the S1 in motor control is revis-
success and failure of the movement’ (Wolpert ited, based on recent animal research on sen-
et al. 2011). In reinforcement-based learning, it sorimotor integration. Secondly, the brain
is the reward or punishment provided by the mechanisms of action observation are
environment as the signals to adjust a move- discussed.
ment, and one’s performance is improved
when the total rewards are maximized
4.4.3.1 The Role of the M1 and the S1 in
(Wolpert et al. 2001). Therefore, sensorimotor
Motor Control
learning based on reinforcement is closely
An interesting finding from neuroimaging
associated with the basal ganglia (Figure 4.2)
studies is that during mastication and swal-
and the circuitry of reward processing (Bostan
lowing, not only the M1 but also the S1 was
and Strick 2018).
consistently found activated (Lin 2018; Sörös
et al. 2009). What are the mechanisms under-
4.4.2.3 Supervised vs. Unsupervised Learning
lying the co-activation of the M1 and the S1?
The error signals of error-based learning are
Recent findings from animal research
considered a teaching signal, such as the com-
revealed that the S1 may play a key role in
ment from a teacher that instructs students
motor control. The researchers found that
how much needs to be improved. This is a form
the activity of the M1 evoked protraction of
of supervised learning, which requires a cor-
rodents’ whiskers, which can be deemed as
responding target value to guide the learning
the major sensor of the animal. Notably, the
process (Bostan and Strick 2018). In terms of
activity of the S1 would drive the retraction
sensorimotor learning, the motor output (e.g.
of their whiskers, suggesting its role in pro-
reaching an object with an arm) is the target
viding rapid feedback for sensorimotor inte-
corresponding to each sensory input.
gration (Matyas et al. 2010). In another study,
Therefore, supervised learning aims to learn
the researchers identified the role of the
‘the mapping from inputs to outputs specified
M1 in controlling rapid feedback. They found
by this teaching signal from the environment’
that when an animal was performing a motor
(Wolpert et al. 2001). In contrast, in unsuper-
action under the perturbation of external
vised learning, there is no target value corre-
forces, the information of shoulder and
sponding to the input data (Wolpert et al. 2001),
elbow motion (which reflects how the pertur-
and the learning process will not be guided by
bation is countered) is integrated into the M1
the external outcome. The Hebbian plasticity
(Pruszynski et al. 2011). In other words, the
is considered as a form of unsupervised learn-
M1 participates in movement execution as
ing (Wolpert et al. 2001), which is further dis-
well as the integration of motion-related
cussed in Chapter 8.
information. The findings suggest that the S1
and the M1 do not work separately for sen-
4.4.3 Sensorimotor Integration
sory and motor processing. Instead, they both
When a movement is made, it is not just the play a key role in the rapid integration of
motor system that would send a command for information during a movement.
4.4.3.2 Action Observation – The Mirror individual performs a given behavior and when
Neuron System an individual observes another person perform-
In the preceding sections, we highlight the ing the same or a similar behavior’ (Rizzolatti
importance of sensorimotor integration of and Sinigaglia 2016), may play a key role in
motor control, i.e. how the sensory feedback is motor learning via observation. In humans, the
integrated for adjusting one motor command. mirror system consists of distributed brain
The framework also applies to sensorimotor regions, including the inferior parietal lobule,
learning, in which both motor prediction and the ventral area of the PMC and the inferior
sensory feedback play a critical role (see frontal gyrus. The core network is associated
Section 4.1). However, we do not always learn with other brain regions, including the primary
how to conduct an action by doing it. Sometimes visual cortex, the cerebellum and the limbic sys-
an action can be learned by watching how it is tem, according to the findings of a neuroimaging
conducted by other individuals. The mental pro- meta-analysis (Molenberghs et al. 2012). Several
cess of observing and learning others’ actions is roles of the mirror neuron network have been
very complicated. A specific set of neurons, proposed. (i) Its activation may represent action
namely the mirror neuron, plays a key role in goals, i.e. ‘an outcome to which the action is
this process. The story of the mirror neuron directed’ (Rizzolatti and Sinigaglia 2016). (ii) It
began in a lab where researchers installed elec- also facilitates the understanding of others’
trodes in the brain of a monkey for an investiga- actions because the network may link the
tion on sensorimotor processing. The researchers observed actions to one’s motor representation,
identified some neural activity when the mon- which is derived from one’s own experience in
key was grasping an object. Accidentally, the executing the same action. (iii) Therefore, the
researchers found that the monkey also showed network may play a key role in emotion and
a pronounced brain activity when it was watch- social interaction (Rizzolatti and Sinigaglia 2016),
ing another guy grasping a cone of gelato and which is markedly associated with the empa-
moving it close to the mouth (Gazzaniga et al. thetic experience (e.g. watching someone being
2019). The researchers started to hypothesize cut and feeling the hurt by oneself).
that the neurons from which activity was
recorded were located in the monkey’s PMC,
4.4.4 Sensorimotor Adaptation
may respond to both action (grasping something
by oneself) and perception (seeing someone In a broader sense, sensorimotor adaptation can
grasping something). In other words, there may be regarded as a form of sensorimotor learning.
exist a system that encodes both the action per- Here, sensorimotor adaptation is defined as ‘a
formed by oneself and the perception of the type of procedural-motor-learning that allows
same action executed by others. maintaining accurate movements in the pres-
Later experiments confirmed that these ‘mir- ence of environmental or internal perturbations
ror’ neurons were activated when an animal by adjusting motor output’ (Della-Maggiore
subject cracked a peanut itself, when it watched et al. 2015). By definition, adaptation refers to not
someone cracked a peanut, or when it heard the just an outcome status but a process in which
sound of cracking a peanut. Notably, the neu- individuals try to cope with a challenge from the
rons were also activated when the monkey environment. For example, when patients have
watched someone was just approaching the pea- their missing teeth restored by a dental bridge,
nuts – i.e. the neurons respond not to a specific their perception of teeth occlusion or pattern of
sensory input (by watching or hearing) but to chewing may be perturbed due to the new bridge.
the intention represented by the action (Umiltà Nevertheless, most people ‘feel better’ some days
et al. 2001). The mirror neuron, which refers to after and can chew just as usual, without even
‘neurons that discharge both when an noticing the presence of a dental bridge.
Therefore, sensorimotor adaptation is an impor- improved after several trials (i.e. adaptation).
tant issue in dental treatment. The methods to When the force field is removed, an after-effect
investigate its brain mechanisms and related (i.e. an overshot in the opposite direction) will
findings are discussed in the following sections. appear immediately. Both tasks show that senso-
rimotor adaptation is associated with a dynamic
4.4.4.1 Experimental Design for Studying adjustment of sensory input and motor output
Sensorimotor Adaptation so that a challenge from the environment (e.g.
In order to study sensorimotor adaptation, a the effect from the prism glasses or the force
common strategy is to perturb the originally field) can be coped with.
learned sensorimotor association and to exam-
ine how this perturbed effect is restored. A 4.4.4.2 Brain Mechanisms
widely used experimental task is the ‘prism of Sensorimotor Adaptation
adaptation’ task, which requires subjects to The current evidence shows that sensorimotor
reach their hands to an object located centrally. adaptation, as observed using the paradigm of
All subjects are expected to reach the target prism adaptation, would be associated with a
accurately once the sensorimotor association cerebello-parietal network (Panico et al. 2020).
between visual and the proprioceptive input and The temporal and prefrontal regions may be
the motor output has been built. The next stage associated with the effects on spatial cogni-
(i.e. the exposure stage) of the experiment is to tion, and the motor cortex may play a key role
have the built association perturbed by experi- in consolidating the adaptive effect (Panico
menters. For example, when subjects wear et al. 2020). Clinical research has consistently
prism glasses that distort the visual input, the shown that patients with deficits in the cere-
object becomes more leftward to its genuine bellum, the prefrontal and the parietal regions
position, and the subjects move their hands left- failed to show the adaptation examined by the
ward, which causes a discrepancy between the prism task (Gazzaniga et al. 2019). If these
visual input and the outcome of movement. regions are essential to sensorimotor adapta-
After several trials, subjects would re-calibrate tion, interference in these regions should
the sensorimotor association for that discrep- influence the individual performance of the
ancy so that the object can be reached accurately. task. The hypothesis was recently tested using
Interestingly, if the prism glasses are removed TDCS. On young adults, researchers found
(i.e. the post-exposure or after-effect stage), sub- that the TDCS interference on their cerebel-
jects will immediately show an ‘overshot’ of lum leads to more deviation of movement not
their hand movement, i.e. moving more right- only when subjects were first exposed to the
ward to the object. This is because the re- prism effect but also when the prism is
calibrated association that they established removed. Therefore, the cerebellum may play
during the exposure stage still has an effect a key role in both stages of sensorimotor adap-
(Fernández-Ruiz and Díaz 1999). After several tation, i.e. the adaptation associated with the
trials of practice, the ‘overshot’ diminishes since early recalibration and the realignment during
subjects have re-established a new association. the after-effect stage (Panico et al. 2016).
Another widely used design is the force-field In addition to the cerebellum, the frontal cor-
paradigm (Della-Maggiore et al. 2015). During tex may play a key role in sensorimotor adapta-
the experiment, a subject’s arm is attached to a tion. The frontal regions may be associated with
velocity-dependent force field, which will per- ‘sensorimotor savings,’ i.e. the faster adaptation
turb the arm’s movement to reach a target. When or ‘relearning’ to the perturbation that one has
the force field is working, the trajectory of the previously adapted to (Seidler et al. 2017). This
subject’s arm movement will show a great degree is demonstrated by a recent study of TDCS stim-
of deviation from the object, and this will be ulation on the prefrontal and motor cortices.
TDCS stimulation on these regions was associ- (Aglioti et al. 2008). The findings suggested that
ated with improvement in the task performance the acquisition of expertise is not only associated
compared to the sham (control) stimulation. with the development of a motor skill but also
However, only the subjects who received TDCS with a strengthened association between observ-
stimulation on the right PFC exhibited greater ing and performing an action.
savings. The findings suggest that the right PFC,
but not the motor cortex, may contribute to sen-
sorimotor adaptation and savings (Seidler 4.4.6 Summary
et al. 2017). There also exists evidence support- ●● Motor learning is generally defined as the
ing the role of the basal ganglia in sensorimotor process of refining the sensorimotor associa-
adaptation. Researchers used fMRI and a joy- tion to adapt sensorimotor transformations
stick aiming task to examine the adaptation of for environmental challenges.
visuomotor rotations and found that brain acti- ●● In error-based learning, a prediction error is
vation at the cerebellum and sensorimotor generated when there exists a mismatch
regions was associated with early adaptive pro- between the predicted and perceived out-
cesses. Moreover, activation was found in the comes. In order to generate a more precise
right globus pallidus and putamen as well as the movement, the system will be tuned so that
right PFC, PMC and parietal cortices, which the error signals are minimized.
may be associated with spatial cognitive pro- ●● In reinforcement-based learning, it is the
cesses of adaptation (Seidler et al. 2006). reward or punishment provided by the envi-
ronment as the signals to adjust a move-
4.4.5 Acquisition of Expert Skills ment, and one’s performance is improved
when the total rewards are maximized.
The previous sections mainly focus on the senso-
●● The mirror neuron responds to both the con-
rimotor learning of a relatively simple move-
ditions when one performs a given behaviour
ment, such as reaching an object or rotating a
and observes another individual performing a
joystick. However, mastering a complex skill,
similar behaviour. It may play a key role in
such as the expertise being a pianist, a gymnast
motor learning via action observation.
or a dentist, may be associated with more com-
●● The experimental task of ‘prism adaptation’
plicated learning. The development of an expert
is widely used for investigating sensorimotor
skill consists of multiple facets of mental func-
adaptation. Sensorimotor adaptation, as
tions. For example, the development of artistic
observed using the paradigm of prism adap-
skills may be associated with cognitive, emo-
tation, would be associated with a cerebello-
tional and motivational processing (Kirsch
parietal network
et al. 2016). The acquisition of an expert skill
●● To master a complicated motor skill, such as
may also be associated with the mirror system.
the expertise being a pianist, a gymnast or a
In a TMS-based study, Aglioti et al. (2008) asked
dentist, may be associated with a more com-
the skilful basketball players, the sports journal-
plicated process, including cognitive pro-
ists who often watched basketball matches, and
cessing, attentional control, motivation and
the novice of basketball to watch the videos
emotional processing.
regarding a basketball throw and a soccer kick of
football. They found that the players and the
journalists showed an increased motor activity Further Readings
compared to the novice when watching the vid-
eos of a basketball throw. Notably, only in the Please see the Companion Website for
basketball players, their cortical activity was able Suggested Readings, Box 4.1 and 4.3 (From
to predict the outcome of the basketball throw Tools to Discovery), and Tables with updated
(i.e. success or failure) depicted in the videos information.
References
Aglioti, S.M., Cesari, P., Romani, M., and Urgesi, Cohen, J.R. and Poldrack, R.A. (2008).
C. (2008). Action anticipation and motor Automaticity in motor sequence learning does
resonance in elite basketball players. Nat. not impair response inhibition. Psychon. Bull.
Neurosci. 11: 1109–1116. Rev. 15: 108–115.
Alberts, M.J., Horner, J., Gray, L., and Brazer, Cuellar, M., Harkrider, A.W., Jenson, D. et al.
S.R. (1992). Aspiration after stroke: lesion (2016). Time-frequency analysis of the EEG
analysis by brain MRI. Dysphagia 7: mu rhythm as a measure of sensorimotor
170–173. integration in the later stages of swallowing.
Aung, P.T., Kato, C., Abe, Y. et al. (2020). Clin. Neurophysiol. 127: 2625–2635.
Functional analysis of rhythmic jaw Debaere, F., Wenderoth, N., Sunaert, S. et al.
movements evoked by electrical stimulation (2003). Internal vs external generation of
of the cortical masticatory area during low movements: differential neural pathways
occlusal loading in growing rats. Front. involved in bimanual coordination
Physiol. 11: 34. performed in the presence or absence of
Avivi-Arber, L. and Sessle, B.J. (2018). Jaw augmented visual feedback. NeuroImage
sensorimotor control in healthy adults and 19: 764–776.
effects of ageing. J. Oral Rehabil. 45: 50–80. Della-Maggiore, V., Landi, S.M., and Villalta,
Avivi-Arber, L., Lee, J.C., and Sessle, B.J. (2015). J.I. (2015). Sensorimotor adaptation: multiple
Dental occlusal changes induce motor cortex forms of plasticity in motor circuits.
neuroplasticity. J. Dent. Res. 94: 1757–1764. Neuroscientist 21: 109–125.
Avivi-Arber, L., Seltzer, Z., Friedel, M. et al. Diedrichsen, J. and Kornysheva, K. (2015).
(2017). Widespread volumetric brain changes Motor skill learning between selection and
following tooth loss in female mice. Front. execution. Trends Cogn. Sci. 19: 227–233.
Neuroanat. 10: 121. Dietsch, A.M., Westemeyer, R.M., Pearson,
Banu, R.F., Veeravalli, P.T., and Kumar, W.G. Jr., and Schultz, D.H. (2019). Genetic
V.A. (2016). Comparative evaluation of taster status as a mediator of neural activity
changes in brain activity and cognitive and swallowing mechanics in healthy adults.
function of edentulous patients, with dentures Front. Neurosci. 13: 1328.
and two-implant supported mandibular Ertekin, C. and Aydogdu, I. (2003).
overdenture-pilot study. Clin. Implant. Dent. Neurophysiology of swallowing. Clin.
Relat. Res. 18: 580–587. Neurophysiol. 114: 2226–2244.
Bostan, A.C. and Strick, P.L. (2018). The basal Fan, X., Qu, F., Wang, J.J. et al. (2017). Decreased
ganglia and the cerebellum: nodes in an gamma-aminobutyric acid levels in the
integrated network. Nat. Rev. Neurosci. 19: brainstem in patients with possible sleep
338–350. bruxism: A pilot study. J. Oral Rehabil. 44:
Chang, W.J., O’connell, N.E., Beckenkamp, 934–940.
P.R. et al. (2018). Altered primary motor cortex Feng, C.Z., Li, J.F., Hu, N. et al. (2019). Brain
structure, organization, and function in chronic activation patterns during unilateral premolar
pain: A systematic review and meta-analysis. occlusion. Cranio 37: 53–59.
J. Pain 19: 341–359. Fernández-Ruiz, J. and Díaz, R. (1999). Prism
Choi, Y.H., Jang, W.H., Im, S.U. et al. (2017). The adaptation and aftereffect: specifying the
brain activation pattern of the medial properties of a procedural memory system.
temporal lobe during chewing gum: a Learn. Mem. 6: 47–53.
functional MRI study. Neural Regen. Res. 12: Filgueiras, A., Quintas Conde, E.F., and Hall,
812–814. C.R. (2018). The neural basis of kinesthetic
and visual imagery in sports: an ALE meta - activity while wearing a palatal plate: an
analysis. Brain Imaging Behav. 12: 1513–1523. functional magnetic resonance imaging study.
Gaser, C. and Schlaug, G. (2003). Brain J. Oral Rehabil. 44: 770–778.
structures differ between musicians and Inamoto, K., Sakuma, S., Ariji, Y. et al. (2015).
non-musicians. J. Neurosci. 23: 9240–9245. Measurement of cerebral blood volume
Gazzaniga, M.S., Ivry, R.B., and Mangun, dynamics during volitional swallowing using
G.R. (2019). Cognitive Neuroscience: The functional near-infrared spectroscopy: an
Biology of the Mind. W. W. Norton & exploratory study. Neurosci. Lett. 588: 67–71.
Company. Ishibashi, R., Pobric, G., Saito, S., and Lambon
Gordon, C.L., Cobb, P.R., and Balasubramaniam, Ralph, M.A. (2016). The neural network for
R. (2018). Recruitment of the motor system tool-related cognition: an activation likelihood
during music listening: an ALE meta-analysis estimation meta-analysis of 70 neuroimaging
of fMRI data. PLoS One 13: e0207213. contrasts. Cogn. Neuropsychol. 33: 241–256.
Hadjikoutis, S., Pickersgill, T.P., Dawson, K., and Isogai, F., Kato, T., Fujimoto, M. et al. (2012).
Wiles, C.M. (2000). Abnormal patterns of Cortical area inducing chewing-like
breathing during swallowing in neurological rhythmical jaw movements and its
disorders. Brain 123 (Pt 9): 1863–1873. connections with thalamic nuclei in Guinea
Hamdy, S., Mikulis, D.J., Crawley, A. et al. pigs. Neurosci. Res. 74: 239–247.
(1999). Cortical activation during human Jang, S.H., Kwon, H.C., Kwon, H.G., and Jang,
volitional swallowing: an event-related fMRI W.H. (2015). The cortical effect of chewing
study. Am. J. Phys. 277: G219–G225. gum during hand movements: A functional
Hardwick, R.M., Caspers, S., Eickhoff, S.B., and MRI study. Somatosens. Mot. Res. 32: 110–113.
Swinnen, S.P. (2018). Neural correlates of Jestrović, I., Coyle, J.L., and Sejdić, E. (2015).
action: comparing meta-analyses of imagery, Characterizing functional connectivity
observation, and execution. Neurosci. patterns during saliva swallows in different
Biobehav. Rev. 94: 31–44. head positions. J. Neuroeng. Rehabil. 12.
Hiraba, H. and Sato, T. (2005). Cortical control Jestrović, I., Coyle, J.L., Perera, S., and Sejdić,
for mastication in cats: changes in masticatory E. (2016). Functional connectivity patterns of
movements following lesions in the normal human swallowing: difference among
masticatory cortex. Somatosens. Mot. Res. 22: various viscosity swallows in normal and
171–181. chin-tuck head positions. Brain Res. 1652:
Hirose, S., Nambu, I., and Naito, E. (2018). 158–169.
Cortical activation associated with motor Jestrović, I., Coyle, J.L., Perera, S., and Sejdić,
preparation can be used to predict the freely E. (2018). Influence of attention and bolus
chosen effector of an upcoming movement volume on brain organization during
and reflects response time: an fMRI decoding swallowing. Brain Struct. Funct. 223: 955–964.
study. NeuroImage 183: 584–596. Jiang, H., Liu, H., Liu, G. et al. (2015). Analysis
Humbert, I.A. and German, R.Z. (2013). New of brain activity involved in chewing-side
directions for understanding neural control in preference during chewing: an fMRI study.
swallowing: the potential and promise of J. Oral Rehabil. 42: 27–33.
motor learning. Dysphagia 28: 1–10. Jing, Y.H., Lin, T., Li, W.Q. et al. (2020).
Humbert, I.A., McLaren, D.G., Kosmatka, Comparison of activation patterns in Mirror
K. et al. (2010). Early deficits in cortical neurons and the swallowing network during
control of swallowing in Alzheimer’s disease. action observation and execution: A task-
J. Alzheimers Dis. 19: 1185–1197. based fMRI study. Front. Neurosci. 14: 867.
Inamochi, Y., Fueki, K., Usui, N. et al. (2017). Kamarunas, E., Mulheren, R., Palmore, K., and
Adaptive change in chewing-related brain Ludlow, C. (2018). Timing of cortical
Løkkegaard, A., Herz, D.M., Haagensen, Martino, R., Foley, N., Bhogal, S. et al. (2005).
B.N. et al. (2016). Altered sensorimotor Dysphagia after stroke: incidence, diagnosis,
activation patterns in idiopathic dystonia-an and pulmonary complications. Stroke 36:
activation likelihood estimation meta-analysis 2756–2763.
of functional brain imaging studies. Hum. Masuda, Y., Kim, S.K., Kato, T. et al. (2005).
Brain Mapp. 37: 547–557. Different corticostriatal projections from two
Lotze, M., Domin, M., and Kordass, B. (2017). parts of the cortical masticatory area in the
Symmetry of fMRI activation in the primary rabbit. Exp. Brain Res. 161: 397–404.
sensorimotor cortex during unilateral Matsuo, K. and Palmer, J.B. (2008). Anatomy and
chewing. Clin. Oral Investig. 21: 967–973. physiology of feeding and swallowing: normal
Lowell, S.Y., Poletto, C.J., Knorr-Chung, and abnormal. Phys. Med. Rehabil. Clin.
B.R. et al. (2008). Sensory stimulation activates N. Am. 19: 691–707, vii.
both motor and sensory components of the Matsuzaka, Y. and Tanji, J. (1996). Changing
swallowing system. NeuroImage 42: 285–295. directions of forthcoming arm movements:
Lowell, S.Y., Reynolds, R.C., Chen, G. et al. neuronal activity in the presupplementary and
(2012). Functional connectivity and laterality supplementary motor area of monkey cerebral
of the motor and sensory components in the cortex. J. Neurophysiol. 76: 2327–2342.
volitional swallowing network. Exp. Brain Res. Matyas, F., Sreenivasan, V., Marbach, F. et al.
219: 85–96. (2010). Motor control by sensory cortex.
Lund, J.P. (1991). Mastication and its control by Science 330: 1240–1243.
the brain stem. Crit. Rev. Oral Biol. Med. Memarian, N., Ferrari, P., Macdonald, M.J. et al.
2: 33–64. (2012). Cortical activity during speech and
Lund, J.P. and Kolta, A. (2006). Generation of non-speech oromotor tasks: a
the central masticatory pattern and its magnetoencephalography (MEG) study.
modification by sensory feedback. Dysphagia Neurosci. Lett. 527: 34–39.
21: 167–174. Miller, A.J. (2008). The neurobiology of
Lund, J.P., Sasamoto, K., Murakami, T., and swallowing and dysphagia. Dev. Disabil. Res.
Olsson, K.A. (1984). Analysis of rhythmical Rev. 14: 77–86.
jaw movements produced by electrical Molenberghs, P., Cunnington, R., and
stimulation of motor-sensory cortex of rabbits. Mattingley, J.B. (2012). Brain regions with
J. Neurophysiol. 52: 1014–1029. mirror properties: a meta-analysis of 125
Luraschi, J., Korgaonkar, M.S., Whittle, T. et al. human fMRI studies. Neurosci. Biobehav. Rev.
(2013). Neuroplasticity in the adaptation to 36: 341–349.
prosthodontic treatment. J. Orofac. Pain 27: Moore, J.D., Kleinfeld, D., and Wang, F. (2014).
206–216. How the brainstem controls orofacial
Maezawa, H., Mima, T., Yazawa, S. et al. (2014). behaviors comprised of rhythmic actions.
Contralateral dominance of corticomuscular Trends Neurosci. 37: 370–380.
coherence for both sides of the tongue during Mosier, K., Patel, R., Liu, W.C. et al. (1999).
human tongue protrusion: an MEG study. Cortical representation of swallowing in
NeuroImage 101: 245–255. normal adults: functional implications.
Marder, E. and Bucher, D. (2001). Central Laryngoscope 109: 1417–1423.
pattern generators and the control of rhythmic Mulheren, R.W. and Ludlow, C.L. (2017).
movements. Curr. Biol. 11: R986–R996. Vibration over the larynx increases
Martin, R.E., Goodyear, B.G., Gati, J.S., and swallowing and cortical activation for
Menon, R.S. (2001). Cerebral cortical swallowing. J. Neurophysiol. 118: 1698–1708.
representation of automatic and volitional Mulheren, R.W., Kamarunas, E., and Ludlow,
swallowing in humans. J. Neurophysiol. 85: C.L. (2016). Sour taste increases swallowing
938–950. and prolongs hemodynamic responses in the
cortical swallowing network. J. Neurophysiol. Panico, F., Rossetti, Y., and Trojano, L. (2020).
116: 2033–2042. On the mechanisms underlying prism
Nachev, P., Kennard, C., and Husain, M. (2008). adaptation: A review of neuro-imaging and
Functional role of the supplementary and neuro-stimulation studies. Cortex 123: 57–71.
pre-supplementary motor areas. Nat. Rev. Park, C.H., Chang, W.H., Lee, M. et al. (2015).
Neurosci. 9: 856–869. Which motor cortical region best predicts
Nagashima, S., Kimoto, K., Ono, Y. et al. (2020). imagined movement? NeuroImage 113: 101–110.
The effect of masticatory behaviour on Perumal, P., Chander, G.N., Anitha, K.V. et al.
generalized attention in heathy volunteers. (2016). Power spectrum density analysis for
Psychogeriatrics 20: 254–261. the influence of complete denture on the
Nakamura, Y. and Katakura, N. (1995). brain function of edentulous patients -pilot
Generation of masticatory rhythm in the study. J. Adv. Prosthodont. 8: 187–193.
brainstem. Neurosci. Res. 23: 1–19. Picard, N. and Strick, P.L. (2001). Imaging the
Narita, N., Ishii, T., Iwaki, S. et al. (2019a). premotor areas. Curr. Opin. Neurobiol. 11:
Prefrontal consolidation and compensation as 663–672.
a function of wearing denture in partially Pruszynski, J.A., Kurtzer, I., Nashed, J.Y. et al.
edentulous elderly patients. Front. Aging (2011). Primary motor cortex underlies
Neurosci. 11: 375. multi-joint integration for fast feedback
Narita, N., Kamiya, K., Makiyama, Y. et al. control. Nature 478: 387–390.
(2019b). Prefrontal modulation during chewing Quintero, A., Ichesco, E., Myers, C. et al.
performance in occlusal dysesthesia patients: a (2013a). Brain activity and human unilateral
functional near-infrared spectroscopy study. chewing: an FMRI study. J. Dent. Res. 92:
Clin. Oral Investig. 23: 1181–1196. 136–142.
Nishino, T. (2012). The swallowing reflex and its Quintero, A., Ichesco, E., Schutt, R. et al.
significance as an airway defensive reflex. (2013b). Functional connectivity of human
Front. Physiol. 3: 489. chewing: an fcMRI study. J. Dent. Res. 92:
Onozuka, M., Fujita, M., Watanabe, K. et al. 272–278.
(2002). Mapping brain region activity during Rizzolatti, G. and Sinigaglia, C. (2016). The
chewing: a functional magnetic resonance mirror mechanism: a basic principle of brain
imaging study. J. Dent. Res. 81: 743–746. function. Nat. Rev. Neurosci. 17: 757–765.
Onozuka, M., Fujita, M., Watanabe, K. et al. Rosenzweig, M.R., Breedlove, S.M., and Leiman,
(2003). Age-related changes in brain regional A.L. (2002). Biological Psychology. Sunderland,
activity during chewing: a functional M.A.: Sinauer Associates, Inc.
magnetic resonance imaging study. J. Dent. Sakamoto, K., Nakata, H., Yumoto, M. et al.
Res. 82: 657–660. (2015). Mastication accelerates go/no-go
Ozdiler, O., Orhan, K., Cesur, E. et al. (2019). decisional processing: an event-related potential
Evaluation of temporomandibular joint, study. Clin. Neurophysiol. 126: 2099–2107.
masticatory muscle, and brain cortex activity Saker, P., Farrell, M.J., Egan, G.F. et al. (2016).
in patients treated by removable functional Overdrinking, swallowing inhibition, and
appliances: a prospective fMRI study. regional brain responses prior to swallowing.
Dentomaxillofac. Radiol. 48 (7): 20190216. Proc. Natl. Acad. Sci. U. S. A. 113:
https://doi.org/10.1259/dmfr.20190216. Epub 12274–12279.
2019 Jul 24. Schmahmann, J.D., Guell, X., Stoodley, C.J., and
Panico, F., Sagliano, L., Grossi, D., and Trojano, Halko, M.A. (2019). The theory and
L. (2016). Cerebellar cathodal tDCS interferes neuroscience of cerebellar cognition. Annu.
with recalibration and spatial realignment Rev. Neurosci. 42: 337–364.
during prism adaptation procedure in healthy Schubotz, R.I. and Von Cramon, D.Y. (2001).
subjects. Brain Cogn. 105: 1–8. Functional organization of the lateral
premotor cortex: fMRI reveals different Suntrup-Krueger, S., Kemmling, A., Warnecke,
regions activated by anticipation of object T. et al. (2017). The impact of lesion location on
properties, location and speed. Brain Res. dysphagia incidence, pattern and complications
Cogn. Brain Res. 11: 97–112. in acute stroke. Part 2: oropharyngeal residue,
Sclocco, R., Beissner, F., Bianciardi, M. et al. swallow and cough response, and pneumonia.
(2018). Challenges and opportunities for Eur. J. Neurol. 24: 867–874.
brainstem neuroimaging with ultrahigh field Tang, Q., Li, G., Liu, T. et al. (2015). Modulation
MRI. NeuroImage 168: 412–426. of interhemispheric activation balance in
Seidler, R.D., Noll, D.C., and Chintalapati, motor-related areas of stroke patients with
P. (2006). Bilateral basal ganglia activation motor recovery: systematic review and
associated with sensorimotor adaptation. Exp. meta-analysis of fMRI studies. Neurosci.
Brain Res. 175: 544–555. Biobehav. Rev. 57: 392–400.
Seidler, R.D., Gluskin, B.S., and Greeley, Teghil, A., Boccia, M., D’Antonio, F. et al. (2019).
B. (2017). Right prefrontal cortex transcranial Neural substrates of internally-based and
direct current stimulation enhances multi-day externally-cued timing: an activation likelihood
savings in sensorimotor adaptation. estimation (ALE) meta-analysis of fMRI
J. Neurophysiol. 117: 429–435. studies. Neurosci. Biobehav. Rev. 96: 197–209.
Sessle, B.J. (2019). Can you be too old for oral Toogood, J.A., Smith, R.C., Stevens, T.K. et al.
implants? An update on ageing and plasticity (2017). Swallowing preparation and execution:
in the oro-facial sensorimotor system. J. Oral insights from a delayed-response functional
Rehabil. 46: 936–951. magnetic resonance imaging (fMRI) study.
Shibusawa, M., Takeda, T., Nakajima, K. et al. Dysphagia 32: 526–541.
(2009). Functional near-infrared spectroscopy Tramonti Fantozzi, M.P., Diciotti, S., Tessa,
study on primary motor and sensory cortex C. et al. (2019). Unbalanced occlusion
response to clenching. Neurosci. Lett. modifies the pattern of brain activity during
449: 98–102. execution of a finger to thumb motor task.
Shimazaki, T., Otsuka, T., Akimoto, S. et al. Front. Neurosci. 13: 499.
(2012). Comparison of brain activation via Trulsson, M. (2006). Sensory-motor function of
tooth stimulation. J. Dent. Res. 91: 759–763. human periodontal mechanoreceptors. J. Oral
Sokolov, A.A., Miall, R.C., and Ivry, R.B. (2017). Rehabil. 33: 262–273.
The cerebellum: adaptive prediction for Tseng, Y.W., Diedrichsen, J., Krakauer, J.W. et al.
movement and cognition. Trends Cogn. Sci. (2007). Sensory prediction errors drive
21: 313–332. cerebellum-dependent adaptation of reaching.
Sörös, P., Inamoto, Y., and Martin, R.E. (2009). J. Neurophysiol. 98: 54–62.
Functional brain imaging of swallowing: an Umiltà, M.A., Kohler, E., Gallese, V. et al. (2001).
activation likelihood estimation meta- I know what you are doing. A
analysis. Hum. Brain Mapp. 30: 2426–2439. neurophysiological study. Neuron 31: 155–165.
Sörös, P., Schafer, S., and Witt, K. (2020). Viggiano, A., Manara, R., Conforti, R. et al.
Model-based and model-free analyses of the (2015). Mastication induces long-term
neural correlates of tongue movements. Front. increases in blood perfusion of the trigeminal
Neurosci. 14: 226. principal nucleus. Neuroscience 311: 75–80.
Stoodley, C.J. (2012). The cerebellum and Wiech, K., Jbabdi, S., Lin, C.S. et al. (2014).
cognition: evidence from functional imaging Differential structural and resting state
studies. Cerebellum 11: 352–365. connectivity between insular subdivisions and
Stoodley, C.J. and Schmahmann, J.D. (2010). other pain-related brain regions. Pain 155:
Evidence for topographic organization in the 2047–2055.
cerebellum of motor control versus cognitive Wilmskoetter, J., Bonilha, L., Martin-Harris,
and affective processing. Cortex 46: 831–844. B. et al. (2019). Mapping acute lesion locations
Dental Neuroimaging: The Role of the Brain in Oral Functions, First Edition. Chia-Shu Lin.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
Companion website: www.wiley.com/go/lin/dental-neuroimaging
Figure 5.1 Processing of oral somatosensory information. Information from individual sensory modalities
is transduced via peripheral receptors at the level of somatosensation and integrated to form a holistic
perceptual experience (e.g. oral stereognosis) at the level of somatoperception. Information is further
integrated, with knowledge and affective–motivational experience, to form a feeling of intraoral condition
at the level of somatorepresentation.
behaviour, i.e. showing a sustained neural (Figure 5.2a). The high sensitivity to the force
response when being activated, has been loaded on a tooth may be associated with the
repeatedly recorded in human subjects fact that during eating, we usually hold the
when their tooth was bearing a force food piece between opposite teeth so that we
(Trulsson and Johansson 1996). can sense its hardness. The force of contrac-
tion of masticatory muscles for jaw motion is
5.1.3.2 Mechanoreceptors of Teeth and Tongue associated with the sensory information fed
Using microneurography, Trulsson and by the periodontal mechanoreceptors. The
Essick (1997) have investigated the signals findings reveal a great diversity of the physi-
from the lingual nerve by assessing different ological properties of oral mechanoreceptors
regions of the human tongue. Two distinct in different intraoral sites.
groups of receptors were identified: the The role of sensory feedback from periodon-
superficial units were located near the sur- tal mechanoreceptors can also be demon-
face of the tongue, which fired signals with a strated by the study by the manipulation task.
low threshold of force and responded to a During the task, subjects were asked to occlude
smaller receptive field. In contrast, the deep and hold a piece of food (e.g. a candy) in
units were located inside the muscle mass of between the upper and lower incisors. They
the tongue. They responded to a larger recep- were asked to split the candy evenly into two
tive field with a higher threshold of force parts. An ideal split means that the candy is
(Trulsson and Essick 1997). Also, using the split into two parts with the same weight
microneurography technique, they found (Figure 5.2b). The larger deviation between the
that the receptive field of periodontal mecha- weight of the larger parts split and the weight
noreceptors extends to the adjacent teeth. from an ideal split is, the worse the perfor-
The periodontal mechanoreceptors respond mance is. Grigoriadis et al. (2017) found that
to force applied in a broad range of directions when dentate subjects were locally anesthe-
(Trulsson 2006). The discharge of neural sig- tized in their incisors (i.e. to reduce sensory
nals is with the highest sensitivity when the feedback from teeth), their performance of the
loading force is below 1 N in anterior teeth splitting task became worse. However, the
Figure 5.2 Experimental methods of investigating oral mechanoreceptors. (a) Recording signals from
periodontal mechanoreceptors. Source: Trulsson (2006). Reproduced with permission of John Wiley and Sons.
(b) The food splitting task. Source: Grigoriadis et al. (2017) with permission of Springer Nature under the
terms of the Creative Commons CC BY 4.0 License.
5.1.5.2 Somaesthesis: The Feeling of the Mouth findings suggest that brain mechanisms may
Somaesthesis refers to a highly integrative and play a key role in the interaction between sen-
multisensory experience that consists of not sory (food content) and affective (pleasantness)
only touch, chemical or gustatory information, experience of eating. In general, somaesthesis is
but also the appetitive status and affective feel- a holistic experience that relates to not only the
ing of the mouth (Haggard and De Boer 2014). A sensory pathway but also the brain mechanisms
pleasant feeling may arise when we are having a of emotion and motivation.
meal. In a functional magnetic resonance imag-
ing (fMRI) study, Grabenhorst and Rolls (2014)
5.1.6 Neuroimaging Research on Oral
investigated brain activation when subjects con-
Somatosensory Processing
sumed high vs. low-fat food with more vs. less
pleasant flavour. They found a strong functional In the domain related to sensory function, there
correlation between the somatosensory cortex have been relatively fewer studies published for
and the orbitofrontal cortex (OFC) when sub- oral somatosensation, compared to orofacial
jects consumed high-fat food with a pleasant fla- pain, probably due to the huge clinical impact of
vour. Moreover, subjective ratings of texture the latter. Still, there have been gradually more
pleasantness were associated with the functional neuroimaging studies published on oral soma-
coupling between the somatosensory cortex and tosensory processing in recent years (Table 5.1),
the OFC (Grabenhorst and Rolls 2014). The as discussed in the following sections.
Table 5.1 Neuroimaging research on brain mechanisms of oral somatosensory processing (since 2015).
Notes: EEG: electroencephalography; EMG: electromyography; fMRI: functional magnetic resonance imaging; fNIRS:
functional near-infrared spectroscopy; MEG: magnetoencephalography; PDAP: persistent dentoalveolar pain disorder.
contrast, the oral task is distinctly associated (Trulsson 2006). In patients who have natural
with the activation of the insula (Fujii teeth replaced by dental implants, the perio-
et al. 2011). The common activation at the pre- dontal mechanoreceptors are mostly removed
motor and the primary sensorimotor area may during extraction. Therefore, the sensory feed-
signify the role of sensorimotor learning (see back during occlusion may be altered. In an
Section 4.1) in the stereognostic test, which earlier study, Grigoriadis et al. (2011) investi-
may be challenging for subjects. The SMG may gated 13 patients with implant-supported
participate multi-modal sensory processing bridges, who had used them for at least one
(see Section 5.4). The activation of the PFC, year, for their performance of chewing hard
including activation at the dorsolateral PFC, is and soft food. They found that the patients
widely found for a task with a greater demand with dental implants and the subjects with
of work memory and attention (Lara and natural teeth did not significantly differ in the
Wallis 2015). Consistently, during the stereog- number of chewing cycles. The dentate sub-
nostic tests, subjects needed to focus on their jects, but not the patients, showed a significant
tactile sensation to give an accurate response. change in the amplitude and velocity of jaw
There is also activation identified in the fron- movement when chewing hard food compared
topolar regions, which are associated with cog- to soft food. The change in jaw movements
nitive control of multiple tasks (Mansouri suggested a better adaptation for food hardness
et al. 2017). The activation may reflect the in dentate subjects, but not patients with den-
nature of stereognosis, which demands an tal implants. Consistently, when chewing hard
evaluation of multiple aspects (e.g. size, shape food, the patients showed a weaker increase in
and texture) of an object. In general, the find- EMG activity, suggesting worse adaptation of
ings suggest that oral stereognosis, as an inte- the muscle activity to food hardness during
grative experience related to oral mastication compared to dentated subjects
somatoperception, may be associated with the (Grigoriadis et al. 2011).
brain mechanisms of both sensory and cogni- Nevertheless, the loss of periodontal mecha-
tive processing. noreceptors may not completely silence any
sensory information from the implant. Sensory
feedback of dental implants may be associated
5.1.7 Clinical Implications
with ‘osseoperception’, which generally refers
The oral somatosensory processing previously to as ‘a conscious perception of external stim-
discussed should not be taken as an independ- uli transmitted via a bone-anchored prosthesis
ent phenomenon unrelated to other oral func- by activation of neural endings and/or recep-
tions. On the contrary, somatosensory tors in the peri-implant environment’ (Abarca
processing is an essential component for more et al. 2006). Therefore, osseoperception is con-
complicated oral functions. For example, the sidered as the assembly of sensory feedback
sensory feedback from teeth and mucosa is derived from mechanoreceptors in muscle,
critical for maintaining efficient mastication. joint, mucosal, cutaneous, and periosteal tis-
In the following sections, we outline three clin- sues (Klineberg et al. 2005). So far, neuroimag-
ical topics that are associated with the process- ing findings regarding osseoperception are still
ing of oral somatosensation, i.e. sensory lacking. Habre-Hallage et al. (2012) compared
feedback of dental implants, occlusal dysesthe- the brain activation of tactile stimulation on
sia and perceptual distortion of the oral cavity. teeth in patients with implants and subjects
with natural teeth. They found that stimulat-
5.1.7.1 Sensory Feedback of Dental Implants ing implants activated an extensive bilateral
The periodontal mechanoreceptors play a key cortical network outside the somatosensory
role in sensing the force loaded on teeth areas, including the parietal, frontal and
insular lobes (Habre-Hallage et al. 2012). A induced by local anaesthesia, on 52 healthy
recent functional near-infrared spectroscopy adults. Local anaesthesia was performed on
(fNIRS) study also revealed activation of the the midface region innervated by the infraor-
S1 when subjects received vibratory tactile bital nerve, and subjects were asked to report
stimuli on both natural teeth and dental the perceived change in the size of the swol-
implants (Sekido et al. 2020). Due to the lack of len area compared to the unaffected side.
psychophysical (e.g. the intensity of loading Based on single-blinded randomization, half
force) or functional (e.g. chewing perfor- of the subjects received inhibitory rTMS at
mance) data, it is difficult to clarify the precise the somatotopic area of the S1 that represents
brain mechanisms of osseoperception. their right face/lip after local anaesthesia. The
researchers found a decreased magnitude
5.1.7.2 Occlusal Dysesthesia in perceptual distortion immediately and
Occlusal dysesthesia (OD) refers to ‘a condi- 20 minutes after inhibitory rTMS compared to
tion in which tooth contacts that are not clini- sham rTMS. Moreover, the effect was not
cally identifiable as premature contacts nor induced by stimulation at the peripheral mus-
associated with other disorders’ but perceived cle or the cortical region other than the face/
as ‘disturbing or unpleasant’ for more than six lip area (Kothari et al. 2020). The findings
months (Imhoff et al. 2020). For diagnosis of strengthen the link between cortical process-
OD, symptoms derived from the disorders of ing and individual experience of body image
the periodontal tissues, the dental pulp, the by highlighting the specificity between the S1
masticatory muscles or the temporomandibu- orofacial area and the location of experienced
lar joint need to be excluded (Imhoff distortion.
et al. 2020). The neural mechanisms of OD
have not been elucidated. Using fNIRS, Narita
5.1.8 Summary
et al. (2019) found that both OD patients and
healthy controls showed increased PFC activ- ●● According to the model of body representa-
ity during chewing compared to rest. However, tion, the processing of oral somatosensa-
OD patients showed a lower PFC activity dur- tion follows a hierarchical organization:
ing chewing compared to healthy controls the processing of somatosensation is asso-
(Narita et al. 2019). Notably, OD may pre- ciated with the transduction of physical
sent comorbidity with psychiatric disorders stimuli via peripheral sensory receptors.
(Shinohara et al. 2020), and sleep restriction Somatoperception is associated with the
would influence the affective experience of perception of an intraoral object via soma-
occlusal sensation (Nishimori et al. 2019). tosensory information.
Therefore, OD may not simply reflect a change ●● Somaesthesis is a highly integrative and
in somatosensory processing. The brain mech- multisensory experience that consists of not
anisms of cognitive–affective processing would only touch, chemical and gustatory informa-
play a key role in the development of OD. tion but also the appetitive status and affec-
tive feeling of the mouth.
5.1.7.3 Orofacial Perceptual Distortion ●● Oral somatosensation, as induced by tactile
Perpetual distortion is a multisensory phe- stimuli, is generally associated with an
nomenon commonly reported by patients as increased brain activation at the S1. Oral
the feeling of ‘swollen’ on their painful face stereognosis, as an integrative experience
area (Dagsdóttir et al. 2018). In a pioneering related to oral somatoperception, may be
study of this topic, Kothari et al. (2020) used associated with the brain mechanisms of
repetitive TMS to manipulate the experience both sensory and attentional–cognitive
of perceptual distortion, as experimentally processing.
(Scott 2005). Evidence from animal research affective–motivational (i.e. valence) level.
also confirmed new percepts of taste. For Consistently, the findings from an earlier
example, taste receptor cells are found associ- meta-analysis also showed consistent activa-
ated with the taste of fat, which may signify the tion at the insula when subjects perceived pure
richness of energy (Gilbertson 1998), and the taste stimuli (Veldhuizen et al. 2011). The neu-
taste of umami, which is induced by amino roimaging findings also confirm the conclu-
acids (Zhao et al. 2003). Finally, researchers sion from animal research, which identified
have identified the insular cortex as the pri- the insular cortex as the location of the gusta-
mary gustatory cortex (Maffei et al. 2012). tory cortex (Maffei et al. 2012). In addition,
Recent findings reveal that in the gustatory neuroimaging studies have provided addi-
cortex, some neurons respond only to taste tional clues about gustatory processing by
stimuli (i.e. a ‘unimodal’ neuron), and there revealing a different pattern of activation in
exist the ‘bimodal neurons’ that respond to the insular subdivisions (Yeung et al. 2018)
both tastants and odorants dissolved in water (Figure 5.3). The meta-analysis revealed that
(Samuelsen and Fontanini 2017). The findings the affective value of taste was associated with
from animal research elucidate the multisen- a consistent pattern of activation in the middle
sory processing of gustatory and olfactory and anterior insula. In contrast, the quality of
information. taste was primarily associated with activation
in the mid-dorsal insula (Yeung et al. 2018). basic qualities of taste (i.e. sweet, salty, bitter
The distinct pattern of activation corresponds and sour), confirming the role of the insula
to the functional and structural connectivity of in gustation. Compared to other taste stimuli,
the insula. The anterior and middle insula has bitter and umami may be associated with
a stronger connection with the PFC and the activation at the primary gustatory cortex
orbitofrontal cortex, primarily for cognitive– and subcortical regions (Meyer-Gerspach
affective processing. In contrast, the posterior et al. 2016). Compared to bitter, sweet may be
and middle insula has a stronger connection associated with activation at the hippocam-
with the area for sensory processing (Wiech pus, somatosensory cortex and orbitofrontal
et al. 2014). Furthermore, the insula does not cortex (Sadler et al. 2020). Notably, the brain
merely receive the gustatory information in mechanisms may vary depending on individ-
bottom-up processing. A recent fMRI study ual dispositions of gustation. For example,
revealed that taste intensity was associated when receiving umami stimuli, subjects with
with top-down (i.e., from the insula to the higher ability of identification of umami
thalamus) information processing (Yeung would show increased activation at the pri-
et al. 2016). Together, the findings highlight mary gustatory cortex compared to those
multidimensional processing, including sen- with a lower ability (Han et al. 2018). fNIRS
sory, cognitive, affective and motivational findings revealed that the addition of soy
aspects, of gustation. sauce and monosodium glutamate (MSG),
which induces umami taste, is associated
with brain activity at the frontal operculum
5.2.3 Recent Neuroimaging Findings
(Onuma et al. 2018). Moreover, the addition
of Gustation
of glutathione enhanced the taste intensity of
Recent neuroimaging research has been focus- the MSG. The effect of glutathione was asso-
ing on the brain mechanisms of the qualitative ciated with activation at the left ventral
experience of taste (e.g. sweet or bitter). As insula (Goto et al. 2016). Therefore, func-
shown in Table 5.2, in the studies, specific tional brain features may reflect the change
tastants were used to induce distinct qualities of composition of taste stimuli and highlight
of taste. In our daily life, the taste is perceived the role of gustation in discriminating and
from consuming real food rather than the selecting food nutrients.
experimental tastants. The research scope has
been extended to the relationship between gus- 5.2.3.2 Brain Mechanisms Associated with the
tation and food processing. Affective–Motivational Experience of Taste
It is noteworthy that in the neuroimaging stud-
5.2.3.1 Brain Mechanisms Associated ies of taste stimuli, brain activation was found
with Taste Stimuli in an extended area beyond the insula and the
Recent neuroimaging findings support the somatosensory areas. The brain mechanisms
notion that brain features may discriminate may reflect the affective–motivational process-
the different taste stimuli (Table 5.2). For ing of gustation. For example, when consum-
example, the brain activation at the insula ing food, individuals may feel an emotional
(Chikazoe et al. 2019) and the EEG activity experience, such as feeling the food tasty or
of the delta-frequency range (Wallroth bland. The emotional experience is associated
et al. 2018) would be associated with the with the motivation to pursue or avoid food. In
identification of taste quality. Using ultra- a neuroimaging meta-analysis, Brown et al.
high field fMRI, Chikazoe et al. (2019) found (2011) analyzed neuroimaging studies focus-
activation in the insula and frontal and pari- ing on aesthetic appraisals of various sensory
etal operculum discriminated between four modalities, including vision, audition,
Table 5.2 Neuroimaging research on brain mechanisms of taste or food stimuli (since 2015).
Eldeghaidy et al. (2021) 12 thermal tasters and 12 fMRI/sweet stimuli delivery at cold vs. ambient
thermal non-tasters temperatures
Sadler et al. (2020) 85 healthy adults fMRI/a probabilistic selection task of sweet and
bitter stimuli
Chikazoe et al. (2019) 31 healthy adults fMRI/sweet, salty, bitter and sour stimuli
Kono et al. (2018) 18 healthy adults EEG/five basic taste stimuli and capsaicin of
different concentrations
Han et al. (2018) 15/15 subjects with high/low fMRI/umami and salty stimuli
umami identification ability
Onuma et al. (2018) 44 healthy adults fNIRS/salty stimuli w/wo umami stimuli or the
odour of soy sauce
Wallroth et al. (2018) 16 healthy adults EEG/salty, sweet, sour or bitter stimuli
Goto et al. (2016) 26 healthy adults fMRI/umami and salty stimuli w/wo glutathione
Meyer-Gerspach et al. (2016) 12 healthy adults fMRI/bitter, sweet, sour, salty and umami
stimuli; working memory processing
Hort et al. (2016) 12 thermal tasters and fMRI/sweet stimuli with no CO2 and low and
12 thermal non-tasters high CO2 levels
Mascioli et al. (2015) 11 healthy adults fMRI/taste and tactile stimulation on the
unilateral side of the tongue
Frank-Podlech et al. (2019) 11 normal-weight adults rs-fMRI/before and after having high-vs.
low-fat yogurt
Smeets and de Graaf (2019) 21 regular beer drinkers fMRI/anticipating and consuming beer,
non-alcoholic beer and carbonated water
Wistehube et al. (2018) 25 patients with chronic brain fMRI/milkshakes with different sugar or fat
lesions (25 healthy controls) content
Eiler II et al. (2018) 74 adults with a positive or fMRI/sweet stimuli of different sucrose
negative family history of concentrations
alcoholism
Shearrer et al. (2018) 53/55 adolescents who were fMRI/milkshakes with different sugar or fat
at high/low risk of obesity content
Dalenberg et al. (2017) 45 healthy adults fMRI/familiar drinking products or unfamiliar
oral nutritional supplements
Eldeghaidy et al. (2016) 17 healthy adults fMRI/flavoured no-fat control stimulus or
flavoured fat stimulus after the prior consumption
of high-fat meal or noncaloric water load
van Rijn et al. (2015) 30 healthy adults fMRI/three food stimuli (sweet/no energy,
non-sweet/energy and sweet/energy) in hunger
and satiety status
Ebisch et al. (2015) 25 healthy adults of high or fMRI/flavour stimuli with affective valences:
low emotion neutral (water), pleasant (apple nectar) or
susceptibility unpleasant (quinine)
Mazzola et al. (2017) 221 patients with epilepsy Stimulation of the insular cortex with
implanted depth electrodes
Notes: EEG: electroencephalography; fMRI: functional magnetic resonance imaging; fNIRS: functional near-infrared
spectroscopy; rs-fMRI: resting-state functional magnetic resonance imaging.
olfaction and gustation. The brain activation Graaf 2019). When investigating the percep-
when subjects appraised each of the sensory tion of sugar and fat, researchers found that
modalities was synthesized. They found that patients with brain damage showed an
when appraising the taste stimuli to be pleas- impaired fat perception but not sugar percep-
ant or attractive, consistent brain activation tion. Moreover, the impaired fat perception
was found not only in the anterior insula was associated with lesions of the anterior
(which has been consistently identified for insula and frontal operculum (Wistehube
processing of taste quality) but also the ante- et al. 2018). Notably, recent neuroimaging
rior cingulate cortex and the lateral/medial findings highlight the importance of individ-
orbitofrontal cortex (Brown et al. 2011). ual factors in the perception of food stimuli.
Notably, the authors found a common activa- When taking the solution of various concen-
tion of the anterior insula across different sen- trations of sucrose, subjects with a family his-
sory modalities, suggesting its role in the tory of alcoholism showed a greater activation
appraisal of the affective valence of sensory at the bilateral amygdala when the sucrose
experience (Brown et al. 2011). concentration was low (Eiler II et al. 2018).
Another fMRI study revealed that adolescents
with a high risk of obesity (based on parental
5.2.4 Neuroimaging Findings
obesity status) showed increased activation in
of Food Perception
caudate, gustatory and oral somatosensory
There has been an increasing number of neu- areas when taking the high-sugar milkshake
roimaging studies on food perception, which is compared to the tasteless solution (Shearrer
associated with the processing of oral soma- et al. 2018). In general, the findings confirm
tosensory and gustatory information. that the insula plays a key role in gustatory per-
Moreover, food perception can be modified by ception (Shearrer et al. 2018; Smeets and De
oral functions (e.g. mastication) (Batisse Graaf 2019; Wistehube et al. 2018) and high-
et al. 2017). Recent neuroimaging studies dis- light the role of other regions in shaping one’s
closed that the brain mechanisms of process- perception of food components, including the
ing food may be associated with the PFC and the amygdala, which both show a
components of food rather than its taste. By substantial functional and structural connec-
investigating the resting-state functional con- tion with the anterior insula (Wiech
nectivity (rsFC) before and after a high-or low- et al. 2014).
fat diet, researchers found that subjects with a Several studies focus on the association
higher sensitivity to fat, as assessed by a stand- between affective–motivational factors and food
ardized oral sensitivity test, showed a stronger perception (Table 5.2). When subjects took the
rsFC between homeostatic regions (e.g. the drinking products that they were familiar with,
hypothalamus) and limbic areas (Frank- the rating of pleasantness was associated with
Podlech et al. 2019). Another study investi- the brain network consisting of the amygdala,
gated the difference of perception and brain ventral prefrontal, insular, striatal and parahip-
activation between drinking alcoholic beer and pocampal area (Dalenberg et al. 2017). Upon
non-alcoholic beer in regular beer drinkers. subjects finished consuming a high-fat meal,
Their ratings of liking and desire to drink did decreased activation in the amygdala was iden-
not significantly differ between these two types tified when more fat stimuli were delivered. The
of beers. However, their brain activation after findings suggest the brain mechanism for an
swallowing revealed increased activation effect of ‘satiety’ after a meal (Eldeghaidy
related to the taste of alcoholic beer at the PFC et al. 2016). Here, the change of insular activa-
and the anterior insula compared to the taste tion may not solely reflect a change in taste pro-
of non-alcoholic beer (Smeets and De cessing but also a change in the motivation for
sections, there exist complicated mechanisms processing. The former highlights the neural
for perceptual processing. processing of intrinsic (personal) factors, such
as goal planning and selective attention (Engel
5.3.2.1 Bottom-up and Top-down Processing et al. 2001). The latter highlights the neural
of Perception processing of extrinsic (environmental) fac-
From an evolutionary perspective, we focus on tors, such as the physical features of stimuli
the information that is useful for increasing (Gazzaniga et al. 2019) (Figure 5.4a). It is note-
the chance to survive. Therefore, there are two worthy that the concept of top-down and bot-
major roles in perceptual processing. Firstly, it tom-up processing was originally proposed in
constructs veridical information about the the literature of visual perception to explain
environment. Secondly, what is perceived how visual information is organized (Rauss
should update our expectation about the envi- and Pourtois 2013). For example, the bottom-
ronment (Press et al. 2020b). By doing so, indi- up processing focuses on detecting the local
viduals can respond effectively to the features of a target stimulus and its difference
environment and adapt efficiently to changes from neighbouring stimuli. The top-down pro-
in the environment. Our mental functions help cessing focuses on our inner goal, which
us to construct a world in our mind based on directs how we perceive the stimuli. The differ-
the information that we sense from the envi- ence between top-down and bottom-up pro-
ronment. However, there can be a gap between cessing does not imply the directionality of
sensing and perceiving. For example, when neural pathways. Both ascending pathways
dental patients notice the presence of a (from peripheral to central nervous system)
‘forceps-like instrument’, they may focus on and descending pathways would jointly par-
the sharp edge of the instrument and associate ticipate in the processing (Rauss and
it with some invasive procedure that they have Pourtois 2013).
experienced. In other words, the perception of
‘what the instrument is like’ is associated with 5.3.2.2 Predictive Coding of Perception
not just sensory information but also one’s The concept of top-down and bottom-up pro-
expectation, memory and knowledge. In terms cessing implies a hierarchical organization of
of cognitive neuroscience, perception is associ- perceptual processing, consisting of a ‘higher-
ated with both top-down and bottom-up level’ processing of integrating information
Figure 5.4 Basic concepts of perceptual processing. (a) Top-down processing highlights the neural
processing of intrinsic (personal) factors, such as one’s goal planning, on the formation of perceptual
experience. The bottom-up processing highlights the neural processing of extrinsic (environmental) factors,
such as the physical features of stimuli, on the formation of perceptual experience. (b) In predictive coding,
the sensory inputs that we receive from the real world are compared with our prediction of the sensory
outcomes. A mismatch (i.e. ‘prediction error’) occurs when our prediction does not fit the outcome we
perceive. The prediction error is associated with attentional bias and learning. For example, we may pay
more attention to an unexpected event compared to an expected one.
(e.g. to expect what to find) and a ‘lower-level’ Sensory inputs that we receive from the real
processing of detailed stimulus information world are compared with the prediction. If a
(e.g. to find the features of an object) (Rauss mismatch (i.e. ‘prediction error’) is detected
and Pourtois 2013) (Figure 5.4a). In recent (i.e. something ‘out of expectation’ occurs), our
years, researchers have formulated theories for attention may be biased to the unexpected sen-
describing the interaction between these two sory inputs (Figure 5.4b). When something
conceptual levels. Cumulating evidence sug- unexpected is detected, the old prediction
gests that the match (or mismatch) between the needs to be updated so that new information
higher- and the lower-level processing plays a can be integrated (Press et al. 2020b). In gen-
dominant role in perception (Figure 5.4b). The eral, such perceptual processing can be deemed
nervous system is critical for matching the pre- as a process of learning driven by prediction
diction generated internally to external stimu- error (Corlett 2020; Press et al. 2020a, b).
lation (Rauss and Pourtois 2013). Our brain is
not merely an organ that passively receives 5.3.2.3 Brain Mechanisms Associated
stimuli from the world. It is a calculator or an with Perceptual Prediction
active ‘inference machine’ (Friston 2010), Our perception of the world is markedly associ-
which anticipates sensory inputs and forms ated with how we act. We can reproduce the
predictions about the incoming events (Barrett experience of an action that we have performed
and Simmons 2015). When we are sensing the by simulating it in our mind and predicting its
world, we pay more attention to the presence of potential outcomes. Just like the mechanisms
the events that we expect to find compared to of motor prediction, such a predictive mecha-
those we do not predict to find (Box 5.1). nism requires an internal forward model
(Wolpert and Flanagan 2001; Wolpert experience (Barrett and Simmons 2015). In
et al. 2011). In the case of perceptual experi- contrast, the granular regions, including the
ence, the forward model is associated with the mid and posterior insula, receive sensory input
function of the lateral premotor cortex (LPMC), from the environment and may generate the
as demonstrated by recent neuroimaging evi- signal of prediction error. Both regions form a
dence (Schubotz and Von Cramon 2003). circuitry that adjusts the internal system of our
Activation at the dorsal LPMC is more relevant body (Barrett and Simmons 2015).
with the spatial feature of an event, and activa-
tion at the ventral LPMC is more relevant with
5.3.3 Attention
the rhythmic and temporal feature of an event
(Schubotz and Von Cramon 2003). Therefore, To most people, the word ‘attention’ is part of
the LPMC may be associated with the predic- our daily language rather than academic jar-
tion of multiple aspects of perceptual experi- gon. Attention is ‘the taking possession by the
ence (Schubotz 2007; Schubotz and Von mind’, as explained more than 100 years ago by
Cramon 2003). The predictive mechanisms of William James (James 2007). From the point of
sensorimotor experience are supported by the cognitive neuroscience, the precise meaning of
histological findings from animal research. the word ‘attention’ is usually interpreted
Cumulating evidence suggests that in the according to how individuals interact with the
agranular cortex (where cortical layers are less environment. For example, when there exist
differentiated), more neurons were found with multiple objects of interest, we can selectively
their activity associated with prediction, and in focus on the object that we are mostly inter-
the granular cortex (where cortical layers are ested in (i.e. selective attention). Such selective
well differentiated), more neurons were found attention is a goal-oriented behaviour, i.e.
with their activity associated with prediction attention is directed voluntarily to our behav-
error (Barrett and Simmons 2015). The cellular ioural goals. For example, when receiving den-
mechanism may account for the role of the pri- tal scaling, patients who are worried about
mary motor cortex, which consists of mostly their periodontal health would focus on a feel-
the agranular cortex, in issuing motor predic- ing of sensitivity on their teeth. In contrast,
tions (Shipp et al. 2013). dentists would focus more on the presence of
Another critical aspect of the predictive gingival bleeding, which may signify a worse
mechanisms is the processing of interoceptive periodontal condition. Attention can also be
experience. Interoception is associated with modulated by external stimuli, such as the
afferent information from the viscera, such as condition when we are distracted by some-
the heart and the gastrointestinal tract thing salient or unexpected. For example, dur-
(Cameron 2001). It refers to ‘perception of that ing dental scaling, patients may reflexively
which is arising from inside the body’ shift their attention from their teeth to the
(Terasawa et al. 2013), such as the feeling of an voice from another room if they hear someone
increased heart rate or the bowel extension. is screaming during treatment! In general,
The processing of interoception is closely attention is associated with how we process
related to the visceromotor cortices, which information, including the sensory inputs
control our autonomic and hormonal from the world and the information from our
responses. Within the visceromotor cortices, memory, and how we respond to the environ-
the brain regions with the agranular cortex, ment (Gazzaniga et al. 2019).
including the anterior cingulate cortex, the
anterior insula, the orbitofrontal cortex and 5.3.3.1 Attention Control
the medial PFC, may be associated with the The concept of attention can be categorized
generation of predicted interoceptive from different aspects. Voluntary attention
refers to the ability to focus on an object inten- during treatment, we would expect the timing
tionally. Therefore, voluntary attention is con- when an instrument is delivered by an assis-
sidered a goal-driven process or ‘endogenous tant. According to Nobre and van Ede (2018),
attention’ (i.e. driven by one’s intrinsic goal or when we anticipate something, we pay atten-
expectation) (Gazzaniga et al. 2019). In con- tion to the temporal structure of how events
trast, reflexive attention occurs when one’s are organized. Firstly, we focus on the factors
attention is captured by an external event. that are predictive of temporal relations.
Therefore, reflexive attention is considered a During dental scaling, patients may pay more
stimulus-driven process or ‘exogenous atten- attention to the cues (e.g. a water spray) that
tion’ (i.e. driven by the stimuli from the forecast the occurrence of a certain stimulus
environment). Voluntary and reflexive atten- (e.g. tooth sensitivity). Secondly, we focus on
tion is also related to top-down and bottom-up the hazard rate that reflects the probability of
processing, respectively. The concept of atten- the occurrence of an event that has not yet
tion control focuses on how attention influ- occurred. Thirdly, we focus on the rhythmic or
ences the flow of information processing. As sequential structure of events. When stimuli
summarized by Petersen and Posner (2012), the are delivered rhythmically or with a fixed
attentional process can be referred to (i) orient- sequence, individuals anticipate the on-beat
ing attention, which selectively focuses on a events or the events that occurred in previous
specific stimulus, similar to the concept of repeats (Nobre and Van Ede 2018). Notably,
selective attention for prioritizing or filtering in mental processes of attention and anticipation
the most relevant information, (ii) executive are associated with other cognitive–affective
attention, which keeps on processing the rele- factors. For example, the fear and anxiety of
vant information regardless of distractors and dental patients relate to how they conceive the
(iii) alerting attention, which maintains an uncertainty of pain, and heightened attention
‘alertness’ or a ready state for incoming stimu- to pain may influence their perceived pain
lus (Petersen and Posner 2012; Torta et al. 2017). (Armfield 2006).
The different aspects of attention control are
commonly seen in the scenario of dental prac- 5.3.3.3 Brain Mechanisms
tice. Firstly, dentists need to focus on their of Attentional Processing
goals, e.g. the tooth to be treated. Secondly, The brain mechanisms of attentional process-
such attention needs to be maintained without ing are associated with two networks. The dor-
being distracted by other unrelated stimuli (e.g. sal attention network mainly consists of the
the walking steps from dental assistants). parietal lobe and the frontal eye field. It is asso-
Thirdly, when focusing on the current treat- ciated with voluntarily selective attention. The
ment, dentists still need to be alerted to new ventral network mainly consists of the tempo-
information that could be important to their roparietal junction and the ventral frontal cor-
goals. For example, dentists are alerted by any tex. It is associated with the involuntary
groan made by patients, which indicates their detection of salient stimuli (Corbetta and
discomfort during treatment. Shulman 2002). Broadly speaking, the dorsal
and the ventral networks, respectively, play a
5.3.3.2 Attention and Anticipation key role in the top-down and bottom-up pro-
When we pay attention to something, we focus cessing of attention (Gazzaniga et al. 2019). For
on different ‘features’ of it. For example, we example, orienting attention is engaged with
search where a dental instrument is placed (i.e. the dorsal attention network, similar to the
the spatial feature) or the size of the instru- brain mechanisms of selective attention of vis-
ment (i.e. the object feature). We also pay ual information (Corbetta and Shulman 2002;
attention to the temporal feature. For example, Raz and Buhle 2006). In contrast, the anterior
cingulate cortex, which is critical to cognitive between an emotional experience (e.g. a fear-
control, is engaged with executive attention ful feeling) and environmental stimuli (e.g.
(Raz and Buhle 2006). For goal-driven atten- encountering a threat). Importantly, as dis-
tional processing, the frontal eye field (FEF) of cussed in the following sections, emotion is
the dorsal attention network plays a critical role not merely a physiological response to the
in attending to the visual features of a stimulus. external stimuli – our emotion is also shaped
For example, when researchers intervened by how we evaluate or appraise the stimuli
the FEF using transcranial magnetic stimula- (Oatley and Johnson-Laird 2014).
tion (TMS), subjects’ motion-processing
area showed increased activity when they 5.3.4.1 Motivation and Reward Processing
were discriminating the motion of a target. At the core of motivation is the concept of
Correspondingly, their fusiform face area (FFA) incentive salience or the feeling of ‘wanting’,
showed increased activity when they were dis- which is mediated by the mesolimbic pathway
criminating gender and faces (Morishima of the brain (Berridge 2018). The pathway
et al. 2009). The findings suggested that the plays a key role in the brain mechanisms of
dorsal attention network is associated with the reward processing, derived from the ventral
top-down control specific to the domain of vis- tegmental area (VTA) to the ventral striatum,
ual features. The PFC also plays a critical role in which includes the nucleus accumbens (NAc).
the dorsal attention network. For example, in The neurotransmitter dopamine plays a major
an MEG study, subjects were asked to watch the role in the mesolimbic pathway. Earlier find-
images of a house and a face overlapped ings from animal research have confirmed the
together. They were told to either attend to the role of dopamine in reward processing.
face image or the house image. The researchers Schultz (1998) demonstrated that activation of
found that attending to faces and houses is, dopaminergic neurons was associated with
respectively, associated with brain activation at the predictability of the presence of reward
the FFA and the parahippocampal place area. rather than the amount of the reward per se.
Notably, the activation in both the cases was Dopamine responses are associated with an
associated with the synchrony of signals with unexpected outcome, such as the reception of
the inferior frontal junction. The findings sug- an unexpected reward or omission of an
gest that the PFC may bias attentional signals to expected reward. In other words, dopamine
different visual features, thus being a candidate responses may signal the prediction error of a
of the source of top-down attentional control reward (Schultz 1998). Reward processing is
(Baldauf and Desimone 2014). critical for goal-directed behaviour, and indi-
viduals are motivated to pursue a reward. For
example, patients may go to see a dentist for
5.3.4 Motivation and Emotion
toothache because the relief of pain is reward-
Just like the word attention, the words ‘moti- ing (Navratilova and Porreca 2014).
vation’ and ‘emotion’ are widely used in our Conversely, reduced interest or pleasure for a
daily language. In terms of motivation, cogni- reward, i.e. anhedonia, may be associated with
tive neuroscientists focus on the association deficits in reward processing (Husain and
between incentive salience and one’s behav- Roiser 2018).
iour of pursuing a desired goal. Increased or
decreased incentive salience relates to the 5.3.4.2 Anxiety, Fear and Threat
behaviour seeking for a reward and the lack While the concept of reward relates to the
of motivation (‘anhedonia’), respectively behaviour of approaching to the desired
(Berridge 2018). In terms of emotion, cogni- object, the concept of threat relates to the
tive neuroscientists focus on the association behaviour of avoiding an aversive object.
Figure 5.5 Experimental design of manipulating threat value associated with pain. (a) The presence of
noxious stimuli is associated with visual cues, which predict low-intensity stimuli (i.e. the square) constantly
or predict high-or low-intensity stimuli (i.e. the circle). The latter evokes higher anxiety related to pain due
to the increased uncertainty (i.e. the stimulus intensity is less predictable). Moreover, the same low-intensity
stimuli would be perceived as more painful in the high-uncertainty condition (i.e. the condition predicted
by a circle). (b) The threat value of pain is associated with the severity of noxious stimuli. Subjects receive
different instructions regarding the severity (e.g. may cause tissue damage or not) of noxious stimuli, which
are delivered at different sites (grey and black). The detection threshold of pain, i.e. the intensity that
subjects feel painful and non-painful for equal times, is determined. When subjects feel a stronger severity
of the stimuli (i.e. feeling more threatened), they would report higher anxiety towards the stimuli. Moreover,
the same stimuli (tuned at the detection threshold) would be perceived as painful more likely in the more
threatening condition (i.e. the condition with more severity) compared to the condition they regard as less
threatening.
feel more anxious about it (Brown et al. 2008; which play a key role in the anticipation
Ploghaus et al. 2001; Asmundson et al. 2007) and attentional processing of pain (Brown
(Figure 5.5a). et al. 2008, Ploghaus et al. 2001, Wiech
Moreover, recent neuroimaging evidence et al. 2010).
has disclosed that increased fear and anxiety of
pain may exacerbate the intensity of pain. An 5.3.5.2 Emotional Valence of Oral
fMRI study revealed that, when receiving the Somatosensory Stimulation
stimuli at the detection threshold of pain, sub- As noted in Section 3.2, quantitative sensory
jects rated the stimuli that they perceived more testing (QST) is widely adopted for assessing
threatening as painful and the stimuli that pain and somatosensory disturbance and has
they perceived less threatening as non-painful been applied in the diagnosis of orofacial pain.
(Wiech et al. 2010) (Figure 5.5b). The low- Conventionally, QST mainly assesses for the
intensity stimuli that were delivered with presence of pain or mechanical sensation via a
unpredictable timing would be perceived as detection test. The emotional experience of a
more painful compared to the same stimuli sensation, such as the pleasant or unpleasant
that were delivered predictably (Brown feeling associated with a stimulus, has been
et al. 2008). Neuroimaging studies have identi- largely neglected. In a recent study, Taneja
fied that increased anxiety mediated by uncer- et al. (2020) investigated the emotional experi-
tainty was associated with the activation of the ence of somatosensory stimulation in 38
anterior insula, the hippocampus and the PFC, healthy subjects who have burning pain in
of bottom-up processing, the stimulus with a subjects to receive touch, painful, auditory and
stronger salience will be aligned with other visual stimuli. They used the multivariate pat-
stimuli with a lesser salience for spatial and tern analysis to identify the pattern of brain
temporal concurrence. However, it is possible activation that can discriminate between the
that multiple stimuli are comparably salient, stimuli from different sensory modalities. The
and they are competing for information pro- results showed that fMRI responses from one
cessing. Top-down modulation, such as indi- primary sensory cortex (e.g. the primary visual
vidual expectation or intention, may facilitate cortex) can discriminate the stimuli between
the integration of multisensory inputs (Talsma other modalities (e.g. tactile vs. auditory).
et al. 2010). As discussed in the following sec- Furthermore, the responses from one primary
tions, both bottom-up and top-down modula- sensory cortex can even discriminate the differ-
tion of information processing (see Section 5.3) ent experimental conditions within the same
are critical to multisensory integration. sensory modality (e.g. to discriminate tactile
stimuli on site A vs. site B by the responses of
the primary visual cortex) (Liang et al. 2013).
5.4.3 Brain Mechanisms
These striking findings suggest that the pri-
of Multisensory Integration
mary sensory cortices may not be unisensory,
Sensory inputs of different modalities are first only for the processing of their corresponding
processed by the corresponding primary sen- sensory modalities. The findings highlight that
sory cortices. For example, the visual and audi- ‘integration’ may occur before the information
tory inputs are proceeded in the primary visual is processed in the higher cortex.
and primary auditory cortices, respectively. In The processing of multisensory integration
multisensory processing, the information from is also associated with sensorimotor learning.
the individual sensory modalities interact with In an earlier fMRI study, Butler et al. (2011)
each other (Talsma et al. 2010). At the cortex, asked subjects to actively learn the visuomotor
some neurons respond to bimodal or multi- associations between novel objects and novel
modal information. Animal studies reveal that sounds or passively observing others learn the
multisensory integration is associated with the association. They found that the brain regions
neuronal firing in response to bimodal (or mul- of audiovisual integration, including the supe-
timodal) sensory stimuli. These neurons would rior temporal sulcus (STS), showed a greater
show the strongest responses to cross-modal multisensory gain when subjects were actively
inputs. In contrast, unimodal sensory input learning than passively observing the visuo-
evokes relatively weak responses (Talsma motor association (Butler et al. 2011).
et al. 2010). In human subjects, using resting- Multisensory integration is also associated
state fMRI, researchers explored that the func- with one’s evaluation of the value of sensory
tional connectivity of the intraparietal sulcus stimuli. In another fMRI study, Pooresmaeili
(IPS) showed distinct clusters associated with et al. (2014) asked subjects to perform an audi-
visual, auditory and somatosensory networks tory and a visual task. Subjects showed an
(Anderson et al. 2010). Moreover, in a task- increased sensitivity of visual perception when
based fMRI study, they found that the IPS the auditory stimuli were associated with a
showed a distinct pattern of activation when high reward. Activation of the superior tempo-
subjects were attending to different sensory ral gyrus/STS, which is associated with multi-
modalities. The findings suggested that the IPS sensory processing, was modulated by the
plays a key role in the top-down control of mul- reward value associated with the auditory
tisensory attention. However, recent fMRI evi- stimuli (Pooresmaeili et al. 2014). The findings
dence has suggested a more complex picture. In suggest that the value of sensory stimuli may
an fMRI study, Liang et al. (2013) asked play a key role in integrating information from
different modalities. In sum, neuroimaging Gustation is associated with not only tactile
studies reveal that sensorimotor learning and sensation but also tongue movement.
reward processing may modulate how cross- Therefore, a key focus of oral multisensory
modal information is integrated. integration is the association between gusta-
tion and the sensorimotor processing of the
oral cavity. Okamoto et al. (2009) have dem-
5.4.4 Research on Multisensory
onstrated an overlap of brain activation
Integration in Oral Sensorimotor Functions
between sweet stimuli and tongue tapping
The issues of multisensory integration have movement as well as speech production.
been a critical research field in neuroscience Using fNIRS, they identified brain activity
because of the cross-modal nature of sensory associated with taste stimuli at the ventral
inputs from the world. Sensory inputs of the part of precentral and postcentral gyri and the
oral cavity also derive from multiple sensory ventrolateral PFC. Moreover, these regions
channels. During eating, the sensations of food were partly overlapped with the motor area
texture, the taste and odour of the food and associated with tongue tapping and the pre-
proprioceptive inputs from masticatory mus- frontal area associated with speech produc-
cles jointly shape our experience of chewing. tion (Okamoto et al. 2009). It is noteworthy
Until now, multisensory integration has not that the spatial overlap of brain activity, as
been a major focus in oral neuroscience observed in the study, may not be attributed to
(Table 5.3). The following section mainly the multimodal responses of neuronal activ-
focuses on the oral multisensory mechanisms ity. On the one hand, the interpretation of
related to gustation and somatosensation. neuroimaging findings at the cellular level
should be carefully made (see Section 2.1 for
5.4.4.1 Multisensory Processing Related limitations of neuroimaging). On the other
to Gustation hand, the spatial overlap may simply reflect
Gustation is associated with direct contact the engagement of a common mental func-
between taste buds and chemical stimuli. tion. For example, both taste and speech
Hort et al. (2016) 12 thermal tasters and 12 fMRI/sweet stimuli with no CO2 and low and high
thermal non-tasters CO2 levels
Henderson et al. (2016) 19 healthy adults fMRI/repetitive open–close jaw movements w/wo
the presence of ongoing pain
Kagawa et al. (2014) 11 healthy adults fNIRS/shape discrimination w/wo a test piece in
the mouth
Hilbert et al. (2014) 13 adults with dental phobia fMRI/visual and auditory stimuli of dental
(13 healthy controls) treatment
Okamoto et al. (2009) 19 healthy adults fNIRS/tasting (sweet), tongue tapping and word
fluency tasks
Ptito et al. (2005) 6 congenitally blind adults PET before and after training/using the tongue in
(5 sighted blindfolded controls) discrimination orientation task
Notes: fMRI: functional magnetic resonance imaging; fNIRS: functional near-infrared spectroscopy; PET: positron
emission tomography.
require cognitive and attentional controls. tongue, the blind subjects showed an increased
Still, the findings highlight that different oral regional cerebral blood flow (rCBF) in the
functions (tongue movement, taste and occipital cortex after the training of an orienta-
speech) may share common brain mecha- tion detection task. In contrast, the sighted
nisms. Hort et al. (2016) investigated the taste subjects did not reveal such a change in occipi-
processing of ‘thermal tasters’, who develop a tal rCBF after training (Ptito et al. 2005).
‘phantom’ taste when receiving thermal stim- Additionally, the rCBF of the occipital region
ulation on the tongue, without receiving a is positively correlated with the rCBF of the
gustatory stimulus. Compared to the non- posterior parietal lobe, an area critical for pro-
tasters, thermal tasters showed a better ability cessing spatial information of vision, in the
to discriminate the high CO2 stimuli, which blind subjects, after training (Ptito et al. 2005).
evokes trigeminal but not gustatory process- The interaction between tactile and auditory
ing. Thermal tasters and non-tasters also dif- perception has also been reported. Changes in
fered in their brain activation to taste stimuli EEG signals were associated with the produc-
when being modulating by different levels of tion of a bilabial sound /pa/ when tactile stim-
CO2 stimuli. The findings highlight the role of uli were applied concurrently on the lip (Shen
cross-modal integration of taste and trigemi- et al. 2018). Together, the findings suggest that
nal processing in thermal tasters (Hort even individuals cannot ‘see’ what happens in
et al. 2016). In another recent fMRI study, their oral cavity, and their perception of the
Eldeghaidy et al. (2021) found increased acti- intraoral conditions may still be associated
vation in the S1 and the S2 when subjects with the processing of visuospatial informa-
were perceiving cold and sweet stimuli, only tion at the cortical level.
in the thermal tasters but not in the non-
tasters. Again, the findings highlight that gus-
5.4.5 Proprioception and Pain
tation is associated with individual variations
in oral somatosensory processing. Pain and proprioception are both key issues to
normal mastication and masticatory dysfunc-
5.4.4.2 Multisensory Processing Related tions. On the one hand, proprioceptive inputs
to Somatosensation play a key role in mastication by participating
As outlined in Section 5.1, oral stereognosis is in reflexive control of jaw movement. On the
the somatosensory perception of the size and other hand, pain may arise when masticatory
shape of an intraoral object. Using fNIRS, muscles are overused, and it plays a protective
Kagawa et al. (2014) investigated the brain acti- role in limiting the activity of jaw muscles
vation of oral stereognosis. They found specific (Murray and Peck 2007) (see Section 8.3).
activation at the occipital lobe, i.e. brain However, the association between propriocep-
regions related to visual processing, when sub- tion and pain in the stomatognathic system
jects were discriminating the shape of intraoral has still remained unclear. Compared to pain,
testing objects and when they were performing there have been fewer neuroimaging studies
a sham-discrimination task (i.e. without focusing on the proprioception of muscles. In
testing objects in their mouth) (Kagawa an earlier study, Goble et al. (2012) stimulated
et al. 2014). The activation at the occipital lobe the tendons of the toes and the crest of the tibia
may suggest a potential tactile-visual cross- bone by pneumatic vibration devices. The
modal processing of oral stereognosis (Kagawa devices evoked activities of both vibrotactile
et al. 2014). Interestingly, an earlier PET study receptors and muscle spindles for the tendon
on individuals with congenital blind has dem- area but only vibrotactile receptors in the bone
onstrated a similar cross-modal mechanism. area (Goble et al. 2012). They found brain
When receiving tactile stimulation on the activation associated with proprioceptive
processing (i.e. stimulation of tendon vs. bone) 5.4.6 Future Directions of Oral
in the sensorimotor cortex and the secondary Multisensory Research
motor area, including the supplementary
At present, in oral neuroscience, fewer studies
motor area (SMA) and the PMC, as well as pre-
have focused on multisensory processing of
frontal and subcortical regions (Goble
the stomatognathic system compared to
et al. 2012). Notably, the pattern of activation is
unisensory processing. The relative lack of
common in both older and younger subjects.
investigation into multisensory issues may
Compared to the younger subjects, older sub-
reflect the challenges in experimental design.
jects showed a lower activation at the puta-
Firstly, when investigating a single sensory
men, and its white matter fractional anisotropy
modality, one only needs to compare the con-
was associated with the subjective rating of
ditions with vs. without the presence of the
joint position sense (Goble et al. 2012). In oral
stimuli. However, when stimuli of two modali-
neuroscience, most neuroimaging studies
ties are investigated, more experimental
focused on the brain activation associated with
parameters need to be considered, such as the
a chewing task, which also revealed consistent
timing and intensity of both types of stimuli.
activation in the sensorimotor cortex, the SMA
Secondly, it is noteworthy that neuroimaging
and the thalamus (Lin 2018). However, because
methods can reveal the spatial features of the
the task design of the chewing studies com-
brain regions associated with different sensory
prised of a variety of sensorimotor components
stimuli. An overlap between the regions of
(e.g. jaw movement, mechanosensation from
activation may suggest, rather than confirm,
the periodontium and mucosa, and propriocep-
the integration of multimodal sensory inputs,
tion from muscles), it is difficult to clarify if the
which should be further examined at the neu-
activation is specific to the activity of muscle
ronal level (e.g. by identifying the pattern of
spindles.
bimodal or multimodal responses) (Talsma
In addition to the central processing of mus-
et al. 2010). Finally, even some brain regions
cle proprioception, the autonomous nervous
sharing the activation from two stimulating
system (ANS) may play a key role in the asso-
tasks are identified, the ‘common regions’ may
ciation between pain and mastication for its
reflect some mental functions general to all
role in regulating physiological activities.
sensory processing. For example, increased
Firstly, in the stomatognathic system, the ANS
activation at the anterior insula may reflect the
plays a key role in regulating the tonicity of
processing of increased salience of sensory
muscle via gamma efferent (Okeson 2019).
input, regardless of its modalities (Legrain
Secondly, neuroimaging evidence has revealed
et al. 2011). In general, the issues of multisen-
that patients with chronic complex regional
sory integration in oral functions demand a
pain syndrome showed structural abnormali-
greater focus in future research.
ties encompassing emotional, autonomic and
pain perception regions, consistent with the
autonomic symptoms found in the patients
5.4.7 Summary
(Geha et al. 2008). Thirdly, the ANS closely
relates to poor sleep and emotional stress, and ●● Multisensory integration refers to the set of
poor sleep quality and psychological distress processes by which information arriving
are commonly seen in patients with temporo- from the individual sensory modalities inter-
mandibular disorders (TMD) related pain acts and influences processing in other sen-
(Slade et al. 2013). The evidence together sug- sory modalities. Our knowledge of the world
gests that an altered status of the ANS may is largely derived from the experiences of
play a key role in both orofacial pain and mas- multiple, rather than a single, sensory
ticatory dysfunction. modality.
References
Abarca, M., Van Steenberghe, D., Malevez, C., and Brown, C.A., Seymour, B., Boyle, Y. et al. (2008).
Jacobs, R. (2006). The neurophysiology of Modulation of pain ratings by expectation and
osseointegrated oral implants. A clinically uncertainty: behavioral characteristics and
underestimated aspect. J Oral Rehabil 33: 161–169. anticipatory neural correlates. Pain 135:
Anderson, J.S., Ferguson, M.A., Lopez-Larson, 240–250.
M., and Yurgelun-Todd, D. (2010). Brown, S., Gao, X., Tisdelle, L. et al. (2011).
Topographic maps of multisensory attention. Naturalizing aesthetics: brain areas for
Proc Natl Acad Sci U S A 107: 20110–20114. aesthetic appraisal across sensory modalities.
Armfield, J.M. (2006). Cognitive vulnerability: a Neuroimage 58: 250–258.
model of the etiology of fear. Clin Psychol Rev Butler, A.J., James, T.W., and James, K.H. (2011).
26: 746–768. Enhanced multisensory integration and motor
Asmundson, G.J.G., Vlaeyen, J.W.S., and reactivation after active motor learning of
Crombez, G. (2007). Understanding and audiovisual associations. J Cogn Neurosci 23:
Treating Fear of Pain. Oxford, U.K.: Oxford 3515–3528.
University Press. Cameron, O.G. (2001). Interoception: the inside
Baldauf, D. and Desimone, R. (2014). Neural story – a model for psychosomatic processes.
mechanisms of object-based attention. Science Psychosom Med 63: 697–710.
344: 424–427. Chikazoe, J., Lee, D.H., Kriegeskorte, N., and
Barrett, L.F. and Simmons, W.K. (2015). Anderson, A.K. (2019). Distinct
Interoceptive predictions in the brain. Nat Rev representations of basic taste qualities in
Neurosci 16: 419–429. human gustatory cortex. Nat Commun
Batisse, C., Bonnet, G., Eschevins, C. et al. 10: 1048.
(2017). The influence of oral health on Corbetta, M. and Shulman, G.L. (2002). Control
patients’ food perception: a systematic review. of goal-directed and stimulus-driven attention
J Oral Rehabil 44: 996–1003. in the brain. Nat Rev Neurosci 3: 201–215.
Berridge, K.C. (2018). Evolving concepts of Corlett, P. (2020). Predicting to perceive and
emotion and motivation. Front Psychol 9: 1647. learning when to learn. Trends Cogn Sci 24:
Breslin, P.A. (2013). An evolutionary perspective 259–260.
on food and human taste. Curr Biol 23: Custead, R., Oh, H., Wang, Y., and Barlow,
R409–R418. S. (2017). Brain encoding of saltatory velocity
through a pulsed pneumotactile array in the Fujii, R., Takahashi, T., Toyomura, A. et al.
lower face. Brain Res 1677: 58–73. (2011). Comparison of cerebral activation
Dagsdóttir, L.K., Bellan, V., Skyt, I. et al. (2018). involved in oral and manual stereognosis.
Multisensory modulation of experimentally J Clin Neurosci 18: 1520–1523.
evoked perceptual distortion of the face. Gazzaniga, M.S., Ivry, R.B., and Mangun,
J Oral Rehabil 45: 1–8. G.R. (2019). Cognitive Neuroscience: The
Dalenberg, J.R., Weitkamp, L., Renken, R.J. et al. Biology of the Mind. W. W. Norton & Company.
(2017). Flavor pleasantness processing in the Geha, P.Y., Baliki, M.N., Harden, R.N. et al.
ventral emotion network. PLoS One 12: e0170310. (2008). The brain in chronic CRPS pain:
Ebisch, S.J., Bello, A., Spitoni, G.F. et al. (2015). abnormal gray-white matter interactions in
Emotional susceptibility trait modulates emotional and autonomic regions. Neuron 60:
insula responses and functional connectivity 570–581.
in flavor processing. Front Behav Gilbertson, T.A. (1998). Gustatory mechanisms
Neurosci 9: 297. for the detection of fat. Curr Opin Neurobiol 8:
Eiler, W.J.A. 2nd, Dzemidzic, M., Soeurt, 447–452.
C.M. et al. (2018). Family history of Goble, D.J., Coxon, J.P., Van Impe, A. et al.
alcoholism and the human brain response to (2012). The neural basis of central
oral sucrose. Neuroimage Clin 17: 1036–1046. proprioceptive processing in older versus
Eldeghaidy, S., Marciani, L., Hort, J. et al. (2016). younger adults: an important sensory role for
Prior consumption of a fat meal in healthy right putamen. Hum Brain Mapp 33: 895–908.
adults modulates the brain’s response to fat. Goto, T.K., Yeung, A.W., Tanabe, H.C. et al.
J Nutr 146: 2187–2198. (2016). Enhancement of combined umami
Eldeghaidy, S., Yang, Q., Abualait, T. et al. and salty taste by glutathione in the human
(2021). Thermal taster status: temperature tongue and brain. Chem Senses 41: 623–630.
modulation of cortical response to sweetness Grabenhorst, F. and Rolls, E.T. (2014). The
perception. Physiol Behav 230: 113266. representation of oral fat texture in the human
Engel, A.K., Fries, P., and Singer, W. (2001). somatosensory cortex. Hum Brain Mapp 35:
Dynamic predictions: oscillations and 2521–2530.
synchrony in top-down processing. Nat Rev Grigoriadis, A., Johansson, R.S., and Trulsson,
Neurosci 2: 704–716. M. (2011). Adaptability of mastication in
Engelen, L., Van Der Bilt, A., and Bosman, people with implant-supported bridges. J Clin
F. (2004). Relationship between oral Periodontol 38: 395–404.
sensitivity and masticatory performance. Grigoriadis, J., Kumar, A., Svensson, P. et al.
J Dent Res 83: 388–392. (2017). Perturbed oral motor control due to
Ettlin, D.A., Zhang, H., Lutz, K. et al. (2004). anesthesia during intraoral manipulation of
Cortical activation resulting from painless food. Sci Rep 7: 46691.
vibrotactile dental stimulation measured by Grigoriadis, A., Kumar, A., Åberg, M.K., and
functional magnetic resonance imaging Trulsson, M. (2019). Effect of sudden
(FMRI). J Dent Res 83: 757–761. deprivation of sensory inputs from
Frank-Podlech, S., Heinze, J.M., Machann, periodontium on mastication. Front Neurosci
J. et al. (2019). Functional connectivity within 13: 1316.
the gustatory network is altered by fat content Grupe, D.W. and Nitschke, J.B. (2013).
and oral fat sensitivity -a pilot study. Front Uncertainty and anticipation in anxiety: an
Neurosci 13: 725. integrated neurobiological and psychological
Friston, K. (2010). The free-energy principle: a perspective. Nat Rev Neurosci 14: 488–501.
unified brain theory? Nat Rev Neurosci 11: Habre-Hallage, P., Hermoye, L., Gradkowski,
127–138. W. et al. (2010). A manually controlled new
device for punctuate mechanical stimulation of Husain, M. and Roiser, J.P. (2018). Neuroscience
teeth during functional magnetic resonance of apathy and anhedonia: a transdiagnostic
imaging studies. J Clin Periodontol 37: 863–872. approach. Nat Rev Neurosci 19: 470–484.
Habre-Hallage, P., Dricot, L., Jacobs, R. et al. Ikebe, K., Amemiya, M., Morii, K. et al. (2007).
(2012). Brain plasticity and cortical correlates Association between oral stereognostic ability
of osseoperception revealed by punctate and masticatory performance in aged
mechanical stimulation of osseointegrated complete denture wearers. Int J Prosthodont
oral implants during fMRI. Eur J Oral 20: 245–250.
Implantol 5: 175–190. Imamura, Y., Shinozaki, T., Okada-Ogawa,
Habre-Hallage, P., Dricot, L., Hermoye, L. et al. A. et al. (2019). An updated review on
(2014). Cortical activation resulting from the pathophysiology and management of burning
stimulation of periodontal mechanoreceptors mouth syndrome with endocrinological,
measured by functional magnetic resonance psychological and neuropathic perspectives.
imaging (fMRI). Clin Oral Investig 18: J Oral Rehabil 46: 574–587.
1949–1961. Imhoff, B., Ahlers, M.O., Hugger, A. et al. (2020).
Haggard, P. and De Boer, L. (2014). Oral Occlusal dysesthesia-A clinical guideline.
somatosensory awareness. Neurosci Biobehav J Oral Rehabil 47: 651–658.
Rev 47: 469–484. Jacobs, R., Bou Serhal, C., and Van Steenberghe,
Han, P., Mohebbi, M., Unrath, M. et al. (2018). D. (1998). Oral stereognosis: a review of the
Different neural processing of umami and literature. Clin Oral Investig 2: 3–10.
salty taste determined by umami James, W. (2007). The Principles of Psychology.
identification ability independent of repeated Cosimo, Inc.
umami exposure. Neuroscience 383: 74–83. Johansson, R.S., Trulsson, M., Olsson, K.A., and
Heft, M.W. and Robinson, M.E. (2010). Age Westberg, K.G. (1988). Mechanoreceptor
differences in orofacial sensory thresholds. activity from the human face and oral mucosa.
J Dent Res 89: 1102–1105. Exp Brain Res 72: 204–208.
Henderson, L.A., Akhter, R., Youssef, A.M. et al. Kagawa, T., Narita, N., Iwaki, S. et al. (2014).
(2016). The effects of catastrophizing on Does shape discrimination by the mouth
central motor activity. Eur J Pain 20: 639–651. activate the parietal and occipital lobes? –
Higaki, N., Goto, T., and Ichikawa, T. (2016). near-infrared spectroscopy study. PLoS One 9:
Periodontal tactile input activates the e108685.
prefrontal cortex. Sci Rep 6. Kawagishi, S., Kou, F., Yoshino, K. et al. (2009).
Hihara, H., Kanetaka, H., Kanno, A. et al. (2020). Decrease in stereognostic ability of the tongue
Somatosensory evoked magnetic fields of with age. J Oral Rehabil 36: 872–879.
periodontal mechanoreceptors. Heliyon 6: e03244. Klineberg, I., Calford, M.B., Dreher, B. et al.
Hilbert, K., Evens, R., Maslowski, N.I. et al. (2005). A consensus statement on
(2014). Fear processing in dental phobia osseoperception. Clin Exp Pharmacol Physiol
during crossmodal symptom provocation: an 32: 145–146.
fMRI study. Biomed. Res. Int.Biomed Res Int Kono, Y., Kubota, A., Taira, M. et al. (2018).
2014: 196353. Effects of oral stimulation with capsaicin on
Hirano, K., Hirano, S., and Hayakawa, I. (2004). salivary secretion and neural activities in the
The role of oral sensorimotor function in autonomic system and the brain. J Dent Sci 13:
masticatory ability. J Oral Rehabil 31: 199–205. 116–123.
Hort, J., Ford, R.A., Eldeghaidy, S., and Francis, Kothari, S.F., Dagsdóttir, L.K., Kothari, M. et al.
S.T. (2016). Thermal taster status: evidence of (2020). Effect of repetitive transcranial
cross-modal integration. Hum Brain Mapp 37: magnetic stimulation on altered perception of
2263–2275. One’s own face. Brain Stimul 13: 554–561.
Lara, A.H. and Wallis, J.D. (2015). The role of Miyamoto, J.J., Honda, M., Saito, D.N. et al.
prefrontal cortex in working memory: a mini (2006). The representation of the human oral
review. Front Syst Neurosci 9: 173. area in the somatosensory cortex: a functional
Legrain, V., Iannetti, G.D., Plaghki, L., and MRI study. Cereb Cortex 16: 669–675.
Mouraux, A. (2011). The pain matrix reloaded: Moana-Filho, E.J., Bereiter, D.A., and Nixdorf,
a salience detection system for the body. Prog D.R. (2015). Amplified Brain Processing of
Neurobiol 93: 111–124. Dentoalveolar Pressure stimulus in persistent
Liang, M., Mouraux, A., Hu, L., and Iannetti, dentoalveolar pain disorder patients. J Oral
G.D. (2013). Primary sensory cortices contain Facial Pain Headache 29: 349–362.
distinguishable spatial patterns of activity for Mobbs, D., Petrovic, P., Marchant, J.L. et al.
each sense. Nat Commun 4: 1979. (2007). When fear is near: threat imminence
Lin, C.S. (2018). Meta-analysis of brain elicits prefrontal-periaqueductal gray shifts in
mechanisms of chewing and clenching humans. Science 317: 1079–1083.
movements. J Oral Rehabil 45: 627–639. Mobbs, D., Yu, R., Rowe, J.B. et al. (2010). Neural
Lin, C.S., Ku, H.L., Chao, H.T. et al. (2014). activity associated with monitoring the
Neural network of body representation differs oscillating threat value of a tarantula. Proc
between transsexuals and cissexuals. PLoS Natl Acad Sci U S A 107: 20582–20586.
One 9: e85914. Morishima, Y., Akaishi, R., Yamada, Y. et al.
Longo, M.R., Azanon, E., and Haggard, P. (2010). (2009). Task-specific signal transmission from
More than skin deep: body representation prefrontal cortex in visual selective attention.
beyond primary somatosensory cortex. Nat Neurosci 12: 85–91.
Neuropsychologia 48: 655–668. Murray, G.M. and Peck, C.C. (2007). Orofacial
Maffei, A., Haley, M., and Fontanini, A. (2012). pain and jaw muscle activity: a new model.
Neural processing of gustatory information in J Orofac Pain 21: 263–278. discussion
insular circuits. Curr Opin Neurobiol 22: 279-88.
709–716. Narita, N., Kamiya, K., Makiyama, Y. et al.
Mansouri, F.A., Koechlin, E., Rosa, M.G.P., and (2019). Prefrontal modulation during chewing
Buckley, M.J. (2017). Managing competing performance in occlusal dysesthesia patients:
goals – a key role for the frontopolar cortex. a functional near-infrared spectroscopy study.
Nat Rev Neurosci 18: 645–657. Clin Oral Investig 23: 1181–1196.
Mascioli, G., Berlucchi, G., Pierpaoli, C. et al. Navratilova, E. and Porreca, F. (2014). Reward
(2015). Functional MRI cortical activations and motivation in pain and pain relief. Nat
from unilateral tactile-taste stimulations of Neurosci 17: 1304–1312.
the tongue. Physiol Behav 151: 221–229. Nishimori, H., Iida, T., Kamiyama, H. et al.
Mazzola, L., Royet, J.P., Catenoix, H. et al. (2019). Effect of sleep restriction on
(2017). Gustatory and olfactory responses to somatosensory sensitivity including occlusal
stimulation of the human insula. Ann Neurol sensation in the orofacial area. J Prosthodont
82: 360–370. Res 63: 193–198.
McNaughton, N. and Corr, P.J. (2004). A Nobre, A.C. and Van Ede, F. (2018). Anticipated
two-dimensional neuropsychology of defense: moments: temporal structure in attention. Nat
fear/anxiety and defensive distance. Neurosci Rev Neurosci 19: 34–48.
Biobehav Rev 28: 285–305. Oatley, K. and Johnson-Laird, P.N. (2014).
Meyer-Gerspach, A.C., Suenderhauf, C., Cognitive approaches to emotions. Trends
Bereiter, L. et al. (2016). Gut taste stimulants Cogn Sci 18: 134–140.
alter brain activity in areas related to Oda, M., Yoshino, K., Tanaka, T. et al. (2014).
working memory: a pilot study. Neurosignals Identification and adjustment of experimental
24: 59–70. occlusal interference using functional
magnetic resonance imaging. BMC Oral probabilistic learning via taste outcomes.
Health 14: 124. Physiol Behav 223: 112962.
Okamoto, M., Dan, H., Clowney, L. et al. (2009). Samuelsen, C.L. and Fontanini, A. (2017).
Activation in ventro-lateral prefrontal cortex Processing of intraoral olfactory and gustatory
during the act of tasting: an fNIRS study. signals in the gustatory cortex of awake rats.
Neurosci Lett 451: 129–133. J Neurosci 37: 244–257.
Okeson, J.P. (2019). Management of Schubotz, R.I. (2007). Prediction of external
Temporomandibular Disorders and Occlusion- events with our motor system: towards a new
E-book. Elsevier Health Sciences. framework. Trends Cogn Sci 11: 211–218.
Onuma, T., Maruyama, H., and Sakai, N. (2018). Schubotz, R.I. and Von Cramon, D.Y. (2003).
Enhancement of saltiness perception by Functional-anatomical concepts of human
monosodium glutamate taste and soy sauce premotor cortex: evidence from fMRI and PET
odor: a near-infrared spectroscopy study. studies. Neuroimage 20 (Suppl 1): S120–S131.
Chem Senses 43: 151–167. Schultz, W. (1998). Predictive reward signal of
Petersen, S.E. and Posner, M.I. (2012). The dopamine neurons. J Neurophysiol 80: 1–27.
attention system of the human brain: 20 years Scott, K. (2005). Taste recognition: food for
after. Annu Rev Neurosci 35: 73–89. thought. Neuron 48: 455–464.
Ploghaus, A., Narain, C., Beckmann, C.F. et al. Sekido, D., Otsuka, T., Shimazaki, T. et al. (2020).
(2001). Exacerbation of pain by anxiety is Comparison of cerebral cortex activation
associated with activity in a hippocampal induced by tactile stimulation between
network. J Neurosci 21: 9896–9903. natural teeth and implants. J Clin Exp Dent 12:
Pooresmaeili, A., Fitzgerald, T.H., Bach, e1021–e1026.
D.R. et al. (2014). Cross-modal effects of value Shearrer, G.E., Stice, E., and Burger, K.S. (2018).
on perceptual acuity and stimulus encoding. Adolescents at high risk of obesity show
Proc Natl Acad Sci U S A 111: 15244–15249. greater striatal response to increased sugar
Press, C., Kok, P., and Yon, D. (2020a). Learning content in milkshakes. Am J Clin Nutr 107:
to perceive and perceiving to learn. Trends 859–866.
Cogn Sci 24: 260–261. Shen, G., Meltzoff, A.N., and Marshall,
Press, C., Kok, P., and Yon, D. (2020b). The P.J. (2018). Touching lips and hearing fingers:
perceptual prediction paradox. Trends Cogn effector-specific congruency between tactile
Sci 24: 13–24. and auditory stimulation modulates
Ptito, M., Moesgaard, S.M., Gjedde, A., and Kupers, N1 amplitude and alpha desynchronization.
R. (2005). Cross-modal plasticity revealed by Exp Brain Res 236: 13–29.
electrotactile stimulation of the tongue in the Shimazaki, T., Otsuka, T., Akimoto, S. et al.
congenitally blind. Brain 128: 606–614. (2012). Comparison of brain activation via
Rauss, K. and Pourtois, G. (2013). What is tooth stimulation. J Dent Res 91: 759–763.
bottom-up and what is top-down in predictive Shinohara, Y., Umezaki, Y., Minami, I. et al.
coding? Front Psychol 4: 276. (2020). Comorbid depressive disorders and
Raz, A. and Buhle, J. (2006). Typologies of left-side dominant occlusal discomfort in
attentional networks. Nat Rev Neurosci 7: patients with phantom bite syndrome. J Oral
367–379. Rehabil 47: 36–41.
Roux, F.E., Djidjeli, I., and Durand, J.B. (2018). Shipp, S., Adams, R.A., and Friston, K.J. (2013).
Functional architecture of the somatosensory Reflections on agranular architecture:
homunculus detected by electrostimulation. predictive coding in the motor cortex. Trends
J Physiol 596: 941–956. Neurosci 36: 706–716.
Sadler, J.R., Shearrer, G.E., Acosta, N.T. et al. Simon, S.A., De Araujo, I.E., Gutierrez, R., and
(2020). Network organization during Nicolelis, M.A. (2006). The neural
Dental Neuroimaging: The Role of the Brain in Oral Functions, First Edition. Chia-Shu Lin.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
Companion website: www.wiley.com/go/lin/dental-neuroimaging
While the framework of pain neuromatrix experimental pain can be generalized to eluci-
highlights the multi-dimensionality of pain, it date the brain mechanisms of chronic pain. In
is not a ‘theory’ with enough details for the following sections, we respectively investi-
explaining all the mental processes about pain gate the brain mechanisms of acute and
(Derbyshire 2000). For example, the neuroma- chronic pain.
trix is conceived to function as a network, but
what are the components included in the net-
6.1.3 Brain Mechanisms Related
work? How does each component connect
to Acute Pain
with the other? These are all the questions that
remained unanswered. The framework was The recent advancement of neuroimaging has
later evolved into the concept of pain matrix, offered an excellent opportunity to directly
which can be loosely defined as ‘a group of observe the brain activation associated with
brain structures jointly activated by painful pain via a non-invasive approach in human
stimuli’ (Garcia-Larrea and Peyron 2013). The subjects. Following the framework of pain
definition provides an operational basis for matrix, researchers can map the brain network
identifying the pain matrix (i.e. by identifying by investigating the loci of brain activation
the loci of brain activation). Still, the associa- when subjects received painful stimuli. In an
tion between these loci and pain, especially its earlier review, Derbyshire reported that the
causal mechanisms, has remained unclear. anterior cingulate cortex (ACC), the primary
Furthermore, though many neuroimaging and secondary somatosensory cortices, the
studies adopted the concept of the pain matrix, insula and the thalamus showed consistent
there exists a great variation of its definition activation across a variety of pain research of
across studies. For example, different neuroim- tonic or phasic pain (Derbyshire 2000).
aging studies may identify the ‘pain matrix’ by Subsequently, several meta-analyses have been
finding brain activation at different regions published by pooling the patterns of brain acti-
(Davis et al. 2017; Derbyshire 2000; Mouraux vation derived from the research on experi-
and Iannetti 2018). In addition to the research mental pain (Burns et al. 2016; Duerden and
purpose, there is also a practical need for eluci- Albanese 2013; Farrell et al. 2005; Jensen
dating the association between pain and the et al. 2016; Peyron et al. 2000; Tanasescu
brain. Clinically, the symptoms and signs of et al. 2016) (Figure 6.1a). In general, the stud-
chronic pain differ from those of acute pain, ies identified that during the stimulation by
and the pain derived from neuropathic mecha- experimental pain, the primary somatosensory
nisms (e.g. trigeminal neuralgia [TN]) would cortex (S1) and secondary somatosensory cor-
differ from that derived from nociceptive tex (S2), the insula, the cingulate cortex and
mechanisms (e.g. a toothache from pulpitis). the thalamus have been consistently identified
The patients with chronic pain show an altered for their activation (Figure 6.1a). The findings
affective–motivational experience compared to are consistent with the results of animal
the patients with acute pain (Malfliet research, which has identified the role of the
et al. 2017). Therefore, it is not surprising that thalamocortical circuitry in pain processing.
the brain mechanisms of acute pain would be Specifically, the lateral pathway of the circuitry
different from those of chronic pain, which participates in the processing of sensory–dis-
may be more associated with affective–cogni- criminative information of pain, including the
tive processing of pain; and the brain mecha- S1 and the S2. The medial pathway, in contrast,
nisms associated with neuropathic pain may participates in the processing of cognitive–
be associated with the mechanisms of neural affective information of pain, including the
injury and repair. Therefore, it would be ques- insula and the cingulate cortex pain (Groh
tionable if the ‘pain matrix’ identified by et al. 2018). The results reveal that the
Figure 6.1 Brain activation associated with pain processing. (a) The brain regions commonly reported in
functional magnetic resonance imaging (fMRI) studies of noxious stimuli. The figure only displays the
relative position and size of the brain regions, not depicting the anatomical details. Note that the insular
cortex (bounded by a dashed line) is covered by the frontal, parietal and temporal operculum. (b) Meta-
analytical findings of the brain activation of experimental orofacial pain in healthy subjects. Increased
activation is consistently found in the posterior mid-cingulate cortex (pMCC), the PPC, the insula, the
thalamus, the S1 and the S2. Decreased activation is consistently found in the primary motor cortex (M1)
and the S1. Source: Ayoub et al. (2018). Reproduced with permission of Elsevier.
processing of acute pain is more likely to are associated with multiple brain functions,
engage with a ‘network’ of multiple brain from attentional control to processing of inter-
regions. However, it does not mean that these oception (Simmons et al. 2013). Notably, there
brain regions are only for pain processing. For is a great variation in the pattern of regional
example, the thalamus relays sensory informa- activation across the studies. For example, acti-
tion from both nociceptive and non-nociceptive vation in the limbic system, such as the
stimuli, and the anterior insula and the ACC hippocampus and the amygdala, was not
consistently identified across all the studies. and attention of pain, may modulate our pain
The inconsistent pattern of activation of these (Wiech 2016). For example, Ploghaus et al.
regions may reflect the fact that people are (2001) found that when subjects received nox-
much different in their cognitive–affective ious stimuli of the same intensity, they per-
experience of pain (e.g. different feelings of ceived greater pain when the stimuli were
threat towards pain). associated with higher anxiety compared to
The same findings have been replicated in the stimuli associated with lower anxiety.
recent neuroimaging meta-analyses about Furthermore, the anxiety-potentiating effect
experimental orofacial (Ayoub et al. 2018; Lin on pain is associated with brain activation at
et al. 2014b). For example, Lin et al. (2014b) the hippocampus (Ploghaus et al. 2001).
have analyzed the pattern of the brain activa- Likewise, Wiech et al. (2010) found that when
tion of six studies of electrical stimulation on subjects received noxious stimuli calibrated at
the teeth of healthy subjects. They found that the detection threshold of pain, they were
activation at the thalamus, the somatosensory more frequently to rate the stimuli as ‘painful’
cortex, the insula, the cingulate cortex and the in the conditions when they felt that the stim-
dorsolateral prefrontal cortex (PFC) were con- uli were threatening compared to the condi-
sistent in the studies. The pattern of activation tions when they felt the stimuli were safe.
generally corresponds to the medial and lateral Increased threat value and increased pain are
nociceptive systems (Groh et al. 2018). When both associated with brain activation at the
healthy adults received acute toothache elicited anterior insula (Wiech et al. 2010). Tracey et al.
by electrical stimuli, the activation of the ante- (2002) reported that when subjects received
rior insula and the ACC reflected the intensity painful thermal stimuli vs. non-painful warm
of pain (Brügger et al. 2012). Furthermore, the stimuli, they perceived the stimuli as less pain-
brain activation of the posterior insula reduced ful when they were not attending to pain com-
when dental pain was relieved by a mental pared to the condition when they were
nerve block (Meier et al. 2015). Notably, the attending to pain. Notably, the periaqueductal
presence of the dorsolateral PFC may signify grey (PAG) showed increased activation in the
the importance of attentional and cognitive non-attentive conditions compared to the
processing during painful stimulation (Lin attentive conditions (Tracey et al. 2002). In the
et al. 2014b). Another meta-analysis by Ayoub functional magnetic resonance imaging (fMRI
et al. (2018) also confirmed the role of similar study by Ploner et al. (2011)), subjects received
brain regions when they investigated orofacial painful stimuli in accompany with the follow-
pain induced by electrical, mechanical, ther- ing conditions: (i) an ‘attention’ condition that
mal or chemical stimuli. In the study, the PFC they counted the number of stimuli they
was not reported as a consistently activated received and (ii) an ‘emotional’ condition that
region, perhaps due to the variations in experi- they watched pictures of different emotional
mental conditions. In general, the findings valence. Subjects perceived greater pain when
revealed that brain regions for sensory, cogni- they attended to pain and watched pictures
tive and emotional processing together take with negative emotional context. Both the
part in the processing of pain, highlighting the attention and emotional effects were associ-
multi-dimensionality of pain (Figure 6.1b). ated with an increased brain activation at the
anterior insula, and the anterior insula
showed a distinct connection to the brain net-
6.1.4 Cognitive–Affective Factors Related
works of attention and emotional processing
to Acute Pain
(Ploner et al. 2011). Together, the findings sug-
It is widely known that cognitive–affective fac- gest that by manipulating the cognitive–
tors, such as the expectation related to pain affective factors, such as the threat value
related to pain and attention of pain, experi- example, the pain derived from empathizing
menters can modulate the pain intensity per- with other’s pain (e.g. watching another person
ceived by subjects. being cut in fingers) is associated with activa-
The brain mechanisms of cognitive and emo- tion at the anterior insula and the ACC (Lamm
tional processing of pain have been gradually et al. 2011). The frontoparietal regions are asso-
clarified by recent neuroimaging meta-analyses ciated with pain empathy of different body parts
(Table 6.1). Some experience of ‘pain’ may arise and the visuospatial perspective (e.g. first or
even without direct noxious stimuli. For third-person) (Jauniaux et al. 2019). Research of
Jauniaux et al. (2019) 86 fMRI studies of pain empathy of different visual cues, ALE
visuospatial and cognitive perspectives
King and Carnahan (2019) 28 fMRI/PET studies of innocuous and noxious cold ALE
exposure
Roberts et al. (2019) 11 fMRI experiments of experimentally induced itch ALE
Ayoub et al. (2019) 21 fMRI/PET/ASL-MRI studies of experimental pain ALE
and 28 studies of chronic pain
Coll (2018) 40 EEG studies of empathy of pain in others Random effects model
Zunhammer et al. (2018) 20 fMRI studies of placebo effects Random effects model
Ayoub et al. (2018) 30 MRI/PET functional and structural studies of ALE
experimental orofacial pain and chronic orofacial pain
Tatu et al. (2018) 55 VBM studies of chronic pain ALE
Chang et al. (2018) 67 studies (TMS/MRI/EEG/MEG/MRS/ PET) of patients Random effects model
with chronic pain
Yuan et al. (2017) 10 VBM studies of chronic back pain Seed-based d mapping
Jia and Yu (2017) 8 sMRI and 5 fMRI studies of migraine ALE
Tanasescu et al. (2016) 266 fMRI studies of experimental cutaneous pain Modified ALE-based
algorithm
Burns et al. (2016) 25 fMRI/PET/TMS/EEG studies of acute muscle pain Random effects model
Jensen et al. (2016) 138 fMRI/PET studies of pain in healthy individuals and ALE
32 studies of pain in patients
Dehghan et al. (2016) 37 MRI/PET/SPECT/DTI/MRS/EEG/MEG studies of ALE
fibromyalgia syndrome
Del Casale et al. (2015) 8 functional neuroimaging studies of pain perception ALE
under hypnosis
Palermo et al. (2015) 19 fMRI studies of pain anticipation ALE
Dai et al. (2015) 9 VBM studies of migraine ES-SDM
Pan et al. (2015) 10 VBM studies of neuropathic pain ES-SDM
Notes: ALE: activation likelihood estimation; ASL: arterial spin labelling; DTI: diffusion tensor imaging; EEG:
electroencephalography; ES-SDM: effect size-signed differential mapping; fMRI: functional magnetic resonance
imaging; MEG: magnetoencephalography; MRI: magnetic resonance imaging; MRS: magnetic resonance
spectroscopy; PET: positron emission tomography; sMRI: structural magnetic resonance imaging; SPECT: single
photon emission computed tomography; TMS: transcranial magnetic stimulation; VBM: voxel-based morphometry.
event-related potential (ERP) also revealed the with pain, we would focus on the link between
association between the frontoparietal regions structural features and patients’ self-reported
and pain empathy (Coll 2018). Notably, a recent ratings. Because pain is a subjective experi-
meta-analysis revealed a weaker association ence, self-report is still the gold standard for
between placebo analgesia and the neurological assessing individual pain (Davis et al. 2017).
signature of nociceptive pain, a neuroimaging Previous studies have revealed that the pat-
marker for nociceptive pain processing (Wager terns of brain activation vary between different
et al. 2013). The findings suggest that the brain types of clinical pain. For example, patients
mechanisms of pain modulation by placebo with chronic back pain (CBP), post-herpetic
may be largely independent of bottom-up noci- neuralgia (PHN), complex regional pain syn-
ceptive processing (Zunhammer et al. 2018). drome (CRPS), osteoarthritis of the knee (OA)
Neuroimaging meta-analysis also identified the and pelvic pain (PP) showed different patterns
brain mechanisms of pain modulation by hyp- of brain activation during imaging (Apkarian
notic suggestions, including hypnotic analgesic, et al. 2009; Mansour et al. 2016). These chronic
pleasant or depersonalization suggestions pain-related patterns consistently showed a
(HASs). When subjects received experimental distinct activation at the PFC (Apkarian
pain, HASs were associated with increased acti- et al. 2009). Baliki et al. (2006) reported that in
vation at the ACC, the insula and PFC, as well the patients with chronic low-back pain, sus-
as decreased activation in the thalamus com- tained high pain was associated with increased
pared to the control condition (Del Casale activation in the medial prefrontal cortex
et al. 2015). Activation of the posterior insula (mPFC), which are widely considered as the
and the parietal lobe can be identified when brain regions associated with affective–
individuals perceived painful vs. non-painful motivational processing. In general, patients
cold stimuli (King and Carnahan 2019). The with chronic pain showed altered brain fea-
insular activation is also found when subjects tures not just in the sensorimotor network but
perceived itch (Roberts et al. 2019). Furthermore, also in the brain network associated with affec-
the activation at the insula, the cingulate cortex tive and motivation processing (Baliki and
and the frontoparietal regions is also identified Apkarian 2015; Mitsi and Zachariou 2016).
when subjects anticipated (rather than received) Chronic pain is also associated with alterations
pain (Palermo et al. 2015). These recent find- in the functional connectivity of brain net-
ings, again, highlight the role of cognitive–affec- works (Figure 6.2). Among these networks, the
tive processing in the brain mechanisms of default mode network (DMN) has been
acute pain. widely investigated for chronic pain. The DMN
mainly includes frontal and posterior midline
(e.g. the ventral mPFC and the posterior
6.1.5 Brain Mechanisms Related
cingulate cortex (PCC)) and the inferior pari-
to Chronic Pain
etal lobule (Buckner et al. 2008). During a
Unlike acute pain, chronic pain is usually resting state, higher functional connectivity
stimulus-independent and occurs spontane- was found between the brain regions of the
ously. Therefore, different approaches have DMN. Interestingly, the DMN connectivity
been adopted to study the brain mechanisms reduces when subjects are actively engaged in
of chronic pain. Firstly, if we want to know the an experimental task. In other words, the DMN
brain activation associated with ongoing pain, shows stronger connectivity when individuals
we may ask patients to continuously rate how are not disturbed by external stimuli (i.e. dur-
pain fluctuates during the imaging session ing a resting state) compared to the condition
(Baliki et al. 2006, 2010). Secondly, if we are when they are interacting with the environ-
interested in the structural features associated ment (e.g. during a task-based fMRI study).
Figure 6.2 Functional networks associated with chronic pain. Source: Davis et al. (2017) with permission of
Springer Nature under the terms of the Creative Commons CC BY 4.0 License.
Baliki et al. (2008a) investigated the DMN in dorsolateral PFC compared to healthy con-
patients with CBP and found that when the trols (Tatu et al., 2018). Patients with
patients were performing an attention task, migraines showed a decreased GMV in the
they showed a reduced deactivation in the posterior insula, the PFC and the ACC (Dai
brain regions of the DMN. Furthermore, a et al. 2015; Jia and Yu 2017). Likewise, in
recent neuroimaging meta-analysis revealed patients with chronic low back pain, there is a
that patients with chronic pain showed con- decreased GMV in the medial frontal cortex
sistently less activation in the right anterior and the ACC (Yuan et al. 2017). In neuro-
hippocampus, and in patients with CBP, there pathic pain, patients showed decreased GMV
was a weaker resting-state functional connec- in multiple regions, including the anterior
tivity between the anterior hippocampus and insula and lateral frontal gyrus. There is also
the mPFC compared to healthy controls an increased GMV in the posterior insula and
(Ayoub et al. 2019). The findings suggest that medial frontal gyrus (Pan et al. 2015). The
changes of the default (i.e. task-free or participation of the primary motor cortex
stimulus-independent) network of the brain (M1) is less clear-cut in patients with chronic
may be associated with persistent and sponta- pain (Chang et al. 2018). In patients with
neous pain in patients with chronic pain. fibromyalgia syndrome, consistent changes
In addition to the fMRI studies, structural are found not only in the somatosensory area
magnetic resonance imaging (sMRI) studies but also in the insula and the amygdala
focus on identifying the association between (Dehghan et al. 2016). Moreover, the experi-
brain features and self-report findings or clin- ence of chronic pain may be associated with
ical symptoms. Meta-analytic findings the plastic effect on grey matter morphology.
revealed that patients with chronic pain In a longitudinal MRI study, Seminowicz
showed an increased grey matter volume et al. (2011) investigated patients with chronic
(GMV) mainly at the bilateral thalamus and low back pain before and six months after
the bilateral basal ganglia, and a decreased receiving treatment. They found that before
GMV mainly at the bilateral ACC and treatment, the patients showed a thinner
cortical thickness at the dorsolateral PFC, the have seen in the preceding sections, nocicep-
insula and the ACC, compared to the healthy tive stimuli can be perceived as stronger or
controls. After treatment, the patients showed weaker pain in different experimental condi-
a regain in the thickness in the dorsolateral tions, which vary between the threat value
PFC, which was associated with a reduction related to pain or attention of pain (Ploner
of pain (Seminowicz et al. 2011). Furthermore, et al. 2011; Tracey et al. 2002; Wiech et al. 2010).
a recent meta-analysis on sMRI studies of Neuroimaging findings have identified activa-
chronic pain revealed that among the brain tion of the brainstem and the medulla, includ-
regions showing alterations between patients ing the PAG, during pain modulation by
and healthy controls, there exist co-alterations cognitive–affective processing (Lin et al. 2014a;
between some pair of regions. Brain networks Tracey et al. 2002). The findings correspond to
were established according to the spatial dis- the descending pathway of pain modulation
tribution of co-altered regions, respectively, identified from animal research (Ossipov
for the regions with increased and decreased et al. 2010). Furthermore, the neuroimaging
GMV. Notably, the researchers also identified studies gradually revealed how the higher cor-
more regions with decreased GMV (in patients tex, including the PFC, participates in pain
vs. healthy controls) than regions with modulation. Neuroimaging research helps to
increased GMV. The anterior insula, the ACC clarify the role of the PFC in pain modulation
and the posterior insula play a key role in the and its contribution to the chronification of
co-altered network of decreased GMV (Tatu pain. For example, an earlier longitudinal study
et al. 2018). The findings suggest that struc- of patients with sub-acute back pain revealed
tural alterations of the salience network may that the white matter fractional anisotropy (FA)
play a key role in chronic pain. assessed at the baseline predicted the persis-
tence of pain over the next year (Mansour
et al. 2013). Moreover, in the subjects whose
6.1.6 Cognitive–Affective Factors Related
pain recovered after one year, the regional FA
to Chronic Pain
of the mPFC was positively correlated with the
Since chronic pain is associated with various FA of the nucleus accumbens (NAc) as well as
cognitive and emotional experiences, it is not the functional connectivity between NAc and
surprising that the pattern of brain activation mPFC. Together, the findings highlight that the
associated with chronic pain will show a mPFC may play a key role in mediating antino-
substantial overlap with the brain activation ciceptive effects, partly via the PAG (Ong
associated with emotional and motivation et al. 2019), and contribute to the chronifica-
processing, which is not necessarily pain- tion of pain.
related (Davis et al. 2017). In the following sec-
tions, we briefly discuss two topics of 6.1.6.2 Chronic Pain and Reward Processing
cognitive–affective processing of pain, i.e. the The mesolimbic pathway, which consists of the
brain mechanisms of pain modulation and NAc and the ventral tegmental area (VTA), is
reward processing of pain. one of the major dopaminergic pathways that
play a key role in reward and motivation pro-
6.1.6.1 Modulation and Chronification of Pain cessing (Serafini et al. 2020). Alterations in the
The pain we perceive does not fully reflect the mesolimbic pathway may contribute to reor-
intensity of nociceptive stimuli. In other words, ganizing the neocortex into a chronic pain state
the stimuli of the same nociceptive intensity (Baliki and Apkarian 2015). The activation of
may be perceived differently due to the modu- reward circuitry may play a key role in the
lation from the central nervous system (CNS), experience of chronic pain (Mitsi and
in which the brain plays a critical role. As we Zachariou 2016). The PFC may relate to the
mesolimbic pathway and plays a key role in during experiments, e.g. perceiving a salient
motivation and reward processing in chronic stimulus of a threat (Mouraux and
pain patients. For example, in patients with Iannetti 2018). More recently, using the mul-
sub-acute pain lasting for one year, decreased tivariate pattern analysis (MVPA), research-
GMV at the NAc and increased functional and ers were able to identify the assembly of
structural connectivity with the PFC have been regional activation that discriminated the
noted (Baliki et al. 2012; Serafini et al. 2020). presence of pain (Brodersen et al. 2012;
When receiving acute painful thermal stimuli, Brown et al. 2011; Wager et al. 2013). The
healthy subjects showed higher activation in MVPA approach has been used to identify the
the NAc compared to patients with CBP (Baliki neurologic pain signatures (NPS), i.e. ‘a mul-
et al. 2010). Notably, in the patients, the func- tivariate brain pattern tracking nociceptive
tional connectivity between the NAc and the pain’ (Wager et al. 2013; Zunhammer
mPFC was positively correlated with the inten- et al. 2018). The methodological advance-
sity of their chronic pain (Baliki et al. 2010). ment greatly improved the low sensitivity/
Together, the findings highlight the key role of specificity of pain prediction and offered an
the interaction between pain modulation and unbiased estimation between the association
reward processing in chronic pain. Patients of pain and brain activation because it did not
with rheumatoid arthritis showed increased pre-select the brain regions for testing
functional connectivity between the SMA, the (Reddan and Wager 2018). Another recent
mid-cingulate cortex, the primary somatosen- study of MVPA reported a new signature, the
sory area, the PFC and the insula (Flodin stimulus intensity-independent pain signa-
et al. 2016). Notably, in the patients with ture-1 (SIIPS1), to predict the trial-by-trial
chronic pain, thinning of cortical thickness at pain ratings. In addition to predicting the
the PFC was noted (Seminowicz et al. 2011). In intensity of the noxious stimulus, its response
general, the findings reveal that the brain can be influenced by affective–cognitive fac-
mechanisms of pain modulation and reward tors of pain, which contributed to pain modu-
processing, which would be associated with the lation (Woo et al. 2017).
PFC and the mesolimbic pathway, play a key Another issue of interest is to focus on the
role in chronic pain (Mitsi and Zachariou 2016). pattern of intrinsic functional connectivity
during the resting state instead of the brain
activation when patients perceive pain. The
6.1.7 Imaging/Neurological
approach helps to elucidate how the ongoing
Pain Signatures
and spontaneous pain disrupts one’s intrin-
Since the earlier days of neuroimaging, sic functional connectivity (Kucyi and
researchers have kept a strong enthusiasm in Davis 2015). The findings broaden the view
finding the imaging makers for chronic pain of ‘neuromatrix’ from a stimuli-based net-
(Box 6.1). The imaging marker, by definition, work to a resting-state network, highlighting
should be highly predictive of the occurrence the dynamic fluctuation between different
of pain, and the absence of the marker will be states in chronic pain patients (Baliki
predictive of the non-occurrence of pain et al. 2008a; Davis et al. 2017). However, it
(Mouraux and Iannetti 2018). As shown in should be noted that these ‘predictors’ –
earlier reviews, while some brain regions either derived from the MVPA approach or a
showed consistent activation across the stud- combination of stimuli-based and resting-
ies, many of these brain regions did not show state research – did not delineate the causal
an activation to pain selectively. Many brain relationship between pain and the neuroim-
regions may show an activation correspond- aging signatures (Davis et al. 2017; Mouraux
ing to the affective–motivational processing and Iannetti 2018).
Box 6.1 From Research to Practice – Revision of the Definition of Chronic Pain by
Neuroimaging Evidence
The recent findings of central sensitization forms of chronic pain, such as fibromyalgia
and neuroimaging of brain plasticity have or nonspecific low-back pain, ‘as a health
raised a critical question in the diagnosis condition in its own right,’ as categorized in
and treatment of chronic pain: if the persis- the subgroup ‘chronic primary pain’. In con-
tent pain leads to changes in the CNS and trast, the other sub-classes are secondary
related behaviour, should it be a disease on to an underlying disease (e.g. chronic
its own right or a symptom secondary to an musculoskeletal pain) (Treede et al. 2019).
underlying aetiology? Cumulating evidence The clarification, on the one hand, keeps
has suggested that chronic pain is associ- consistent with the view that pain is a
ated with changes in the brain in multiple symptom secondary to disease. On the
aspects, including structural, functional and other hand, it highlights the possibility that
neurochemical changes in the brain (Tracey the persistence of pain per se should be
and Bushnell 2009). The IASP Task Force focused as the target of treatment, even
has refined the definition for each sub- when no underlying disease is identified, as
class of chronic pain and pointed that some in the case of chronic primary pain.
a strong negative effect on patients’ feeding pains that occur on at least 50% of the days dur-
behaviour and psychosocial well-being (Nasri- ing at least 3 months’. It is noteworthy that
Heir et al. 2016). While acute pain is sometimes inconsistency between the site of pain and the
considered as ‘good pain’ because it signifies source of pain is common for chronic orofacial
ongoing tissue damage and motivates patients pain. For example, patients with neurovascular
to seek medical care, chronic pain exacerbates orofacial pain may perceive toothache (i.e. the
patients’ quality of life and disrupts their daily site of pain) without any relevant dental lesion.
functions (Sharav and Benoliel 2008). This sec- The toothache derives from migraine (i.e. the
tion outlines recent neuroimaging findings on source of pain) (Benoliel et al. 2010). In gen-
the brain mechanisms of chronic pain. eral, in chronic orofacial pain, the association
Specifically, we focus on the mechanisms of between clinical symptoms and signs and the
central sensitization, which have been widely underlying mechanisms is not clear-cut.
studied in recent years, as a potential mecha- Therefore, the diagnosis of chronic orofacial
nism for persistent pain. At the neurophysiolog- has long been a major challenge in dentistry.
ical level, central sensitization is a manifestation
of the plasticity of the somatosensory nervous
6.2.3 Chronic Pain and Neural Plasticity
system (Latremoliere and Woolf 2009). We
briefly discuss the role of neural plasticity in The reason why the classification and diagno-
pain and behaviour in this section. sis of chronic pain are so challenging can be
partly explained by the definition of pain.
According to the definition by IASP, pain is
6.2.2 Definition and Classification
associated with actual or potential tissue dam-
of Chronic Pain
age (IASP 2020). For acute pain, this associa-
According to the International Classification of tion between pain and tissue damage cannot
Diseases (ICD) version 11, chronic pain is be neglected. For example, to relieve tooth-
defined as ‘persistent or recurrent pain lasting ache, the lesions from dental pulp or periodon-
longer than 3 months’ (Treede et al. 2015; tium need to be managed first. However,
Treede et al., 2019). This general definition of chronic pain is a long-term experience. In
chronic pain does not specify the pathological addition to the identification of the tissue dam-
mechanisms of chronic pain, partly due to the age or peripheral lesions related to pain, we
complexity of the underlying mechanisms. The also need to consider how the (painful) experi-
Task Force for the Classification of Chronic ence is shaped and what factors contribute to
Pain, organized by the IASP, established the the chronification of pain.
classification of chronic pain according to mul-
tiple criteria, including the aetiology of pain, 6.2.3.1 Neural Plasticity
the underlying pathophysiological mecha- The shaping of individual experience may be
nisms and the body site of pain. In the latest associated with the brain mechanisms of
version of the ICD (ICD-11), chronic pain is plasticity, which is generally defined as ‘the
classified into seven major categories: (i) susceptibility of human behavior to modifica-
chronic primary pain, (ii) chronic cancer- tion’ at the behavioural level (Pascual-Leone
related pain, (iii) chronic postsurgical and post- et al. 2005). In terms of neuroscience, the
traumatic pain, (iv) chronic neuropathic pain, concept of neural plasticity specifically refers
(v) chronic secondary headache and orofacial to the phenomenon that ‘neuronal and synap-
pain, (vi) chronic secondary visceral pain and tic functions are capable of being moulded or
(vii) chronic secondary musculoskeletal pain shaped so that they influence subsequent
(Treede et al. 2015). Chronic headache and oro- perceptual experiences’ (Melzack et al. 2001).
facial pain is defined as ‘headaches or orofacial The term ‘neural plasticity’ needs further
Figure 6.3 Mechanisms of peripheral and central sensitization. (a) In the normal status, signals induced by
noxious stimuli and non-noxious (e.g. tactile) stimuli are transduced via the pathway of nociceptive and
tactile processing, respectively. (b) In peripheral sensitization, neurons would show increased
responsiveness to noxious stimuli. For example, the inflammation at the peripheral site (i.e. the light red
area) may reduce the threshold to evoke a response. The signals for subsequent nociceptive processing are
amplified. Therefore, peripheral sensitization may be associated with hyperalgesia. (c) In secondary
hyperalgesia (in contrast to primary hyperalgesia (b)), noxious stimulation to the area surrounding the site
of injury or inflammation (i.e. the black arrow) elicits an amplified pain. The amplification is mediated by
the central neurons (i.e. the red circle), which have been sensitized by constantly receiving nociceptive
inputs from the primary lesion (i.e. the circled light grey area). (d) In allodynia, non-noxious tactile stimuli
are conveyed by the Aβ fibre elicit pain. Note that at the central level, both the nociceptive and tactile
pathways converge on central nociceptive neurons. The central neurons (the red square) have been
sensitized by constantly receiving nociceptive inputs from the primary lesion.
Central sensitization is defined by the IASP higher level in the CNS. For example, in
as ‘increased responsiveness of nociceptive patients with neuropathic pain, the thalamus
neurons in the central nervous system to showed a high rate of spontaneous firing (Lenz
their normal or subthreshold afferent input’ et al. 1989). As discussed in Section 6.3, the
(IASP 2020). In the case of central sensitiza- neuropathic pain of orofacial regions may be
tion, persistent stimulation to the central associated with central sensitization of the
nervous system (CNS) neurons would trigeminal pathway from the brainstem to the
amplify their signals (Figure 6.3c). Such an thalamus (Park et al. 2006).
amplification occurs when an excitatory syn-
apse is facilitated or an inhibitory synapse is 6.2.4.3 Hyperalgesia
depressed (Woolf and Salter 2000). Therefore, Hyperalgesia can be differentiated into primary
the sensitization of the nervous system is and secondary hyperalgesia according to the
associated with alterations in the experience relationship between the site of tissue damage
of pain. and pain (Johanek et al. 2006). Primary hyper-
algesia occurs at the same site where damage
6.2.4.2 Allodynia occurs (Schmidt and Willis 2007). In contrast,
Until now, most of our understanding of cen- secondary hyperalgesia develops when noxious
tral sensitization has been based on animal stimuli are applied in the area surrounding the
research. A key feature of central sensitization site of tissue damage (Figure 6.3c). Secondary
is the amplification of the neuronal responses hyperalgesia is associated with central sensiti-
to stimuli, which can be noxious or non- zation and is commonly reported by patients
noxious. The condition that pain is perceived with chronic pain. Increased pain is mediated
‘due to a stimulus that does not normally pro- by the CNS neurons that are sensitized due
voke pain’ is termed allodynia (IASP 2020). to continuous nociceptive inputs from the
Some forms of allodynia, such as ‘thermal allo- site of primary hyperalgesia (Schmidt and
dynia’, derive from changes at the peripheral Willis 2007).
level (Sharav and Benoliel 2008). However, the
symptom of mechanical allodynia should be
6.2.5 Neuroimaging Research
carefully examined because it may indicate
on Central Sensitization
central sensitization. Mechanical allodynia is
observed when a non-noxious stimulus (e.g. As a critical mechanism of chronic pain, cen-
cotton-swapping), which normally induces Aβ tral sensitization has been a widely investi-
fibre (for tactile sensation), evokes pain. In gated topic of neuroimaging research in recent
normal status, signals from Aβ fibre do not years. The recent findings of pain sensitization
excite the CNS neuron of nociceptive process- from experimental research (i.e. pain sensitiza-
ing, which has a higher threshold of response tion induced by experimental methods) or
(Figure 6.3a). However, in central sensitiza- clinical research (i.e. chronic pain associated
tion, the CNS neuron of nociceptive processing with sensitization) are summarized in
may respond to subthreshold afferent input. Table 6.2.
As shown in Figure 6.3d, in the condition of
central sensitization, the signals conveyed by 6.2.5.1 Research on Sensitization Induced by
the Aβ fibre may excite the nociceptive neu- Experimental Methods
rons in the CNS lead to pain. Therefore, the For healthy subjects, central sensitization can
experience of mechanical allodynia is also be studied using the model of secondary
termed as ‘Aβ pain’ (Sharav and Benoliel 2008). hyperalgesia, as noted above, which refers to
It is noteworthy that beyond the spinal cord the pain derived from stimuli in the area
and the brainstem, sensitization may occur at a surrounding the site of an injury or
Table 6.2 Neuroimaging research on brain mechanisms of central sensitization and pain (since 2015).
Experimental pain
Torta et al. (2020) 60 healthy adults, secondary hyperalgesia induced EEG/mechanical pinprick
by low-frequency electrical stimulation of the skin stimuli and cognitive tasks of
increasing difficulty
Hansen et al. (2019) 115 healthy adults, secondary hyperalgesia rs-fMRI/functional connectivity
induced by thermal sensitization
Hansen et al. (2018) 121 healthy adults, secondary hyperalgesia sMRI/FreeSurfer
induced by thermal sensitization
Furman et al. (2018) 44 healthy adults, prolonged pain induced by EEG before and after capsaicin
capsaicin-heat pain incubation/heat stimuli
Van Den Broeke et al. 16 healthy adults, secondary hyperalgesia EEG before and 20 min after
(2017) induced by HFS HFS/pinprick stimuli
Liang et al. (2016) 12 healthy adults, hyperalgesia induced by EEG/intraepidermal electrical
intraepidermal injection of capsaicin stimulation
Van Den Broeke et al. 19 healthy adults, secondary hyperalgesia EEG before and 30 min after
(2015) induced by intradermal capsaicin capsaicin injection/pinprick
stimuli
Asghar et al. (2015) 40 healthy adults, secondary hyperalgesia sMRI, fMRI before and after
induced by a first-degree burn injury induction/mechanical noxious
stimuli
Disease-related pain: headache
Chong et al. (2020) 18 patients with cluster headache and 19 sMRI/FreeSurfer
patients with migraine (22 healthy controls)
Aguila et al. (2016) 20 patients with migraine (20 healthy controls) MRS
1
Becerra et al. (2016) 32 patients with migraine (33 healthy controls) H-MRS
Chen et al. (2015) 15/19 patients of migraine w/wo cutaneous fMRI/transcutaneous electrical
allodynia (20 healthy controls) nerve stimulation at the forearm
Chen et al. (2017) 27 patients with medication-overuse headache sMRI/volumetric analysis
(27 normal controls)
Disease-related pain: fibromyalgia
Hazra et al. (2020) 50 patients of fibromyalgia (50 healthy controls) fNIRS/cold pressure test
Kaplan et al. (2020) 27/27 patients with rheumatoid arthritis w/wo fMRI
concomitant fibromyalgia
Hadanny et al. (2018) 30 patients with fibromyalgia syndrome and a dMRI, SPECT
history of childhood sexual abuse receiving
hyperbaric oxygen therapy and psychotherapy
Bosma et al. (2016) 14 patients with fibromyalgia syndrome (15 fMRI/heat pain stimuli
pain-free controls)
Lim et al. (2015) 17 patients with fibromyalgia (21 healthy MEG/median nerve
controls) stimulation at the wrist
(Continued)
Table 6.2 (Continued)
Notes: dMRI: diffusion magnetic resonance imaging; EEG: electroencephalography; fMRI: functional magnetic
resonance imaging; fNIRS: functional near-infrared spectroscopy; HFS: high-frequency electrical stimulation;
HF-SCS: high-frequency spinal cord stimulation; MEG: magnetoencephalography; MRI: magnetic resonance
imaging; MRS: magnetic resonance spectroscopy; rs-fMRI: resting-state functional magnetic resonance imaging;
SPECT: single photon emission computed tomography; sMRI: structural magnetic resonance imaging.
not associated with the volume of brain pain with motor syndrome (Nishioka
regions of pain processing. Instead, the et al. 2019). Most of the studies made a com-
increased size of secondary hyperalgesia was parison between a clinical group and a group of
associated with changes in functional connec- non-painful healthy controls (Table 6.2). A
tivity of the sensorimotor network and the variety of brain features were investigated,
DMN (Hansen et al. 2019). Using electroen- including the morphology of grey matter, the
cephalography (EEG), researchers also identi- morphology of white matter, functional con-
fied the neural activity associated with nectivity and structural connectivity. However,
secondary hyperalgesia when subjects were the neuroimaging studies do not show very
receiving mechanical or electrical stimuli consistent findings. For example, Chen et al.
(Liang et al. 2016; Van Den Broeke et al. 2017; (2017) investigated the patients with migraine
Van Den Broeke et al. 2015). For example, and medical-overuse headache and found an
healthy subjects received intradermal capsai- increased thalamic volume in patients com-
cin to induce an area of secondary hyperalge- pared to controls. De Kruijf et al. (2016) investi-
sia. Mechanical stimuli (i.e. pinprick stimuli) gated GMV of community-dwelling subjects.
were applied to the area to elicit secondary They found that in female subjects individuals
hyperalgesia. After capsaicin injection, EEG with chronic joint pain showed a smaller GMV
also revealed the pinprick-evoked potentials, at the hippocampus, the temporal and the fron-
which may reflect pain processing associated tal lobes (De Kruijf et al. 2016). The diversity of
with central sensitization (Van Den Broeke results may be largely attributed to the different
et al. 2015). A recent study revealed that pathological mechanisms of chronic pain. In
hypersensitivity induced by mechanical stim- terms of functional connectivity, an alteration
uli may be less pronounced when subjects in the DMN and the salience network is con-
were engaged with a cognitive–demanding sistently reported in several studies (De Groote
task, suggesting a modulatory effect of cogni- et al. 2020; Gupta et al. 2015; Karafin et al. 2019;
tive processing on sensitization (Torta Zempsky et al. 2017). Moreover, a recent study
et al. 2020). In addition to mechanical stimuli, on patients with chronic low back pain revealed
prolonged pain can also be elicited by heat that deficient mesocorticolimbic connectivity,
stimuli with the induction of capsaicin. The which may signify dysfunctions reward net-
pain intensity was correlated with the EEG work, may play a key role in increased pain sen-
signatures of the pain-free period preceding sitivity (Yu et al. 2020). In patients with chronic
pain induction (Furman et al. 2018). Heat spinal pain, the features of grey matter mor-
stimuli were also used to induce hyperalgesia, phology are significantly correlated with the
as shown in the experimental model of first- emotional representations, chronicity and pain
degree burn injury (Asghar et al. 2015). catastrophizing of female patients (Malfliet
et al. 2019). Therefore, the diversity of brain
6.2.5.2 Research on Patients with Chronic Pain mechanisms identified in these studies may
In addition to the experimental model, central reflect the difference in cognitive–affective pro-
sensitization has also been investigated for cessing of pain.
a variety of chronic pain, including fibromyal- It should be noted that in most of the studies
gia (Bosma et al. 2016; Hazra et al. 2020; a direct comparison between the clinical and
Kaplan et al. 2020; Lim et al. 2015), migraine the control groups was made to disclose the
(Aguila et al. 2016; Becerra et al. 2016; difference in the brain mechanisms associated
Chong et al. 2020), osteoarthritis (Soni with central sensitization. However, these dif-
et al. 2019), dysmenorrhea (Böttcher et al. 2019; ferences may not be fully attributed to central
Liu et al. 2017), whiplash pain (Coppieters sensitization. Because clinical symptoms (e.g.
et al. 2017; Coppieters et al. 2018) and chronic mechanical allodynia) are highly associated
with central sensitization, it is critical to inves- nociceptive neurons in the central nervous
tigate the association between the clinical system to their normal or subthreshold affer-
symptoms and the brain on an individual basis. ent input’.
For example, Chen et al. (2015) reported that ●● The symptom of mechanical allodynia
the patients with allodynia showed reduced should be carefully examined because it may
activation of the inferior parietal lobule com- indicate central sensitization. Secondary
pared to those without allodynia. Soni et al. hyperalgesia develops in the area surround-
(2019) assessed preoperative neuropathic-like ing the injury, which is also associated with
pain of patients with knee osteoarthritis using central sensitization.
a standard questionnaire. They found that ●● Neuroimaging findings have disclosed a
higher neural activity in the rostral ventrome- complicated pattern of the brain mecha-
dial medulla (RVM), a region known for pain nisms related to chronic pain. To affirm the
modulation, was associated with moderate-to- role of central sensitization in chronic pain,
severe pain after arthroplasty (Soni et al. 2019). the association between the brain and the
In another study, Aguila et al. (2016) investi- clinical symptoms of pain sensitization
gated the central sensitization of migraine would require further investigation.
patients using standard questionnaires. They
assessed the level of glutamate and γ-
aminobutyric acid (GABA), a critical modula- 6.3 Brain Mechanisms of Chronic
tor of pain, using magnetic resonance
Orofacial Pain
spectroscopy (MRS). They reported a positive
association between the GABA levels and the
6.3.1 Introduction
central sensitization scores (Aguila et al. 2016).
In sum, neuroimaging findings have disclosed As outlined in the preceding sections, chronic
a complicated pattern of the brain mechanisms orofacial pain is detrimental to individual well-
related to chronic pain. To affirm the role of being and therefore poses a great challenge in
central sensitization in chronic pain, the asso- healthcare. It is estimated that in the US, 5% of
ciation between the brain and the clinical adults have reported chronic pain in the face or
symptoms of pain sensitization would require jaw during a 3-month period (Benoliel
further investigation. et al. 2019). According to the latest classifica-
tion by the ICD-11, orofacial pain is catego-
rized into the ‘primary’ or ‘secondary’ domain.
6.2.6 Summary
Chronic primary orofacial pain includes
●● The classification of chronic pain is gener- chronic primary TMD pains, chronic burning
ally based on multiple criteria, including the mouth and pain with an idiopathic origin (e.g.
aetiology of pain, the underlying pathophys- persistent idiopathic dentoalveolar pain).
iological mechanisms and the body site Chronic secondary orofacial pain includes
of pain. chronic dental pain (derived from pulpal or
●● For patients with chronic pain, clinicians periapical lesions) and chronic neuropathic
need to consider not only the association orofacial pain (derived from lesions or diseases
between pain and tissue damage but also the related to the trigeminal nerve system)
cognitive–affective factors that contribute to (Benoliel et al. 2019). It is noteworthy that this
the chronic experience of pain. classification, just like the current classifica-
●● Sensitization may occur at the peripheral or tion of chronic pain in general, does not reflect
the central level (e.g. the spinal cord or the a clear association between clinical symp-
brainstem). Central sensitization is defined toms and pathological mechanisms (Treede
by the IASP as ‘increased responsiveness of et al. 2015, 2019). For dental pain with an
intraoral lesion (e.g. pulpitis or periodontitis), structural and functional features related to
treating the lesion would predict an improve- the thalamocortical system. A general hypoth-
ment of clinical symptoms. In contrast, pain esis is that compared to healthy controls
related to TMD or trigeminal neuropathy patients with TMD-related pain would show
shows a great diversity in clinical symptoms in alterations in the brain regions associated with
terms of individual variability in sensory and the thalamus and the somatosensory cortices,
emotional experiences (Nasri-Heir et al. 2016). which have been identified in various neuro-
Since pain is a multi-dimensional construct imaging studies of acute pain (see Section 6.1).
related to cognitive–affective processing (see However, the meta-analytic findings from neu-
Section 6.1), the role of our brain in the clinical roimaging studies of TMD-related pain reveal
diversity of chronic orofacial pain should not a different story. An earlier meta-analytic study
be neglected. This section focuses on the recent (Lin 2014) reveals that in the neuroimaging
neuroimaging findings on the brain mecha- research on TMD-related pain, (i) structural
nisms of chronic orofacial pain. Particularly, difference (i.e. GMV) in the thalamus and the
we focus on the potential brain mechanisms of somatosensory cortices is not consistently
pain related to TMD and trigeminal neuropa- found across the studies. (ii) In contrast, the
thy, two major challenges in the clinical man- brain regions with a difference in GMV are
agement of orofacial pain. found in a widespread area of the brain, includ-
ing the insula, the cingulate cortex and the
PFC. (iii) Both increased and decreased GMV
6.3.2 TMD-related Pain
are identified in the studies. It should be noted
Pain is one of the major chief complaints that the included studies are heterogenous in
reported by patients with TMD. Moreover, in their sub-classification of diseases (e.g. myo-
the patients, pain is usually associated with fascial TMD or idiopathic TMD) and experi-
functional limitations, such as a reduced mental methods (e.g. pain evoked by pressure
amount of maximal mouth opening. Therefore, or evoked by clenching). Therefore, it is not
TMD has a negative impact on patients’ quality surprising that the patterns of the brain mech-
of life, not only for pain but also for dysfunc- anisms vary across the studies. Still, the find-
tions for feeding. There has been a large body ings reveal more complicated brain
of neuroimaging evidence on the brain mecha- mechanisms of chronic orofacial pain com-
nisms of TMD-related pain. However, incon- pared to acute dental pain (Lin 2014). The
sistent findings have been reported, partly complex pattern of the brain mechanisms
because of the pathological complexity and associated with chronic orofacial pain may
great inter-individual heterogeneity in clinical reflect the diversity of cognitive–affective
symptoms and signs. The neuroimaging find- experiences of pain.
ings from recent studies are summarized in the In a recent study, Yin et al. (2020) have
following sections. reviewed 25 studies regarding the brain mecha-
nisms of TMD-related pain. In all the included
6.3.2.1 Findings from Meta-analytic Studies studies, patients were diagnosed with Research
Animal research has identified the role of the Diagnostic Criteria for TMD (RDC/TMD) or
thalamocortical system in pain processing. It Diagnostic Criteria for TMD (DC/TMD).
has been conceived that the lateral and the Among the included studies, the average dura-
medial pathways, respectively, take part in the tion of symptoms varied greatly, from 14 months
processing of sensory–discriminative informa- to 12 years, similar to the findings from the
tion and the emotional-motivational informa- earlier review (Lin 2014). Both reviews identi-
tion of pain (Groh et al. 2018). Therefore, fied a great number of positive (i.e. TMDs > HCs)
the earlier neuroimaging studies focused on and negative (i.e. TMDs < HCs) changes in
GMV. The study found a relatively consistent showed a reduced GMV (Younger et al. 2010)
structural difference between TMDs and con- and an increased cortical thickness (Moayedi
trols regarding the trigeminal nerve, including a et al. 2011), and the inconsistency was also
lower FA and decreased GMV of the nerve root found in the brain regions regarding the cogni-
(Yin et al. 2020). However, inconsistent findings tive–affective processing of pain, including an
emerged when it comes to subcortical and corti- increased GMV (Younger et al. 2010) and a
cal regions. For example, the trigeminal princi- decreased GMV (Gerstner et al. 2011) reported
pal sensory nucleus (Vp), a critical nucleus for in the anterior insula and the PFC.
nociception of the orofacial area, showed a
decreased GMV (Wilcox et al. 2015b) and an 6.3.2.2 Recent Neuroimaging Findings
increased GMV (Younger et al. 2010). The diver- of TMD-related Pain
sity of structural difference was also found at The recent neuroimaging findings of TMD-
the cortical level. For example, the S1 both related pain are summarized in Table 6.3. While
Structural
Wilcox et al. (2015b) 22 patients with painful TMD (40 healthy sMRI, dMRI/VBM and tractography
controls)
Functional, task-based
Vuong et al. (2020) 26/16 patients of chronic fatigue fMRI/the Valsalva manoeuvre
syndrome w/wo TMD
Ernst et al. (2020) 13 TMD patients receiving a mandibular fMRI before, within and after tx./
splint therapy occlusal movements
Wang et al. (2019a) 18 patients of chronic jaw pain (based on EEG/thermal stimuli at the right
TMD pain screening) (16 healthy controls) forearm
Roy et al. (2018) 16 patients of chronic jaw pain (based on fMRI/a force-production task and
TMD pain screening) (15 healthy thermal stimuli at the right forearm
controls)
Harper et al. (2016) 10 patients with myofascial-type TMD (10 fMRI/pressure pain in the thumbnail
healthy controls) and in the anterior temporalis
Functional, connectivity
Mills et al. (2020) 16 patients with painful TMD (45 healthy rs-fMRI/functional connectivity
controls)
Barkhordarian et al. (2020) 11 patients with comorbid TMD and rs-fMRI dMRI, ASL-MRI, fMRI/
systemic/neurologic conditions insertion of dental orthotics
Zhang et al. (2018a) 8 TMD patients with synovitis pain (10 rs-fMRI with mouth closed and
healthy controls) mouth open conditions/ReHo
He et al. (2018) 30 TMD patients (20 healthy controls) rs-fMRI/functional connectivity
Others
1
Harfeldt et al. (2018) 36 patients with generalized or regional H-MRS
chronic TMD pain
Notes: ASL-MRI: arterial spin labelling arterial spin labelling; dMRI: diffusion magnetic resonance imaging; EEG:
electroencephalography; fMRI: functional magnetic resonance imaging; MRS: magnetic resonance spectroscopy;
ReHo: regional homogeneity; rs-fMRI: resting-state functional magnetic resonance imaging; sMRI: structural
magnetic resonance imaging; TMD: temporomandibular disorders; VBM: voxel-based morphometry.
earlier neuroimaging studies mainly focused on decrease in pain over time. Notably, although
the structural brain features of TMD-related patients reported decreased pain during the
pain (Ayoub et al. 2018; Lin 2014; Yin treatment period, changes in kinematic and
et al. 2020), recent studies focused on the func- electromyography of jaw movement were not
tional features, including the intrinsic func- significant (Ernst et al. 2020). Finally, Vuong
tional connectivity of TMD-related pain. For et al. (2020) investigated the association between
structural brain mechanisms, Wilcox et al. TMD and the autonomic nervous system by
(2015b) investigated the grey matter and white comparing 26 patients with chronic fatigue syn-
matter morphology in TMD-related pain. In drome (CFS) and TMD and 16 CFS patients
TMD patients, they found a decreased GMV in without TMD. They found that when perform-
the medullary dorsal horn and alterations in ing the Valsalva manoeuvre (which activates
white matter in many regions of the descending the ANS), the CFS patients with TMD showed
pathway of pain modulation, including the greater activation in the left insular cortex com-
PAG. Moreover, decreased FA was found in the pared to the CFS patients without TMD (Vuong
trigeminal and the trigeminothalamic tracts in et al. 2020). The findings highlight the potential
the patients (Wilcox et al. 2015b). The findings role of the ANS in chronic orofacial pain, which
suggest an association between TMD-related has not been fully elucidated.
pain and alterations in the pathway of pain Recent studies have gradually focused on the
modulation. In terms of functional neuroimag- intrinsic functional connectivity of TMD-
ing research, Harper et al. (2016) investigated related pain. Zhang et al. (2018a) reported that
the fMRI data of patients with myofascial-type TMD patients with synovitis pain showed
TMD and healthy controls, who received pres- weaker connectivity between the anterior
sure pain on the anterior temporalis and the insula and the middle cingulate cortex, which
thumbnail, using MVPA. The pattern of brain are part of the salience network, compared to
activation can discriminate the pain derived healthy controls. Consistently, He et al. (2018)
from the temporalis and the thumb with a found that TMD patients showed decreased
better-than-average accuracy, only in the TMD connectivity between the ventral and dorsal
patients (Harper et al. 2016). Another task- corticostriatal circuitries. In addition, Mills
based fMRI study showed that patients with et al. (2020) investigated the pain-modulating
chronic jaw pain differed from healthy controls network at the subcortical level. They found
in the brain activation of the dorsolateral PFC, that TMD patients, compared to healthy con-
the ventral PMC and the inferior parietal lobule trols, showed a greater intrinsic functional
when they received heat stimuli at the forearm connectivity between the RVM and the trigem-
(Roy et al. 2018). Furthermore, an EEG study inal nucleus. However, the connectivity
showed that patients with chronic jaw pain between other brainstem regions (e.g. the PAG)
showed altered neural responses to heat stimuli was not significantly different between the
applied at the forearm compared to healthy con- groups (Mills et al. 2020). Barkhordarian et al.
trols (Wang et al. 2019a). It should be noted that (2020) reported that the alterations in brain
in both studies pain was elicited in the forearm activation differed between the TMD patients
but not the orofacial region. Therefore, the find- with and without dental orthotics. In patients
ings suggest that in the patients pain may be without dental orthotics, increased connectiv-
modulated at the CNS level, regardless of the ity in the salience network was identified
peripheral site of nociceptive input. A recent (Barkhordarian et al. 2020). In addition to
longitudinal fMRI on 13 TMD patients who the alterations in the salience network,
received splint therapy, Ernst et al. (2020) found researchers have identified increased func-
decreased activation in the anterior insula and tional connectivity of the DMN associated
the cerebellum, which were associated with a with TMD-related pain (Kucyi et al. 2014).
Alterations in the DMN have been widely within the patients, stronger mPFC connectiv-
reported in a variety of chronic pain. For exam- ity was associated with a higher score of pain
ple, Baliki et al. (2008a, 2008b) found that the rumination (Kucyi et al. 2014). Together, the
DMN which normally showed reduced con- findings suggest that TMD-related pain, simi-
nectivity in a task condition, showed reduced lar to other types of orofacial pain, is associ-
deactivation in patients with chronic pain ated with alterations of the intrinsic functional
when they were performing an attention task. connectivity of the DMN, and the alteration
Moreover, the activity of the mPFC, as a key may reflect individual differences in cognitive–
component of the DMN, was correlated with affective processing of pain.
the year of pain (Baliki et al. 2008a, 2008b).
Alterations in the DMN and the salience net-
6.3.3 Trigeminal Neuropathic Pain
work are also reported in recent neuroimaging
and Trigeminal Neuralgia
studies of chronic pain. Kucyi et al. (2014)
reported that during a resting state TMD Recent neuroimaging findings of trigeminal
patients showed a stronger intrinsic functional neuropathic pain (TNP) and TN are summa-
connectivity between the mPFC and the PCC rized in Table 6.4. In general, many studies
compared to healthy controls. Moreover, focused on the structural difference between
Table 6.4 (Continued)
Notes: ALFF: amplitude of low-frequency fluctuation; CTN: classic trigeminal neuralgia; dMRI: diffusion magnetic
resonance imaging; fMRI: functional magnetic resonance imaging; ITN: idiopathic trigeminal neuralgia; MRI:
magnetic resonance imaging; MS: multiple sclerosis; MVD: microvascular decompression; NVC: neurovascular
compression; PDAP: persistent dentoalveolar pain disorder; ReHo: regional homogeneity; rs-fMRI: resting-state
functional magnetic resonance imaging; sMRI: structural magnetic resonance imaging; TBSS: tract-based spatial
statistics; TN: trigeminal neuralgia; VBM: voxel-based morphometry.
TN/TNP patients and healthy controls. Several anterior cerebellum. The PAG, as a key compo-
studies focused on the features of grey matter nent for pain modulatory pathway, found a
and white matter, using voxel-based morpho- consistent increase in GMV (Henssen
metry (VBM) and diffusion magnetic reso- et al. 2019). The findings are consistent with
nance imaging (dMRI, respectively). For the results from an earlier review, showing that
functional features, many studies focused on more brain regions were associated with a neg-
the difference in intrinsic functional connec- ative alteration (i.e. TN/TNPs < healthy con-
tivity between TN/TNP patients and healthy trols) in GMV compared to a positive alteration
controls. In addition, there has been an increas- (Lin 2014). In patients with TN/TNP, the wide-
ing number of studies on burning mouth syn- spread decrease in GMV, especially in the fron-
drome (BMS), as summarized in Table 6.5. tal lobe, has also been reported in the
meta-analyses of structural features of other
6.3.3.1 Structural Features of Trigeminal chronic pain, including neuropathic pain (Pan
Neuropathic Pain or Trigeminal Neuralgia et al. 2015) and migraine (Jia and Yu 2017).
In a recent meta-analysis, Henssen et al. (2019) The pattern of decreased GMV is also identified
have reviewed eight studies that investigated in recent neuroimaging findings (Table 6.4).
the difference of GMV between patients with Compared to healthy controls, TN patients
TNP/TN and healthy controls. Using activa- showed a decreased local gyrification index
tion likelihood estimation (ALE), they found a (LGI) in the left insula (Wang et al. 2018),
widespread decrease in GMV of the thalamo- decreased GMV in the insula, the secondary
cortical pathway (i.e. including the thalamus somatosensory cortex and the ACC (Wang
and the S1), the network of cognitive–affective et al. 2017a, 2019b). Furthermore, an investi-
processing of pain (i.e. including the ACC, the gation with ultra-high field MRI revealed
bilateral insula and the middle frontal gyrus) decreased GMV and thickness in the caudal-
and brain regions related to motor functions, anterior and posterior cingulate cortex rather
including the M1, bilateral striatum and the than the rostral-anterior cingulate cortex
Table 6.5 Neuroimaging research on brain mechanisms of burning mouth syndrome (since 2015).
Kohashi et al. (2020) 15 patients of primary BMS (15 healthy fMRI/thermal stimuli on the right
controls) palm and right lower lip
Tan et al. (2019) 26 BMS patients (27 healthy controls) sMRI, DTI, rs-fMRI/VBM, functional
connectivity
Lee et al. (2019) 12 BMS patients (14 healthy controls) sMRI, ASL-MRI
Yoshino et al. (2017) 27 BMS patients (21 healthy controls) fMRI/tactile stimuli, angry vs.
neutral face pictures
Wada et al. (2017) 14 BMS patients (14 healthy controls) dMRI/tractography, graph-based
network analysis
Shinozaki et al. (2016) 16 primary BMS patients (15 healthy controls) fMRI/heat pain stimuli on the right
palm or right lower lip
Sinding et al. (2016) 12 idiopathic BMS patients and 17 subjects sMRI/VBM
with dysgeusia (13 healthy controls)
Notes: ASL-MRI: arterial spin labelling magnetic resonance imaging; BMS: burning mouth syndrome; dMRI:
diffusion magnetic resonance imaging; DTI: diffusion tensor imaging; fMRI: functional magnetic resonance
imaging; rs-fMRI: resting-state functional magnetic resonance imaging; sMRI: structural magnetic resonance
imaging; VBM: voxel-based morphometry.
(Moon et al. 2018). A decreased GMV was also zone (Lutz et al. 2016). In patients with multi-
found in the cerebellum (Li et al. 2017; Tsai ple sclerosis and TN, lesions in the pontine
et al. 2018). In addition to the brain regions trigeminal afferents were identified (Fröhlich
associated with sensorimotor processing, et al. 2018). Consistently, a recent study of the
reduction in hippocampal/parahippocampal TN patients with solitary pontine lesions
volume has been identified in the TN patients revealed that the lesions were associated with
compared to healthy controls (Hayes lower FA and higher axial diffusivity (AD),
et al. 2017; Li et al. 2017; Vaculik et al. 2019; radial diffusivity (RD) and mean diffusivity
Wang et al. 2019b). In addition, some neuroim- (MD), compared to the unaffected side
aging findings further revealed the potential (Tohyama et al. 2020). In general, the findings
pathological mechanisms of TN. Neurovascular suggest that in TN patients morphological fea-
compression (NVC) is considered one of the tures of the trigeminal nerve can be identified
major pathophysiological mechanisms of TN using dMRI. Likewise, dMRI revealed that
(Jones et al. 2019). A recent study reported that patients with painful trigeminal neuropathy
the posterior fossa volume was significantly showed higher FA and lower MD within the
smaller in patients without NVC compared to subnucleus oralis of the spinal trigeminal
those with NVC (Hardaway et al. 2019). Using nucleus (Wilcox et al. 2015a). Using 7-Tesla
VBM and surface-based morphometry, Wu MRI, Rutland et al. (2019) found that TN
et al. (2020) found that TN patients with NVC patients showed a lower FA, higher MD and
showed a lower GMV and cortical thickness in higher RD, in the tract between the thalamus
a widespread area compared to TN patients and the S1 ipsilateral to their NVC sites,
without NVC. The TN patients who did not compared to healthy controls. Together, the
respond to surgical treatment (i.e. non- findings suggest altered white matter micro-
responders to microvascular decompression) structure in the pathway of orofacial sensory
showed a larger trigeminal nerve volume and processing. However, the pattern of alterations
hippocampus volume compared to the in the brain may be more complicated.
responders (Danyluk et al. 2020). Notably, a In TN patients compared to healthy controls,
retrospective study on MRI pathologies Tian et al. (2016) reported a widespread
revealed that, in 125 patients with orofacial increase in AD in the tracts associated with
pain, trigeminal neurovascular contact was sensory and cognitive–affective processing of
identified in 22.4% of the patients (Devine pain. Wang et al. (2017a) reported an increased
et al. 2019). The findings suggest a potential MD in the corpus callosum, the corona radiata,
link between the brain and the deficits in the the superior longitudinal fasciculus and the
trigeminal nerve, which requires further internal and external capsule. Consistently,
clarification. using a machine-learning approach, Zhong
In recent years, more findings have been et al. (2018) found that white matter connectiv-
reported on the alterations of white matter ity of multiple regions associated with sensory
morphology, based on the research on and cognitive–affective processing of pain,
dMRI. Recent studies focused on the structural including the insula and the parietal lobules,
difference of the trigeminal nerve system can accurately differentiate between TN
between TNP/TN patients and healthy con- patients and healthy controls. Notably, indi-
trols. The TN patients showed higher FA in the vidual variability of the structural features may
cisternal segment and lower FA in the root be associated with therapeutic effects. For
entry zone (REZ) of the trigeminal nerve example, DeSouza et al. (2015) investigated TN
(Chen et al. 2016). TN patients also showed a patients and found thinner cortical thickness
lower FA within the zone of the affected at the ventral anterior insula as well as lower
trigeminal nerve compared to the contralateral FA and higher MD, RD and AD in trigeminal
REZ, compared to healthy controls. After Alshelh et al. (2016) found an increased infra-
microvascular decompression or gamma knife slow oscillatory activity in the ascending pain
radiosurgery (GKRS), the cortical thickness of pathway (i.e. the spinal trigeminal nucleus, the
the ventral anterior insula, as well as abnor- somatosensory thalamus and the S1) in patients
malities of white matter, became reversed to with orofacial neuropathic pain. In another
the same level as healthy controls (Desouza study of T2 relaxometry, the researchers found
et al. 2015). Hung et al. (2017) further reported decreased T2 relaxation times, an index indica-
that an increased AD in the trigeminal pontine tive of astrogliosis, in ascending pain pathways
fibres was associated with TN patients who did in TNP patients (Alshelh et al. 2018). The
not respond to surgical therapy. In contrast, increased oscillatory activity and prolonged
the responders to GKRS showed significantly astrocyte activation may relate to thalamocorti-
lower FA at the affected nerve compared to the cal dysrhythmia, which plays a key role in the
unaffected nerve (Tohyama et al. 2018). constant perception of pain (Alshelh
et al. 2016, 2018). Patients with chronic neuro-
6.3.3.2 Functional Features of Trigeminal pathic pain at the jaw showed increased func-
Neuropathic Pain or Trigeminal Neuralgia tional connectivity between the pain-modulatory
Earlier neuroimaging research focuses on the regions of the brainstem, including the connec-
brain activation associated with TNP and TN, tivity between the RVM and the spinal trigemi-
when patients were receiving peripheral nal nucleus as well as the connectivity between
stimuli. For example, Gustin et al. (2012) have the PAG and cortical regions, including the hip-
identified the re-organization of the soma- pocampus, NAc and ACC (Mills et al. 2018). At
tosensory region by recording the brain activa- the cortical level, the connectivity of the DMN,
tion responding to tactile stimuli (Gustin i.e. the connectivity between the PCC and the
et al. 2012). In their study, the organization of mPFC, was weaker in TN patients (Zhang
the S1 was quantified by the distance between et al. 2019). TN patients also showed altered
the central sulcus and loci of peak activation to functional connectivity in the brain regions
sensory stimuli (i.e. brushing at the lip, thumb associated with cognitive–affective processing
and little finger). They found that a significant of pain, including the connections between the
difference in the distance between patients with amygdala and the PFC. Alterations of the
painful trigeminal neuropathy and healthy con- amygdala-PFC connectivity were associated
trols suggests S1 re-organization associated with pain relief, as reported in longitudinal
with neuropathic pain. Notably, the re- research (Zhang et al. 2018b). Just recently, sev-
organization was not found in patients with eral studies investigated the amplitude of low-
TMD-related pain (Gustin et al. 2012). In a frequency fluctuations (ALFF) associated with
recent fMRI study on patients with persistent chronic orofacial pain. The ALFF is used to
dentoalveolar pain disorder (PDAP), the quantify spontaneous neuronal activity during
researchers applied pressure stimuli on the den- the resting state. They found that classical TN is
toalveolar site on patients and healthy controls. associated with a lower ALFF in multiple brain
They found increased activation in the S1, S2 regions, including the insula and the PFC
and the PFC (Moana-Filho et al. 2015). The (Zhang et al. 2019). In contrast, a higher ALFF
findings suggest that somatosensory processing in multiple regions, including the precentral
may play a key role in the altered nociceptive gyrus, the temporal and occipital lobes and the
processing in the neuropathic condition. Recent cerebellum, was also reported (Chen et al. 2019).
neuroimaging research has further extended Other studies reported both a higher ALFF in
our understanding of the functional features of temporal, occipital and frontal regions and a
TNP and TN from the perspective of the func- lower ALFF in the mPFC associated with the
tional connectome. Using resting-state fMRI, DMN (Wang et al. 2017b). Another commonly
used index of functional connectivity is regional lobule, and notably, the S1 activation was
homogeneity (ReHo), an index of local func- stronger when they received the stimulation
tional connectivity that reflects the synchroni- while watching angry (vs. neutral) facial
zation of brain activation (i.e. the BOLD expressions (Yoshino et al. 2017). The findings
timeseries) of the voxels within a focal region suggest a potential modulatory role of emo-
(Jiang and Zuo 2016). Alterations in ReHo were tional context in BMS. In terms of the struc-
found in multiple brain regions, between tural features of the brain, a lower GMV was
patients with classical TN and healthy controls identified in the thalamus and middle tempo-
and between TN patients before and after ral gyrus (Lee et al. 2019) and the ventromedial
receiving therapy (Dou et al. 2016; Xiang PFC (Tan et al. 2019) in BMS patients com-
et al. 2019). Alterations in ReHo were also found pared to healthy controls. Notably, when com-
in multiple brain regions, including the frontal pared to patients with dysgeusia, BMS patients
and temporal lobes and the cerebellum (Wang showed a lower GMV in the temporal lobe, the
et al. 2015b; Yuan et al. 2018). Consistently, a PCC and the cerebellum, and a higher GMV
recent neuroimage meta-analysis reported con- was found in the insula and the dorsolateral
sistent brain regions with an increased ReHo at PFC (Sinding et al. 2016). Alterations in struc-
the bilateral middle and superior frontal gyri tural connectivity, as assessed using diffusion
and the PCC (Henssen et al. 2019). Notably, tensor imaging (DTI), were found in the brain
brain regions of sensory processing, such as the regions of cognitive–affective processing
S1 and the S2, are less identified for changes of (including the OFC and the ACC) but not the
ALFF and ReHo. In general, all the indices of regions of sensory processing (Wada
functional connectivity reveal a different pat- et al. 2017).
tern of connectivity between patients and
healthy controls. Altered functional connectiv-
6.3.4 Neurochemical Features Related
ity within subcortical regions and between sub-
to Chronic Orofacial Pain
cortical and cortical regions (especially the PFC)
have both been identified. Evidence from animal research suggests that
neurotransmitters of the CNS, such as GABA,
6.3.3.3 Burning Mouth Syndrome play a critical role in modulating pain process-
Recent neuroimaging findings of the BMS are ing (Goudet et al. 2009). Glutamate and GABA
summarized in Table 6.5. In two fMRI studies, are major excitatory and inhibitory neuro-
the researchers investigated the brain activa- transmitters, respectively, in the mammalian
tion of thermal stimuli applied on the palm CNS (Goudet et al. 2009). In human subjects,
and the lower lip (Kohashi et al. 2020; alterations in the level of neurotransmitters are
Shinozaki et al. 2016). They found that the rep- mainly assessed using MRS, which measures
etition of noxious heat stimulus was associated the chemical composition of brain tissues and
with greater activation in the insula, the orbit- estimates the concentration of neurotransmit-
ofrontal cortex (OFC), the PFC and the ACC, ters (Gazzaniga et al. 2019). Using single-voxel
in BMS patients, compared to healthy controls proton MRS (1H-MRS), Gerstner et al. (2012)
(Shinozaki et al. 2016). Furthermore, reduced investigated the level of neurotransmitters in
activation from the first to the fourth sequence TMD patients with myofascial pain before and
of stimulation, which may indicate temporal after a pressure pain test. They found a higher
habituation, was found in the ACC in the BMS level of N-acetylaspartate (NAA) and choline
patients (Shinozaki et al. 2016). In another in TMD patients, compared to healthy con-
study, tactile stimuli were applied intraorally trols, before pressure pain at the anterior tem-
in BMS patients. In the patients, activation was poralis. After the pressure pain test, there was
found in the S1, the SMA and superior parietal a significant decrease in the glutamate level.
Moreover, in TMD patients, a higher NAA findings on dental practice, especially for diag-
level of the left posterior insula was associated nosing and classifying chronic orofacial pain,
with a longer pain symptom duration (Gerstner are discussed in the following sections.
et al. 2012). The findings suggest that altera-
tions in the level of neurotransmitters of pain 6.3.5.1 Brain Mechanisms Between Acute
modulation may be associated with individ- and Chronic Pain
ual variability in TMD-related pain. In To the general public, chronic pain is some-
patients with neuralgia or neuropathy, Gustin times conceived as a longer version of acute
et al. (2011) found a significant decrease in pain, and the approach of managing acute pain
the NAA/creatine ratio, i.e. a neurochemical should naturally apply to chronic pain. For
marker of neural viability, compared to healthy example, to treat acute pulpitis, one needs to
controls. Notably, the decrease in NAA/cre- tackle the infection/inflammation of pulpal
atine ratio was identified in the thalamic tissues. When the same logic is applied to
region, which also showed a smaller GMV in TMD-related pain, the peripheral sites with tis-
the patients with neuropathic pain (but not sue damage need to be tackled first in order to
TMD patients) compared to healthy controls relieve pain. However, more and more clinical
(Gustin et al. 2011). Furthermore, they found a and neuroimaging evidence has revealed that
significantly lower GABA level in the patients chronic pain should be taken as an entity with
with painful trigeminal neuropathy compared its specific mechanisms rather than merely the
to healthy controls in the thalamus (Henderson pain lasting for a longer period (Treede
et al. 2013). The findings highlight that neuro- et al. 2019).
chemical modulation in pain processing, espe- For example, neuroimaging meta-analyses
cially at the thalamus, may play a key role in have revealed that acute toothache induced
orofacial neuropathic pain. Extending these experimentally via electrical stimuli is associ-
findings, Harfeldt et al. (2018) found that a ated with a consistent activation of the thala-
higher glutamate level was associated with the mus, the somatosensory cortex, the insula and
temporal summation to painful stimuli in 36 the cingulate cortex (Lin et al. 2014b).
patients with generalized or regional TMD- Activation of these brain regions was also iden-
related pain. In addition, a higher level of cho- tified in a recent meta-analysis that includes
line was associated with a decreased maximal acute orofacial pain induced by electrical, ther-
range of mouth opening and pressure pain mal, chemical or mechanical stimuli (Ayoub
threshold of the hand (Harfeldt et al. 2018). et al. 2018). The pattern of brain activation is,
Together, the findings revealed that variations however, much different from the pattern of
in the neurochemical features may be associ- chronic orofacial pain, which included activa-
ated with individual differences in the pain- tion not only in the sensory regions but also in
related symptoms. the widespread brain regions of cognitive–
affective processing of pain (Lin 2014). Ayoub
et al. (2018) investigated the pattern of struc-
6.3.5 Applications of Neuroimaging
tural and functional features (including results
on Chronic Orofacial Pain
of task-based fMRI and intrinsic functional
Conclusions from the recent neuroimaging connectivity) of patients with chronic orofacial
findings and previous reviews/meta-analyses pain (including patients with TNP, BMS and
provide a complex picture of the brain mecha- TMD-related pain). The meta-analysis of ten
nisms of chronic orofacial pain. There exists a studies of VBM revealed a consistently larger
pattern of widespread alterations in structural GMV in the right ventral thalamus and poste-
features and intrinsic functional connectivity. rior putamen in patients compared to healthy
The clinical implications of the neuroimaging controls. Another meta-analysis of eight
studies of functional features revealed a con- Consistent activation in the parietal and frontal
sistent higher activity in the left medial and regions, which have been reported in acute
posterior thalamus and lower activity in the left toothache (Ayoub et al. 2018, Lin et al. 2014b),
posterior insula in patients compared to healthy was not identified in the results of chronic oro-
controls (Ayoub et al. 2018) (Figure 6.4a). facial pain. The thalamus and the posterior
Figure 6.4 Brain features associated with chronic orofacial pain. (a) Brain activation associated with
chronic orofacial pain. Meta-analytical findings reveal a consistent pattern of higher brain activation in the
left medial and posterior thalamus and lower brain activation in the left posterior insula in patients with
chronic orofacial pain (COFP) compared to healthy controls. Source: Ayoub et al. (2018). Reproduced with
permission of Elsevier. (b) Functional connectivity of patients with temporomandibular disorder (TMD)-
related pain. The left panel reveals that TMD patients showed enhanced functional connectivity (FC)
between the medial prefrontal cortex (mPFC) and the posterior cingulate cortex (PCC)/precuneus
(PCu)/retrosplenial cortex (RSC) compared to healthy controls. The right panel reveals that functional
connectivity between the mPFC and medial thalamus/PAG was positively correlated with pain rumination
in TMD patients. Source (insets): Kucyi et al. (2014), p.3969-3975 with permission of the Society for
Neuroscience under the terms of the Creative Commons Attribution 4.0 International License.
insula have been identified for both acute and individual variability in cognitive–affective
chronic orofacial pain, suggesting their roles in processing of pain. For example, Kucyi et al.
the processing of the sensory information of (2014) found increased functional connectivity
pain. In contrast, the diverse findings of corti- of the DMN in patients with TMD-related pain
cal brain regions (e.g. the S1/S2, the anterior compared to healthy controls (Figure 6.4b).
insula and the parietal and prefrontal regions) Notably, they found that individual variations
may imply that in patients with chronic pain a in this strength of functional connectivity
greater individual difference in symptoms may between the mPFC and the PAG, a key region
be associated with the diversity of the brain of pain modulation, were associated with the
mechanisms of the cognitive–affective experi- score of pain rumination (Kucyi et al. 2014).
ence of pain. In contrast, in healthy controls, greater ability
of spontaneous mind-wandering (which
6.3.5.2 Brain Mechanisms Related reflects the disengagement of the attention to
to Individual Differences in Chronic pain) was associated with the connectivity of
Orofacial Pain the DMN (Kucyi et al. 2013). The findings
In neuroimaging research on chronic orofacial show that excessive attention to pain, as repre-
pain, most studies adopted a cross-sectional sented by the higher score of pain rumination,
design to compare a patient group with a con- may play a key role in chronic pain. Together,
trol group. The studies identify the brain the findings suggest that the variability in
regions where structural or functional fea- structural or functional features may reflect
tures show a significant between-group the individual difference in cognitive–affective
difference. The findings reveal the brain processing of pain.
mechanisms associated with pain but do not
necessarily reveal the brain mechanisms of
6.3.6 Summary
individual variability in clinical symptoms
within the patient group. In order to evaluate ●● In patients with TMD-related pain, a wide-
the individual difference in clinical symptoms spread area of the brain shows difference in
(e.g. duration or perceived intensity of pain), GMV, including the insula, the cingulate
more analyses should be focused on the asso- cortex and the PFC. Both increased and
ciation between the brain features and clinical decreased GMV are identified in the studies.
symptoms at the individual level. For exam- ●● In patients with TN or TNP, a widespread
ple, recent studies show that in TN patients decrease in GMV is identified in the thalam-
the variability of pain duration was associated ocortical pathway (i.e. including the thala-
with structural features, such as regional brain mus and the S1) and the network of
volume and cortical thickness (Moon cognitive–affective processing of pain, based
et al. 2018), and functional features, such as on meta-analytic evidence. Alterations of
the intrinsic functional connectivity of sub- the trigeminal nerve system are also identi-
cortical regions (Tsai et al. 2019). Higher pain fied using dMRI.
intensity was associated with weaker intrinsic ●● Different indices of intrinsic functional con-
functional connectivity within the DMN in nectivity (including the inter-regional con-
patients with classic TN (Zhang et al. 2019) nectivity, ALFF and ReHo) convergently
and a greater ReHo value in the cerebellum reveal a different pattern of connectivity
and the middle temporal gyrus (Yuan between patients and healthy controls.
et al. 2018) and in the left precentral gyrus Altered functional connectivity both within
(Wang et al. 2015b) in patients with idiopathic subcortical regions and between subcortical
TN. The association between pain-related and cortical regions (especially the PFC) is
symptoms and brain features may reflect identified.
●● Variations in the neurochemical features, as Finally, from the point of translational applica-
quantified by MRS, may be associated with tion, the framework of experimental design is
individual differences in the pain-related discussed.
symptoms.
●● In patients with chronic orofacial pain, the
6.4.2 Fear, Anxiety and Phobia
individual difference in symptoms may be
associated with the diversity of the brain The terms dental fear, dental anxiety and den-
mechanisms in the cortical regions, which tal phobia are used interchangeably in the lit-
play a key role in the processing of the cogni- erature. An obvious overlap of all the terms is
tive–affective experience of pain. the adjective ‘dental’, which links fear, anxiety
and phobia, to a dental setting. For example,
dental anxiety refers to ‘a state of anxiety elic-
6.4 Brain Mechanisms of Dental ited by the provision of dental care’
(Weiner 2011). It is dental-care-related anxi-
Fear and Anxiety
ety because anxiety is aroused specifically in a
dental setting (McNeil and Randall 2014).
6.4.1 Introduction
The ‘dental’ anxiety is different from the anxi-
The history of patients’ dread of dental treat- ety in a more generalized form or trait anxi-
ment may be as long as the history of dentistry ety, which represents a disposition to feel
itself. In one of the earliest articles on fearful anxious in any situation non-specific to a den-
dental patients, the author described a group tal setting. What needs to be clarified is the
of dental patients known as ‘nervous patients’, difference between fear, anxiety and phobia
who were characterized by feeling ‘the dread (Table 6.6). The difference between these
of the operation almost as terrible as the opera- terms should be examined from multiple
tion itself’ (Lewis 1899). The author suggested aspects, including cognitive, emotional,
that dentists should not suppress patients’ behavioural and physiological dimensions
expectation of pain, highlighting the role of (Armfield 2010). For example, both fear and
cognitive–affective processing (e.g. anticipat- anxiety are psychological constructs related
ing a threat) in such dread. In terms of clinical to the perception of a threat (see Section 5.3).
management, the author further addressed They are associated with similar emotional
that it may be ‘difficult to outline a specific for- experiences (e.g. feeling threatened) and
mula for all cases’ (Lewis 1899), which implies behavioural responses (e.g. trying to avoid
great individual differences in the patients’ the threat) towards a threat. However,
experience and behavioural responses. Over a fear and anxiety are different in that fear
century, though technologies of dental treat- usually relates to an identifiable immediate
ment have been advancing greatly, the man- threat, and anxiety relates to the uncertain
agement of fearful patients poses a major nature of a threat (Asmundson et al. 2007;
challenge to the dental practice today. Milgrom et al. 1995; Öst and Skaret 2013).
In this chapter, we summarize recent neuro- Distinguishing between fear and anxiety
imaging findings of the brain mechanisms would be critical for interpreting the results
related to dental fear/anxiety. Firstly, we clarify of clinical assessment (Armfield 2010). For
the definition of fear, anxiety and phobia in the example, in the Dental Anxiety Scale
context of dental treatment. Secondly, the (Corah 1969), the questions about the feeling
structural and functional features related to when individuals are waiting for coming
highly anxious or dental-phobic patients are treatment focus on the anticipation of a future
outlined. We focus on the brain features related and uncertain threat. Patients’ responses to
to the individual variability of fear and anxiety. these questions may reflect their anxiety
about dental treatment. The questions about (Oosterink et al. 2009). Therefore, the term
the feeling when individuals are having their ‘phobia’ (and its adjective form, phobic) should
teeth drilled focus on the experience towards be used carefully because it relates to the diag-
an immediate threat. Patients’ responses to nosis of mental disorders. It is noteworthy that
these questions may reflect their fear of den- there may exist a qualitative difference, in addi-
tal treatment. Distinguishing between these tion to a quantitative difference, between pho-
concepts is also critical for clinical manage- bia and fear/anxiety. While fear and anxiety are
ment because patients with different reasons emotional experiences about perceiving a
for their emotional disturbance would require threat, specific phobia is associated with ‘unrea-
different strategies for behavioural manage- sonable or irrational fear related to a specific
ment (Milgrom et al. 1995). object or situation’ (LeBeau et al. 2010).
While fear and anxiety are commonly used Quantitatively, a patient with dental phobia
to describe individual emotional experiences may perceive much stronger negative emotion
and behavioural responses towards a threat, the towards dental settings compared to the patients
term ‘phobia’ is used under the scope of psychi- who feel anxious/fearful about dental treat-
atry. As shown in Table 6.6, dental phobia is ment. In addition to this quantitative difference,
defined based on the Diagnostic and Statistical patients with dental phobia would engage
Manual (DSM) by American Psychiatric themselves with more maladaptive behaviour,
Association (LeBeau et al. 2010). As a common such as refusing treatment when it is necessary
sub-class of specific phobia, dental phobia or avoiding the information about oral health-
shows a higher prevalence (3.7%) than the other care. For phobic patients, extreme fear can be
types of specific phobia (e.g. height or spider triggered merely by thinking about dental-
phobia), according to a survey of Dutch adults related information.
6.4.3 Brain Mechanisms of Dental Fear number of publications of both structural and
and Anxiety functional features of dental fear/anxiety in
recent years. Notably, many of these studies
Pain and fear/anxiety are highly associated
focus on the difference in brain mechanisms
with dental treatment (Lin et al. 2017).
between patients with dental phobia and other
However, the brain mechanisms associated
types of specific phobia (Table 6.7). The recent
with dental fear/anxiety have been less investi-
neuroimaging findings of brain mechanisms
gated in neuroimaging research compared to
of dental fear, anxiety and phobia are discussed
the brain mechanisms of acute and chronic
in the following sections.
orofacial pain. There has been an increasing
Structural
Hilbert et al. (2015) 26 DP and 33 SP adults (37 healthy sMRI/VBM
controls)
Lueken et al. (2015) 26 DP and 33 SP adults (37 healthy sMRI/classification
controls)
Schienle et al. (2013) 45 DP adults (41 healthy controls) sMRI/VBM
Functional, fMRI
Yeung et al. (2019b) 20 healthy adults fMRI/audiovisual stimuli of dental
treatment
Said Yekta-Michael 20 dental students (20 controls) fMRI/video clips presenting a dental
et al. (2019) treatment
Stefanescu et al. (2018) 94 subjects (including DP and SP fMRI/visual stimuli, auditory stimuli
patients and healthy controls)
Feldker et al. (2017) 20 patients with panic disorder, 20 fMRI/disorder-related picture stimuli
SAD patients, 16 DP patients and 11
PTSD patients (67 healthy controls)
Henderson et al. (2016) 19 healthy adults fMRI/repetitive open–close jaw movements
w/wo the presence of ongoing pain
Halsband and Wolf (2015) 12 DP adults (12 healthy controls) fMRI after hypnotic intervention/
audiovisual stimuli
Lin et al. (2015) 17 healthy adults fMRI, rs-fMRI/picture stimuli of dental
treatment
Yu et al. (2015) 7 healthy adults fMRI/recordings of the noise from the
same dental instrument
Scharmüller et al. (2015) 20 DP adults (20 healthy controls) fMRI/visual stimuli
Schienle et al. (2014) 40 healthy adults fMRI/picture stimuli, under the
instruction of different tasks (distraction,
classification and fear of pain)
Meier et al. (2014) 15 healthy adults fMRI/fear conditioning (painful electric
stimuli on the right maxillary canine vs.
the right tibia)
Scharmüller et al. (2014b) 20 DP adults (20 healthy controls) fMRI/picture stimuli, under the
instruction of different tasks (distraction,
classification and fear of pain)
(Continued)
Table 6.7 (Continued)
Hilbert et al. (2014) 13 DP adults (13 healthy controls) fMRI/visual and auditory stimuli of dental
treatment
Lueken et al. (2014) 13 DP and 13 SP adults (13 healthy fMRI/picture stimuli presented in an
controls) anticipation and immediate perception
phase
Scharmüller et al. (2014a) 45 DP adults (41 healthy controls) fMRI/picture stimuli of dental treatment
Lin et al. (2013b) 15 healthy adults fMRI/fear conditioning (painful electric
stimuli on the upper incisor)
Lin et al. (2013a) 15 healthy adults fMRI/fear conditioning (painful electric
stimuli on the upper incisor)
Schienle et al. (2013) 45 DP adults (41 healthy controls) fMRI/picture stimuli of dental treatment
(phobic, disgust, fear and neutral content)
Hermann et al. (2013) 24 DP adults (21 healthy controls) fMRI/picture stimuli (phobic, disgust, fear
and neutral content)
Lueken et al. (2011) 12 DP and 12 SP patients fMRI/video stimuli of dental treatment
(17 healthy controls)
Said Yekta et al. (2009) 20 healthy adults fMRI/video stimuli of dental treatment
Functional, other methods
Alexopoulos et al. (2019) 16 DP adults (16 healthy controls) MEG/neutral, phobia-irrelevant and
phobia-relevant pictures
Racek et al. (2015) 21 patients with hypersensitive fNIRS/cold stimulation vs. non-painful
teeth control of a hypersensitive tooth
Wang et al. (2015a) 24 patients receiving orthodontic EEG before and after initial archwire
treatment (12 patients receiving placement
cognitive behavioural therapy and
12 controls)
Leutgeb et al. (2014) 23 DP adults (23 healthy controls) EEG/picture stimuli, under the
instruction of different tasks (distraction,
classification and fear of pain)
Scharmüller et al. (2012) 14 DP and 14 SP adults (14 healthy EEG/picture stimuli (disorder-relevant
controls) and affectively neutral content)
Schienle et al. (2011) 30 DP adults (30 healthy controls) EEG/picture stimuli of dental treatment
(phobic, disgust, fear and neutral content)
Köchel et al. (2011) 25 DP adults (24 healthy controls) fNIRS/auditory stimuli of dental drilling,
pleasant and neutral content
Leutgeb et al. (2011) 18 DP adults (18 healthy controls) EEG/picture stimuli of dental treatment
(phobic, disgust, fear and neutral content)
Notes: DP: dental phobia; EEG: electroencephalography; fMRI: functional magnetic resonance imaging; fNIRS:
functional near-infrared spectroscopy; MEG: magnetoencephalography; MRI: magnetic resonance imaging; PTSD:
post-traumatic stress disorder; rs-fMRI: resting-state functional magnetic resonance imaging; SAD: social anxiety
disorder; sMRI: structural magnetic resonance imaging; SP: specific phobia; VBM: voxel-based morphometry.
6.4.3.1 Research on Symptom-Provoking Tasks use a vicarious stimulus to mimic a threat related
Because fear and anxiety are both emotional to dental care. For example, if dental treatment is
experiences towards a threat, to elicit fear and something scary to patients, then during an
anxiety, a common experimental method is to experiment, the stimuli associated with dental
treatment will be presented to the patients. These conditioning paradigm that associated the
stimuli may be presented as the picture of the presence of visual cues and different levels of
dental instrument (i.e. a visual stimulus), the painful stimuli. Some painful stimuli can be
sound of a handpiece (i.e. an auditory stimulus, precisely predicted by a visual cue, while some
e.g., [Köchel et al. 2011; Yu et al. 2015]) or a video painful stimuli may occur unpredictably. In
of dental procedures (i.e. an audiovisual stimu- other words, the experimenters manipulated
lus). Furthermore, fear and anxiety elicited by both the severity of a threat (i.e. pain intensity)
the stimuli can be rated by subjects, and the asso- and the uncertainty that the threat will occur
ciation between brain features and subjective rat- (i.e. if pain is predictable or not). Notably, the
ings of fear and anxiety can be investigated. In uncertainty of a threat is highly associated
recent years, such a symptom-provocation para- with the anxiety one would perceive.
digm has been widely used to investigate the Individuals may feel less anxious when a vis-
brain mechanisms of dental fear/anxiety ual cue that always predicts mild pain is pre-
(Hermann et al. 2013; Lin et al. 2015; Said Yekta sented. In contrast, when the cue is not
et al. 2009; Said Yekta-Michael et al. 2019; predictive of the intensity of incoming pain
Scharmüller et al. 2012; Scharmüller et al. 2014b; (i.e. the following stimulus may be mild or
Schienle et al. 2014; Yu et al. 2015). In a recent strong pain), they may feel greater anxiety
neuroimaging meta-analysis, Yeung et al. (2019a) upon the presence of the cue (Ploghaus
investigated seven fMRI studies using the et al. 2001). Using this approach, researchers
symptom-provocation paradigm for non-phobic found that subjects perceived stronger pain
subjects. They found that the stimuli mimicking when they felt the stimuli are more unpredict-
dental treatment were associated with a consist- able compared to the condition when the stim-
ent pattern of brain activation at the bilateral uli were predictable (Lin et al. 2013b; Ploghaus
anterior insula (extending to the putamen and et al. 2001). The neuroimaging findings sug-
claustrum) and the right dorsal ACC (within the gest that pain can be potentiated by the anxiety
mPFC) (Yeung et al. 2019a). Notably, the sensory of a threat. The results highlight the impor-
cortices (e.g. the visual or the auditory cortex) do tance of managing patients’ anxiety in pain
not show a consistent activation because the management during dental practice.
studies adopted different stimulating protocols As noted above, the conditioning paradigms
(e.g. visual-only or auditory-only stimuli). In con- focus on the association between environ-
trast, the activation of the anterior insula and the mental cues and a threat. A similar paradigm
ACC may suggest that the processing of threat- can be used to investigate the mechanisms of
related information, as shown in previous neuro- how fear and anxiety are developed during
imaging findings of threat perception (Wiech clinical settings. Using fMRI, Meier et al.
et al. 2010), may play a key role in fear and anxi- (2014) investigated the brain activation asso-
ety, regardless of the materials for symptom- ciated with fear conditioning in 21 healthy
provoking tasks. adults. For the fear-conditioning paradigm,
subjects received painful electrical stimuli
6.4.3.2 Research on Pain paired with visual cues (i.e., conditioned
and Conditioning Tasks stimuli). Notably, the unconditioned stimuli
Except for the symptom-provoking design, were performed at two sites: the right maxil-
some studies adopted painful stimuli as a lary canine and the right tibia of the subjects.
threat to elicit fear and anxiety. In these stud- They found that subjects showed a higher
ies, experimenters manipulated not just the skin conductance response (as an index of
intensity of pain but also the expectancy of emotional arousal) when fear conditioning
pain by adopting a fear-conditioning task. For was performed for the canine compared to the
example, Lin et al. (2013a, 2013b) adopted a tibia. Moreover, for the canine but not the
increased activation at the PFC and the OFC 6.4.5 Factors Associated with Dental
when viewing symptom-provoking videos Fear and Anxiety
related to each phobia (Lueken et al. 2011).
Both results from the structural and func-
However, when viewing symptom-provoking
tional neuroimaging suggest that dental fear
pictures, only patients with snake phobia
and anxiety are associated with an extensive
(but not dental phobia) showed a shift of acti-
network of the brain, including multiple
vation from the cortical area to the midbrain
regions associated with cognitive–affective
area (Lueken et al. 2014), which may signify
processing. The great diversity of the pattern
the activation of the defensive fear system
of structural and functional features may not
(Mobbs et al. 2009). Additionally, when being
be surprising if we consider the variability of
provoked by the materials specific to their
emotional experiences across dental patients
disorders, only patients with spider phobia
(Milgrom et al. 1995; Weiner 2011). For exam-
(but not dental phobia) showed a greater
ple, some patients are scared of the presence
EEG density of the insula and the cingulate
of needle injection. Such a fear directly links
cortex (Scharmüller et al. 2012). In general,
to the detection of an immediate threat, for
the findings reveal a distinct pattern of func-
which the amygdala plays a key role
tional features between dental phobia and
(Ohman 2005). In contrast, other patients may
animal phobia.
be more anxious about whether the injection
is painful or not. Anxiety is associated with
6.4.4.2 Structural Features of Dental Phobia
one’s anticipation of a future threat, which
By comparing patients with dental phobia,
occurrence is unpredictable. The anterior
patients with other types of specific phobia
insula may play a more critical role in the
and healthy controls, Hilbert et al. found an
increased anxiety modulated by the uncer-
increased GMV in widespread areas, includ-
tainty of a threat (Simmons et al. 2013; Lin
ing the insula, the ACC, the PFC and the OFC
et al. 2013a; Wiech et al. 2010). Therefore,
(Hilbert et al. 2015). Notably, the pattern of
neuroimaging findings may provide more
structural features is similar to the pattern of
cues for clinicians to understand the psycho-
brain activation identified in symptom-
logical process related to patients’ fear and
provoking studies. Furthermore, when view-
anxiety. Firstly, individual differences in den-
ing the pictures with aversive context, female
tal fear and anxiety may be associated with the
subjects showed a greater activation at the
processing of threat-related information,
caudate nucleus, and the female subjects
which is associated with strong attention to a
with dental phobia, compared to the male
present threat or anticipation of a future
subjects, showed a greater GMV in the cau-
threat. As shown in a functional near-infrared
date nucleus (Schienle et al. 2013). The find-
spectroscopy (fNIRS) study of tooth stimula-
ings suggest a gender difference in the brain
tion, the prefrontal activity may reflect the dif-
mechanisms of dental phobia. In a recent
ference of attentional/cognitive processing
study, Lueken et al. (2015) used a machine-
between different sessions (e.g. before receiv-
learning approach to classify different sub-
ing/anticipating stimuli vs. after receiving
types of phobia, including dental and snake
stimuli) (Racek et al. 2015). Secondly, individ-
phobia, based on the GMV and white matter
ual differences in fear and anxiety may be
volume (WMV) of 55 brain regions from
associated with the personal traits related to
sMRI research. They found that both the data
threat perception. Using the fear-conditioning
from GMV and WMV discriminated between
paradigm, Lin et al. (2013a) reported that acti-
the sub-types of phobia with high accuracy
vation at the anterior insula, rather than the
(Lueken et al. 2015).
posterior insula, was associated with the activation in the ventromedial PFC and the lat-
increased anxiety during conditioning. When eral OFC (Hermann et al. 2013). The activation
the stimuli were delivered unpredictably, sub- of mPFC and the OFC is consistent with their
jects may feel increased pain due to uncer- roles in regulating negative emotions, such as
tainty of intensity (Lin et al. 2014a). The using skills of cognitive appraisal to re-appraise
findings echoed the role of anticipation of the threat value of a stimulus (Diekhof
pain in anxiety and also the importance of et al. 2011).
attention processing of dental stimuli (Racek Scharmüller et al. (2014b) investigated the
et al. 2015). Furthermore, such a feeling of brain mechanisms of distraction, another
uncertainty is associated with the degree of common cognitive–behavioural approach for
pain catastrophizing, i.e. a tendency to develop managing dental fear and anxiety. They
a catastrophic thought about pain (e.g. ‘the found that a pain-focusing task (i.e. to focus
pain must get worse’) (Lin et al. 2013b). on the fear elicited by pain) was associated
with higher activation in the amygdala com-
pared to a classification task (i.e. to decide if
6.4.6 Brain Mechanisms Associated
the content of stimuli is dental-related or
with Anxiety Management
not). When dental phobic subjects were dis-
Because dental fear and anxiety pose a great tracted from the fearful stimuli, the subjects
challenge in clinical management, it is impor- with a higher dental anxiety level showed a
tant that the brain mechanisms identified by greater activation at the amygdala and the
neuroimaging research should help to establish insula (Scharmüller et al. 2014b). The find-
better strategies for managing patients’ fear/ ings support the role of the amygdala and the
anxiety. In fact, one of the earliest neuroimag- insula in processing threat-related informa-
ing studies of pain and anxiety of dental treat- tion (McNaughton and Corr 2004; Mobbs
ment, published in 1981, was for the effect of et al. 2009). In general, the neuroimaging
management. In this pioneering study, Chen findings suggest two possible mechanisms
et al. (1981) reported the changes of the cortical for manipulating dental fear and anxiety.
power spectrum, recorded by EEG, during a Firstly, a stimulus with a lower threat value
hypnotic intervention on a single patient who may reduce patients’ fear and anxiety, which
received dental surgery. In an earlier study, EEG may be associated with activation of the ante-
was also used to investigate the brain activity rior insula and the amygdala. Secondly,
associated with the pain-relieving effect of cog- patients may develop some skills for regulat-
nitive–behavioural therapy during orthodontic ing their emotions, such as reappraising the
treatment (Wang et al. 2015a). Recently, neuro- threat value of a stimulus. The regulating
imaging research has been used to elucidate the process may be associated with the mPFC
brain mechanisms underlying cognitive–behav- and the OFC.
ioural therapy of dental phobia, such as the use
of cognitive appraisal strategy (Armfield and
6.4.7 Future Directions of Neuroimaging
Heaton 2013). Hermann et al. (2013) found that
of Dental Fear/Anxiety Research
when watching symptom-provoking pictures,
dental phobic patients showed a reduced activa- As a relatively young topic of oral neurosci-
tion over time in multiple regions, including the ence, the brain mechanisms of dental fear and
mPFC and ventrolateral PFC, which are associ- anxiety have been investigated in recent
ated with regulating emotional experiences 10 years. Still, the brain mechanisms underly-
(Diekhof et al. 2011). Notably, the dental phobic ing dental fear and anxiety have not yet been
subjects with a higher ability of cognitive fully elucidated. In the following sections, we
appraisal showed more reduction of activation focus on the experimental design of neuroim-
in the dorsomedial PFC and less reduction of aging research on dental fear and anxiety.
References
Aguila, M.R., Rebbeck, T., Leaver, A.M. et al. nociception: a meta-analytic and functional
(2016). The association between clinical connectivity study. Pain 160: 1245–1260.
characteristics of migraine and brain GABA Baliki, M.N. and Apkarian, A.V. (2015).
levels: an exploratory study. J. Pain 17: Nociception, pain, negative moods, and
1058–1067. behavior selection. Neuron 87: 474–491.
Alexopoulos, J., Steinberg, C., Liebergesell- Baliki, M.N., Chialvo, D.R., Geha, P.Y. et al.
Kilian, N.E. et al. (2019). Biased emotional (2006). Chronic pain and the emotional brain:
attention in patients with dental phobia. Eur. specific brain activity associated with
J. Neurosci. 49: 290–302. spontaneous fluctuations of intensity of
Alshelh, Z., Di Pietro, F., Youssef, A.M. et al. chronic back pain. J. Neurosci. 26:
(2016). Chronic neuropathic pain: it’s about 12165–12173.
the rhythm. J. Neurosci. 36: 1008–1018. Baliki, M.N., Geha, P.Y., Apkarian, A.V., and
Alshelh, Z., Di Pietro, F., Mills, E.P. et al. (2018). Chialvo, D.R. (2008a). Beyond feeling: chronic
Altered regional brain T2 relaxation times in pain hurts the brain, disrupting the default-
individuals with chronic orofacial neuropathic mode network dynamics. J. Neurosci. 28:
pain. Neuroimage Clin. 19: 167–173. 1398–1403.
Apkarian, A.V., Baliki, M.N., and Geha, Baliki, M.N., Geha, P.Y., Jabakhanji, R. et al.
P.Y. (2009). Towards a theory of chronic pain. (2008b). A preliminary fMRI study of
Prog. Neurobiol. 87: 81–97. analgesic treatment in chronic back pain and
Armfield, J.M. (2006). Cognitive vulnerability: a knee osteoarthritis. Mol. Pain 4: 47.
model of the etiology of fear. Clin. Psychol. Baliki, M.N., Geha, P.Y., Fields, H.L., and
Rev. 26: 746–768. Apkarian, A.V. (2010). Predicting value of
Armfield, J.M. (2010). How do we measure pain and analgesia: nucleus accumbens
dental fear and what are we measuring response to noxious stimuli changes in the
anyway? Oral Health Prev. Dent. 8: 107–115. presence of chronic pain. Neuron 66: 149–160.
Armfield, J.M. and Heaton, L.J. (2013). Baliki, M.N., Petre, B., Torbey, S. et al. (2012).
Management of fear and anxiety in the dental Corticostriatal functional connectivity
clinic: a review. Aust. Dent. J. 58: 390–407. predicts transition to chronic back pain. Nat.
quiz 531. Neurosci. 15: 1117–1119.
Asghar, M.S., Pereira, M.P., Werner, M.U. et al. Barkhordarian, A., Demerjian, G., and
(2015). Secondary hyperalgesia phenotypes Chiappelli, F. (2020). Translational research of
exhibit differences in brain activation during temporomandibular joint pathology: a
noxious stimulation. PLoS One 10: e0114840. preliminary biomarker and fMRI study.
Asmundson, G.J.G., Vlaeyen, J.W.S., and J. Transl. Med. 18: 22.
Crombez, G. (2007). Understanding and Becerra, L., Veggeberg, R., Prescot, A. et al.
Treating Fear of Pain. Oxford, UK: Oxford (2016). A “complex” of brain metabolites
University Press. distinguish altered chemistry in the cingulate
Ayoub, L.J., Seminowicz, D.A., and Moayedi, cortex of episodic migraine patients.
M. (2018). A meta-analytic study of Neuroimage Clin. 11: 588–594.
experimental and chronic orofacial pain Benoliel, R., Sharav, Y., and Eliav, E. (2010).
excluding headache disorders. Neuroimage Neurovascular orofacial pain. J. Am. Dent.
Clin. 20: 901–912. Assoc. 141: 1094–1096.
Ayoub, L.J., Barnett, A., Leboucher, A. et al. Benoliel, R., Svensson, P., Evers, S. et al. (2019).
(2019). The medial temporal lobe in The IASP classification of chronic pain for
predict outcome of surgical treatment for state fMRI study. Medicine (Baltimore)
trigeminal neuralgia. Cephalalgia 40: 586–596. 95: e5193.
Davis, K.D., Flor, H., Greely, H.T. et al. (2017). Draganski, B., Gaser, C., Busch, V. et al. (2004).
Brain imaging tests for chronic pain: medical, Neuroplasticity: changes in grey matter
legal and ethical issues and recommendations. induced by training. Nature 427: 311–312.
Nat. Rev. Neurol. 13: 624–638. Duerden, E.G. and Albanese, M.C. (2013).
De Groote, S., Goudman, L., Peeters, R. et al. Localization of pain-related brain activation: a
(2020). Magnetic resonance Imaging meta-analysis of neuroimaging data. Hum.
exploration of the human brain during 10 kHz Brain Mapp. 34: 109–149.
spinal cord stimulation for failed back surgery Ernst, M., Schenkenberger, A.E., Domin,
syndrome: a resting state functional magnetic M. et al. (2020). Effects of centric mandibular
resonance imaging study. Neuromodulation splint therapy on orofacial pain and cerebral
23: 46–55. activation patterns. Clin. Oral Investig. 24:
De Kruijf, M., Bos, D., Huygen, F.J. et al. (2016). 2005–2013.
Structural brain alterations in community Farrell, M.J., Laird, A.R., and Egan, G.F. (2005).
dwelling individuals with chronic joint pain. Brain activity associated with painfully hot
AJNR Am. J. Neuroradiol. 37: 430–438. stimuli applied to the upper limb: a meta-
Dehghan, M., Schmidt-Wilcke, T., Pfleiderer, analysis. Hum. Brain Mapp. 25: 129–139.
B. et al. (2016). Coordinate-based (ALE) Feldker, K., Heitmann, C.Y., Neumeister, P. et al.
meta-analysis of brain activation in patients with (2017). Transdiagnostic brain responses to
fibromyalgia. Hum. Brain Mapp. 37: 1749–1758. disorder-related threat across four psychiatric
Del Casale, A., Ferracuti, S., Rapinesi, C. et al. disorders. Psychol. Med. 47: 730–743.
(2015). Hypnosis and pain perception: an Flodin, P., Martinsen, S., Altawil, R. et al. (2016).
activation likelihood estimation (ALE) Intrinsic brain connectivity in chronic pain: a
meta-analysis of functional neuroimaging resting-state fMRI study in patients with
studies. J. Physiol. Paris 109: 165–172. rheumatoid arthritis. Front. Hum. Neurosci. 10: 107.
Derbyshire, S.W. (2000). Exploring the pain Fröhlich, K., Winder, K., Linker, R.A. et al.
“neuromatrix”. Curr. Rev. Pain 4: 467–477. (2018). Supratentorial lesions contribute to
Desouza, D.D., Davis, K.D., and Hodaie, trigeminal neuralgia in multiple sclerosis.
M. (2015). Reversal of insular and Cephalalgia 38: 1326–1334.
microstructural nerve abnormalities following Furman, A.J., Meeker, T.J., Rietschel, J.C. et al.
effective surgical treatment for trigeminal (2018). Cerebral peak alpha frequency predicts
neuralgia. Pain 156: 1112–1123. individual differences in pain sensitivity.
Devine, M., Rahman, N., Connor, S.E.J. et al. Neuroimage 167: 203–210.
(2019). Evaluation of routine magnetic Garcia-Larrea, L. and Peyron, R. (2013). Pain
resonance imaging of patients with chronic matrices and neuropathic pain matrices: a
orofacial pain. Int. J. Oral Maxillofac. Surg. review. Pain 154 (Suppl 1): S29–S43.
48: 48–55. Gazzaniga, M.S., Ivry, R.B., and Mangun,
Diekhof, E.K., Geier, K., Falkai, P., and Gruber, G.R. (2019). Cognitive Neuroscience: The
O. (2011). Fear is only as deep as the mind Biology of the Mind. W. W. Norton & Company.
allows: a coordinate-based meta-analysis of GBD 2015 Disease and Injury Incidence and
neuroimaging studies on the regulation of Prevalence Collaborators (2016). Global,
negative affect. Neuroimage 58: 275–285. regional, and national incidence, prevalence,
Dou, Z., Zhang, X., Yang, L. et al. (2016). and years lived with disability for 310 diseases
Alternation of regional homogeneity in and injuries, 1990–2015: a systematic analysis
trigeminal neuralgia after percutaneous for the Global Burden of Disease Study 2015.
radiofrequency thermocoagulation: a resting Lancet 388: 1545–1602.
Gerstner, G., Ichesco, E., Quintero, A., and Halsband, U. and Wolf, T.G. (2015). Functional
Schmidt-Wilcke, T. (2011). Changes in regional changes in brain activity after hypnosis in
gray and white matter volume in patients with patients with dental phobia. J. Physiol. Paris
myofascial-type temporomandibular disorders: 109: 131–142.
a voxel-based morphometry study. J. Orofac. Hansen, M.S., Asghar, M.S., Wetterslev, J. et al.
Pain 25: 99–106. (2018). The association between areas of
Gerstner, G.E., Gracely, R.H., Deebajah, A. et al. secondary hyperalgesia and volumes of the
(2012). Posterior insular molecular changes in caudate nuclei and other pain relevant brain
myofascial pain. J. Dent. Res. 91: 485–490. structures-A 3-tesla MRI study of healthy
Giesler, G.J. Jr., Katter, J.T., and Dado, men. PLoS One 13: e0201642.
R.J. (1994). Direct spinal pathways to the Hansen, M.S., Becerra, L., Dahl, J.B. et al. (2019).
limbic system for nociceptive information. Brain resting-state connectivity in the
Trends Neurosci. 17: 244–250. development of secondary hyperalgesia in
Goldberg, D.S. and McGee, S.J. (2011). Pain as a healthy men. Brain Struct. Funct. 224:
global public health priority. BMC Public 1119–1139.
Health 11: 770. Hardaway, F.A., Holste, K., Ozturk, G. et al.
Goudet, C., Magnaghi, V., Landry, M. et al. (2019). Sex-dependent posterior fossa
(2009). Metabotropic receptors for glutamate anatomical differences in trigeminal neuralgia
and GABA in pain. Brain Res. Rev. 60: 43–56. patients with and without neurovascular
Groh, A., Krieger, P., Mease, R.A., and compression: a volumetric MRI age-and
Henderson, L. (2018). Acute and chronic pain sex-matched case-control study. J. Neurosurg.
processing in the thalamocortical system of 132: 631–638.
humans and animal models. Neuroscience Harfeldt, K., Alexander, L., Lam, J. et al. (2018).
387: 58–71. Spectroscopic differences in posterior insula
Gupta, A., Rapkin, A.J., Gill, Z. et al. (2015). in patients with chronic temporomandibular
Disease-related differences in resting-state pain. Scand J Pain 18: 351–361.
networks: a comparison between localized Harper, D.E., Shah, Y., Ichesco, E. et al. (2016).
provoked vulvodynia, irritable bowel Multivariate classification of pain-evoked
syndrome, and healthy control subjects. Pain brain activity in temporomandibular disorder.
156: 809–819. Pain Rep. 1: e572.
Gustin, S.M., Peck, C.C., Wilcox, S.L. et al. Hayes, D.J., Chen, D.Q., Zhong, J. et al. (2017).
(2011). Different pain, different brain: Affective circuitry alterations in patients with
thalamic anatomy in neuropathic and trigeminal neuralgia. Front. Neuroanat. 11: 73.
non-neuropathic chronic pain syndromes. Hazra, S., Venkataraman, S., Handa, G. et al.
J. Neurosci. 31: 5956–5964. (2020). A cross-sectional study on central
Gustin, S.M., Peck, C.C., Cheney, L.B. et al. sensitization and autonomic changes in
(2012). Pain and plasticity: is chronic pain fibromyalgia. Front. Neurosci. 14: 788.
always associated with somatosensory cortex He, S., Li, F., Gu, T. et al. (2018). Reduced
activity and reorganization? J. Neurosci. 32: corticostriatal functional connectivity in
14874–14884. temporomandibular disorders. Hum. Brain
Hadanny, A., Bechor, Y., Catalogna, M. et al. Mapp. 39: 2563–2572.
(2018). Hyperbaric oxygen therapy can induce Henderson, L.A., Peck, C.C., Petersen, E.T. et al.
neuroplasticity and significant clinical (2013). Chronic pain: lost inhibition?
improvement in patients suffering from J. Neurosci. 33: 7574–7582.
fibromyalgia with a history of childhood Henderson, L.A., Akhter, R., Youssef, A.M. et al.
sexual abuse-randomized controlled trial. (2016). The effects of catastrophizing on
Front. Psychol. 9: 2495. central motor activity. Eur. J. Pain 20: 639–651.
Henssen, D., Dijk, J., Knepflé, R. et al. (2019). pain and healthy controls. Pain 157:
Alterations in grey matter density and 1279–1286.
functional connectivity in trigeminal Jia, Z. and Yu, S. (2017). Grey matter alterations
neuropathic pain and trigeminal neuralgia: a in migraine: a systematic review and meta-
systematic review and meta-analysis. analysis. Neuroimage Clin. 14: 130–140.
Neuroimage Clin. 24: 102039. Jiang, L. and Zuo, X.N. (2016). Regional
Hermann, A., Leutgeb, V., Scharmüller, W. et al. homogeneity: a multimodal, multiscale
(2013). Individual differences in cognitive neuroimaging marker of the human
reappraisal usage modulate the time course of connectome. Neuroscientist 22: 486–505.
brain activation during symptom provocation Johanek, L., Shim, B., and Meyer, R.A. (2006).
in specific phobia. Biol. Mood Anxiety Primary hyperalgesia and nociceptor
Disord. 3: 16. sensitization. Handb. Clin. Neurol. 81: 35–47.
Hilbert, K., Evens, R., Maslowski, N.I. et al. Jones, M.R., Urits, I., Ehrhardt, K.P. et al. (2019).
(2014). Fear processing in dental phobia A comprehensive review of trigeminal
during crossmodal symptom provocation: an neuralgia. Curr. Pain Headache Rep. 23: 74.
fMRI study. Biomed. Res. Int. 2014: 196353. Kaplan, C.M., Schrepf, A., Ichesco, E. et al.
Hilbert, K., Evens, R., Maslowski, N.I. et al. (2015). (2020). Association of Inflammation with
Neurostructural correlates of two subtypes of pronociceptive brain connections in
specific phobia: a voxel-based morphometry rheumatoid arthritis patients with
study. Psychiatry Res. 231: 168–175. concomitant fibromyalgia. Arthritis
Hung, P.S., Chen, D.Q., Davis, K.D. et al. (2017). Rheumatol. 72: 41–46.
Predicting pain relief: use of pre-surgical Karafin, M.S., Chen, G., Wandersee, N.J. et al.
trigeminal nerve diffusion metrics in trigeminal (2019). Chronic pain in adults with sickle cell
neuralgia. Neuroimage Clin. 15: 710–718. disease is associated with alterations in
Iannetti, G.D., Zambreanu, L., Wise, R.G. et al. functional connectivity of the brain. PLoS One
(2005). Pharmacological modulation of 14: e0216994.
pain-related brain activity during normal and King, M. and Carnahan, H. (2019). Revisiting the
central sensitization states in humans. Proc. brain activity associated with innocuous and
Natl. Acad. Sci. U. S. A. 102: 18195–18200. noxious cold exposure. Neurosci. Biobehav.
IASP (1994). Part III: pain terms, a current list Rev. 104: 197–208.
with definitions and notes on usage. In: Köchel, A., Plichta, M.M., Schäfer, A. et al.
Classification of Chronic Pain (eds. (2011). Auditory symptom provocation in
H. Merskey and N. Bogduk). Seattle: dental phobia: a near-infrared spectroscopy
IASP Press. study. Neurosci. Lett. 503: 48–51.
IASP. 2020. IASP Terminology [Online]. Kohashi, R., Shinozaki, T., Sekine, N. et al.
Available: https://www.iasp-pain.org/ (2020). Time-dependent responses in brain
terminology?navItemNumber=576 activity to ongoing hot stimulation in burning
[Accessed]. mouth syndrome. J. Oral Sci. 62: 170–174.
Jauniaux, J., Khatibi, A., Rainville, P., and Kucyi, A. and Davis, K.D. (2015). The dynamic
Jackson, P.L. (2019). A meta-analysis of pain connectome. Trends Neurosci. 38: 86–95.
neuroimaging studies on pain empathy: Kucyi, A., Salomons, T.V., and Davis,
investigating the role of visual information K.D. (2013). Mind wandering away from pain
and observers’ perspective. Soc. Cogn. Affect. dynamically engages antinociceptive and
Neurosci. 14: 789–813. default mode brain networks. Proc. Natl. Acad.
Jensen, K.B., Regenbogen, C., Ohse, M.C. et al. Sci. U. S. A. 110: 18692–18697.
(2016). Brain activations during pain: a Kucyi, A., Moayedi, M., Weissman-Fogel,
neuroimaging meta-analysis of patients with I. et al. (2014). Enhanced medial
two subtypes of specific phobia. Neuroimage McNeil, D.W., Au, A.R., Zvolensky, M.J. et al.
56: 363–372. (2001). Fear of pain in orofacial pain patients.
Lueken, U., Hilbert, K., Stolyar, V. et al. (2014). Pain 89: 245–252.
Neural substrates of defensive reactivity in Meier, M.L., De Matos, N.M., Brügger, M. et al.
two subtypes of specific phobia. Soc. Cogn. (2014). Equal pain-unequal fear response:
Affect. Neurosci. 9: 1668–1675. enhanced susceptibility of tooth pain to fear
Lueken, U., Hilbert, K., Wittchen, H.U. et al. conditioning. Front. Hum. Neurosci. 8: 526.
(2015). Diagnostic classification of specific Meier, M.L., Widmayer, S., Abazi, J. et al. (2015).
phobia subtypes using structural MRI data: a The human brain response to dental pain
machine-learning approach. J. Neural Transm. relief. J. Dent. Res. 94: 690–696.
(Vienna) 122: 123–134. Melzack, R. (1990). Phantom limbs and the
Lutz, J., Thon, N., Stahl, R. et al. (2016). concept of a neuromatrix. Trends Neurosci.
Microstructural alterations in trigeminal 13: 88–92.
neuralgia determined by diffusion tensor Melzack, R. (1999). From the gate to the
imaging are independent of symptom neuromatrix. Pain Suppl. 6: S121–S126.
duration, severity, and type of neurovascular Melzack, R. (2001). Pain and the neuromatrix in
conflict. J. Neurosurg. 124: 823–830. the brain. J. Dent. Educ. 65: 1378–1382.
Malfliet, A., Coppieters, I., Van Wilgen, P. et al. Melzack, R., Coderre, T.J., Katz, J., and
(2017). Brain changes associated with Vaccarino, A.L. (2001). Central neuroplasticity
cognitive and emotional factors in chronic and pathological pain. Ann. N. Y. Acad. Sci.
pain: a systematic review. Eur. J. Pain 21: 933: 157–174.
769–786. Milgrom, P., Weinstein, P., and Getz, T. (1995).
Malfliet, A., De Pauw, R., Kregel, J. et al. (2019). Treating Fearful Dental Patients: A Patient
Gender differences in the association of brain Management Handbook. University of
gray matter and pain-related psychosocial Washington, Continuing Dental Education.
characteristics. Pain Physician 22: E191–E203. Mills, E.P., Di Pietro, F., Alshelh, Z. et al. (2018).
Mansour, A.R., Baliki, M.N., Huang, L. et al. Brainstem pain-control circuitry connectivity
(2013). Brain white matter structural in chronic neuropathic pain. J. Neurosci. 38:
properties predict transition to chronic pain. 465–473.
Pain 154: 2160–2168. Mills, E.P., Akhter, R., Di Pietro, F. et al. (2020).
Mansour, A., Baria, A.T., Tetreault, P. et al. Altered brainstem pain modulating circuitry
(2016). Global disruption of degree rank functional connectivity in chronic painful
order: a hallmark of chronic pain. Sci. Rep. temporomandibular disorder. J. Pain 13:
6: 34853. 2223–2235.
McNaughton, N. and Corr, P.J. (2004). A Mitsi, V. and Zachariou, V. (2016). Modulation of
two-dimensional neuropsychology of defense: pain, nociception, and analgesia by the brain
fear/anxiety and defensive distance. Neurosci. reward center. Neuroscience 338: 81–92.
Biobehav. Rev. 28: 285–305. Moana-Filho, E.J., Bereiter, D.A., and Nixdorf,
McNeil, D.W. and Randall, C.L. (2014). Dental D.R. (2015). Amplified brain processing of
fear and anxiety associated with oral health dentoalveolar pressure stimulus in persistent
care: conceptual and clinical issues. In: dentoalveolar pain disorder patients. J. Oral
Behavioral Dentistry, 2e (eds. D.I. Mostofsky, Facial Pain Headache 29: 349–362.
A.G. Forgione and D.B. Giddon). Ames Moayedi, M., Weissman-Fogel, I., Crawley,
(IA): Wiley. A.P. et al. (2011). Contribution of chronic pain
McNeil, D.W., Rainwater, A.J., and 3rd (1998). and neuroticism to abnormal forebrain gray
Development of the fear of pain matter in patients with temporomandibular
questionnaire – III. J. Behav. Med. 21: 389–410. disorder. Neuroimage 55: 277–286.
Mobbs, D., Marchant, J.L., Hassabis, D. et al. Park, S.J., Zhang, S., Chiang, C.Y. et al. (2006).
(2009). From threat to fear: the neural Central sensitization induced in thalamic
organization of defensive fear systems in nociceptive neurons by tooth pulp stimulation
humans. J. Neurosci. 29: 12236–12243. is dependent on the functional integrity of
Moon, H.C., Park, C.A., Jeon, Y.J. et al. (2018). 7 trigeminal brainstem subnucleus caudalis but
Tesla magnetic resonance imaging of caudal not subnucleus oralis. Brain Res. 1112: 134–145.
anterior cingulate and posterior cingulate Pascual-Leone, A., Amedi, A., Fregni, F., and
cortex atrophy in patients with trigeminal Merabet, L.B. (2005). The plastic human brain
neuralgia. Magn. Reson. Imaging 51: 144–150. cortex. Annu. Rev. Neurosci. 28: 377–401.
Mouraux, A. and Iannetti, G.D. (2018). The Peyron, R., Laurent, B., and Garcia-Larrea,
search for pain biomarkers in the human L. (2000). Functional imaging of brain
brain. Brain 141: 3290–3307. responses to pain. A review and meta-analysis
Nasri-Heir, C., Epstein, J.B., Touger-Decker, R., (2000). Neurophysiol. Clin. 30: 263–288.
and Benoliel, R. (2016). What should we tell Ploghaus, A., Narain, C., Beckmann, C.F. et al.
patients with painful temporomandibular (2001). Exacerbation of pain by anxiety is
disorders about what to eat? J. Am. Dent. associated with activity in a hippocampal
Assoc. 147: 667–671. network. J. Neurosci. 21: 9896–9903.
Nishioka, K., Suzuki, M., Nakajima, M. et al. Ploner, M., Lee, M.C., Wiech, K. et al. (2011).
(2019). Painful legs and moving toes Flexible cerebral connectivity patterns
syndrome evaluated through brain single subserve contextual modulations of pain.
photon emission computed tomography: a Cereb. Cortex 21: 719–726.
case series. J. Neurol. 266: 717–725. Racek, A.J., Hu, X., Nascimento, T.D. et al.
Ohman, A. (2005). The role of the amygdala in (2015). Different brain responses to pain and
human fear: automatic detection of threat. its expectation in the dental chair. J. Dent. Res.
Psychoneuroendocrinology 30: 953–958. 94: 998–1003.
Ong, W.Y., Stohler, C.S., and Herr, D.R. (2019). Reddan, M.C. and Wager, T.D. (2018). Modeling
Role of the prefrontal cortex in pain pain using fMRI: from regions to biomarkers.
processing. Mol. Neurobiol. 56: 1137–1166. Neurosci. Bull. 34: 208–215.
Oosterink, F.M., De Jongh, A., and Hoogstraten, Roberts, C.A., Giesbrecht, T., Stancak, A. et al.
J. (2009). Prevalence of dental fear and phobia (2019). Where is itch represented in the brain,
relative to other fear and phobia subtypes. Eur. and how does it differ from Pain? An
J. Oral Sci. 117: 135–143. activation likelihood estimation meta-analysis
Ossipov, M.H., Dussor, G.O., and Porreca, of experimentally-induced itch. J. Invest.
F. (2010). Central modulation of pain. J. Clin. Dermatol. 139 (2245–2248): e3.
Invest. 120: 3779–3787. Roy, A., Wang, W.E., Ho, R.L.M. et al. (2018).
Öst, L.-G. and Skaret, E. (2013). Cognitive Functional brain activity during motor control
Behavioral Therapy for Dental Phobia and and pain processing in chronic jaw pain. Pain
Anxiety. Chichester, UK: Wiley. 159: 2547–2564.
Palermo, S., Benedetti, F., Costa, T., and Rutland, J.W., Huang, K.H., Gill, C.M. et al.
Amanzio, M. (2015). Pain anticipation: an (2019). First application of 7-T ultra-high field
activation likelihood estimation meta-analysis diffusion tensor imaging to detect altered
of brain imaging studies. Hum. Brain Mapp. microstructure of thalamic-somatosensory
36: 1648–1661. anatomy in trigeminal neuralgia.
Pan, P.L., Zhong, J.G., Shang, H.F. et al. (2015). J. Neurosurg.: 1–9. https://doi.org/10.3171/
Quantitative meta-analysis of grey matter 2019.6.JNS19541.
anomalies in neuropathic pain. Eur. J. Pain 19: Said Yekta, S., Vohn, R., and Ellrich, J. (2009).
1224–1231. Cerebral activations resulting from virtual
dental treatment. Eur. J. Oral Sci. 117: Serafini, R.A., Pryce, K.D., and Zachariou,
711–719. V. (2020). The mesolimbic dopamine system
Said Yekta-Michael, S., Schuppen, A., Gaebler, in chronic pain and associated affective
A.J. et al. (2019). Expertise modulates comorbidities. Biol. Psychiatry 87: 64–73.
students’ perception of pain from a self- Sharav, Y. and Benoliel, R. (2008). Orofacial Pain
perspective: quasi-experimental study. J. Med. and Headache. Elsevier Health Sciences.
Internet Res. 21: e10885. Shinozaki, T., Imamura, Y., Kohashi, R. et al.
Scharmüller, W., Leutgeb, V., Schäfer, A., and (2016). Spatial and temporal brain responses
Schienle, A. (2012). Investigating phobic to noxious heat thermal stimuli in burning
specificity with standardized low resolution mouth syndrome. J. Dent. Res. 95: 1138–1146.
brain electromagnetic tomography Simmons, W.K., Avery, J.A., Barcalow, J.C. et al.
(SLORETA). Brain Res. 1477: 74–82. (2013). Keeping the body in mind: insula
Scharmüller, W., Leutgeb, V., Schöngassner, functional organization and functional
F. et al. (2014a). Altered functional connectivity connectivity integrate interoceptive,
of basal ganglia circuitry in dental phobia. Soc. exteroceptive, and emotional awareness. Hum.
Cogn. Affect. Neurosci. 9: 1584–1588. Brain Mapp. 34: 2944–2958.
Scharmüller, W., Ubel, S., Leutgeb, V. et al. Sinding, C., Gransjoen, A.M., Schlumberger,
(2014b). Do not think about pain: neural G. et al. (2016). Grey matter changes of the
correlates of attention guiding during visual pain matrix in patients with burning mouth
symptom provocation in dental phobia – an syndrome. Eur. J. Neurosci. 43: 997–1005.
fMRI study. Brain Res. 1566: 69–76. Soni, A., Wanigasekera, V., Mezue, M. et al.
Scharmüller, W., Wabnegger, A., and Schienle, (2019). Central sensitization in knee
A. (2015). Functional brain connectivity osteoarthritis: relating presurgical brainstem
during fear of pain: a comparison between neuroimaging and paindetect-based patient
dental phobics and controls. Brain Connect. 5: stratification to arthroplasty outcome.
187–191. Arthritis Rheumatol. 71: 550–560.
Schienle, A., Köchel, A., and Leutgeb, V. (2011). Stefanescu, M.R., Endres, R.J., Hilbert,
Frontal late positivity in dental phobia: a K. et al. (2018). Networks of phobic fear:
study on gender differences. Biol. Psychol. 88: functional connectivity shifts in two subtypes
263–269. of specific phobia. Neurosci. Lett. 662:
Schienle, A., Scharmüller, W., Leutgeb, V. et al. 167–172.
(2013). Sex differences in the functional and Tan, Y., Wu, X., Chen, J. et al. (2019). Structural
structural neuroanatomy of dental phobia. and functional connectivity between the
Brain Struct. Funct. 218: 779–787. amygdala and orbital frontal cortex in burning
Schienle, A., Wabnegger, A., Schöengassner, F., mouth syndrome: an fMRI study. Front.
and Scharmüller, W. (2014). Neuronal Psychol. 10: 1700.
correlates of three attentional strategies Tanasescu, R., Cottam, W.J., Condon, L. et al.
during affective picture processing: an fMRI (2016). Functional reorganisation in chronic
study. Cogn. Affect. Behav. Neurosci. 14: pain and neural correlates of pain
1320–1326. sensitisation: a coordinate based meta-
Schmidt, R. and Willis, W. (2007). Encyclopedia analysis of 266 cutaneous pain fMRI studies.
of Pain. Berlin. Heidelberg: Springer. Neurosci. Biobehav. Rev. 68: 120–133.
Seminowicz, D.A., Wideman, T.H., Naso, L. et al. Tatu, K., Costa, T., Nani, A. et al. (2018). How do
(2011). Effective treatment of chronic low morphological alterations caused by chronic
back pain in humans reverses abnormal brain pain distribute across the brain? A meta-
anatomy and function. J. Neurosci. 31: analytic co-alteration study. Neuroimage Clin.
7540–7550. 18: 15–30.
Tian, T., Guo, L., Xu, J. et al. (2016). Brain white subfield volume reductions in classic
matter plasticity and functional reorganization trigeminal neuralgia. Neuroimage Clin.
underlying the central pathogenesis of 23: 101911.
trigeminal neuralgia. Sci. Rep. 6: 36030. Van Den Broeke, E.N., Mouraux, A., Groneberg,
Tohyama, S., Hung, P.S., Zhong, J., and Hodaie, A.H. et al. (2015). Characterizing pinprick-
M. (2018). Early postsurgical diffusivity evoked brain potentials before and after
metrics for prognostication of long-term pain experimentally induced secondary
relief after Gamma Knife radiosurgery for hyperalgesia. J. Neurophysiol. 114: 2672–2681.
trigeminal neuralgia. J. Neurosurg. 131: Van Den Broeke, E.N., De Vries, B., Lambert,
539–548. J. et al. (2017). Phase-locked and non-phase-
Tohyama, S., Hung, P.S., Cheng, J.C. et al. (2020). locked EEG responses to pinprick stimulation
Trigeminal neuralgia associated with a before and after experimentally-induced
solitary pontine lesion: clinical and secondary hyperalgesia. Clin. Neurophysiol.
neuroimaging definition of a new syndrome. 128: 1445–1456.
Pain 161: 916–925. Vuong, Q.C., Allison, J.R., Finkelmeyer, A. et al.
Torta, D.M., De Laurentis, M., Eichin, K.N. et al. (2020). Brain responses in CFS and TMD to
(2020). A highly cognitive demanding autonomic challenges: an exploratory fMRI
working memory task may prevent the study. JDR Clin. Trans. Res. 5: 224–232.
development of nociceptive hypersensitivity. Wada, A., Shizukuishi, T., Kikuta, J. et al. (2017).
Pain 161: 1459–1469. Altered structural connectivity of pain-related
Tracey, I. and Bushnell, M.C. (2009). How brain network in burning mouth syndrome-
neuroimaging studies have challenged us to investigation by graph analysis of probabilistic
rethink: is chronic pain a disease? J. Pain 10: tractography. Neuroradiology 59: 525–532.
1113–1120. Wager, T.D., Atlas, L.Y., Lindquist, M.A. et al.
Tracey, I., Ploghaus, A., Gati, J.S. et al. (2002). (2013). An fMRI-based neurologic signature of
Imaging attentional modulation of pain in the physical pain. N. Engl. J. Med. 368: 1388–1397.
periaqueductal gray in humans. J. Neurosci. Wang, J., Wu, D., Shen, Y. et al. (2015a).
22: 2748–2752. Cognitive behavioral therapy eases
Treede, R.D., Rief, W., Barke, A. et al. (2015). A orthodontic pain: EEG states and functional
classification of chronic pain for ICD-11. Pain connectivity analysis. Oral Dis. 21: 572–582.
156: 1003–1007. Wang, Y., Zhang, X., Guan, Q. et al. (2015b).
Treede, R.D., Rief, W., Barke, A. et al. (2019). Altered regional homogeneity of spontaneous
Chronic pain as a symptom or a disease: the brain activity in idiopathic trigeminal
IASP classification of chronic pain for the neuralgia. Neuropsychiatr. Dis. Treat. 11:
international classification of diseases 2659–2666.
(ICD-11). Pain 160: 19–27. Wang, Y., Cao, D.Y., Remeniuk, B. et al. (2017a).
Tsai, Y.H., Yuan, R., Patel, D. et al. (2018). Altered brain structure and function
Altered structure and functional connection associated with sensory and affective
in patients with classical trigeminal neuralgia. components of classic trigeminal neuralgia.
Hum. Brain Mapp. 39: 609–621. Pain 158: 1561–1570.
Tsai, Y.H., Liang, X., Yang, J.T., and Hsu, Wang, Y., Xu, C., Zhai, L. et al. (2017b). Spatial-
L.M. (2019). Modular organization of brain temporal signature of resting-state Bold
resting state networks in patients with signals in classic trigeminal neuralgia. J. Pain
classical trigeminal neuralgia. Neuroimage Res. 10: 2741–2750.
Clin. 24: 102027. Wang, Y., Zhang, Y., Zhang, J. et al. (2018).
Vaculik, M.F., Noorani, A., Hung, P.S., and Structural and functional abnormalities of the
Hodaie, M. (2019). Selective hippocampal insular cortex in trigeminal neuralgia: a
Zempsky, W.T., Stevens, M.C., Santanelli, J.P. et al. Zhang, Y., Mao, Z., Pan, L. et al. (2019). Frequency-
(2017). Altered functional connectivity in sickle specific alterations in cortical rhythms and
cell disease exists at rest and during acute pain functional connectivity in trigeminal neuralgia.
challenge. Clin. J. Pain 33: 1060–1070. Brain Imaging Behav. 13: 1497–1509.
Zhang, J., Li, X., Jin, Z. et al. (2018a). Zhong, J., Chen, D.Q., Hung, P.S. et al. (2018).
Spontaneous brain activity and connectivity in Multivariate pattern classification of brain
female patients with temporomandibular joint white matter connectivity predicts classic
synovitis pain: a pilot functional magnetic trigeminal neuralgia. Pain 159: 2076–2087.
resonance imaging study. Oral Surg. Oral Med. Zunhammer, M., Bingel, U., Wager, T.D., and
Oral Pathol. Oral Radiol. 126: 363–374. Placebo Imaging, C. (2018). Placebo effects on
Zhang, Y., Mao, Z., Pan, L. et al. (2018b). the neurologic pain signature: a meta-analysis
Dysregulation of pain-and emotion-related of individual participant functional magnetic
networks in trigeminal neuralgia. Front. Hum. resonance Imaging data. JAMA Neurol. 75:
Neurosci. 12: 107. 1321–1330.
Part III
Dental Neuroimaging: The Role of the Brain in Oral Functions, First Edition. Chia-Shu Lin.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
Companion website: www.wiley.com/go/lin/dental-neuroimaging
the total cortical GMV at the rate of 1.89 cm3/ the thickness and surface area. For example, as
year, in the global average thickness at the rate age increases, the temporal lobe showed an
of 0.004 mm/year and in the total surface accelerating reduction, while the PFC showed
area at the rate of 3.68 cm2/year (Lemaitre a decelerating reduction (Storsve et al. 2014).
et al. 2012). There also exists a strong associa- Consistently, a longitudinal study (over a one-
tion between different morphometric features. to eight-year interval) on adults aged from 20
For example, an accelerating reduction in the to 85 years revealed different patterns of age-
volume was positively correlated with an accel- related reduction in the regional brain volume.
erating reduction in the cortical thickness. In The age-related reduction of the frontal regions
contrast, more reduction in the cortical thick- and the hippocampus fits a cubic model of
ness may be associated with less reduction in regression, while the reduction of the primary
the surface area in occipital-parietal regions somatosensory area (i.e. the precentral and
(Storsve et al. 2014). Together, the findings sug- postcentral cortices) fits a linear model of
gest that changes in brain structure at the regression (Pfefferbaum et al. 2013). These
whole-brain scale may occur as age increases. findings suggest that the age-related effect on
The changes in the overall architecture of the structural features varies between different
brain may be associated with some systemic brain regions.
factors related to aging. For example, individ-
ual differences in GMV were associated with 7.1.2.2 White Matter Morphometrics
the degree of physical activity and cardiorespi- Consistent with the changes in the cortical grey
ratory fitness (Erickson et al. 2014). matter, cross-sectional research has revealed an
Notably, the reduction of morphological fea- age-related difference in the morphometrics of
tures at the whole-brain scale does not mean a white matter, including a difference in the total
homogenous reduction in all brain regions. white matter volume (WMV) (mean ± standard
Region-wise, some brain regions show more deviation) between the younger group
age-related reduction than others (Lemaitre (20–49 years, 499.9 ± 73.7 ml) and the older group
et al. 2012; Pfefferbaum et al. 2013; Storsve (50–86 years, 445.3 ± 56.0 ml) (Ge et al. 2002).
et al. 2014; Thambisetty et al. 2010). For exam- Longitudinal research revealed an annual
ple, based on 216 healthy volunteers without a decrease in fractional anisotropy (FA) and an
history of neurological or psychiatric disor- increase in diffusivity (Sexton et al. 2014). The
ders, Lemaitre et al. (2012) found that the changes also showed a region-specific pattern,
age-related differences in the morphological with the frontal and parietal regions showing a
features are not equal between different brain greater degree of reduction (Sexton et al. 2014).
regions. The prefrontal cortex (PFC), including In middle-aged adults, a significant WM contrac-
the superior frontal gyrus (SFG), the middle tion in frontal, temporal and cerebellar regions
frontal gyrus (MFG) and the frontal pole, can be identified in a longitudinal study over four
showed the greatest reduction in volume than years (Ly et al. 2014). The FA of white matter
the other regions, and the SFG also showed the regions (including the entorhinal area, genu and
greatest reduction in the cortical thickness splenium) estimated at the baseline is associated
(Lemaitre et al. 2012). Longitudinal studies with the rate of atrophy of other tracts, including
also revealed age-related atrophy in GMV and the superior longitudinal fasciculus and anterior
cortical thinning in various regions (Taki corona radiata (Ly et al. 2014). Notably, the
et al. 2013; Thambisetty et al. 2010). A longitu- changes can also be identified in very old sub-
dinal study based on 207 healthy adults aged jects. For example, in the very old subjects aged
between 23 and 87 years revealed a substantial between 81 and 103 years, a decrease in mean FA
regional difference in the annual changes in and an increase in mean diffusivity (MD) were
identified for all tracts in longitudinal research matter (see Chapter 2), such as the correlation
with a period of 2.3 years (Lövdén et al. 2014). of the cortical thickness between different
Notably, the changes were associated with the regions across individuals, was used to esti-
decreased performance of a perceptual speed mate the inter-regional association. A recent
task, suggesting an association between altera- study based on 257 healthy adults aged
tions of white matter and age-related changes in 20–87 years reported that structural covari-
sensorimotor functions (Lövdén et al. 2014). ance of grey matter declined as age increased,
signifying a ‘disintegration’ of grey matter net-
7.1.2.3 Structural Connectivity work (Koini et al. 2018). Consistently, the
In addition to the regional changes in mor- findings revealed that in normal aging, the
phological features, aging is also associated structural changes may not happen in an iso-
with brain connectivity. Diffusion tensor lated region.
imaging (DTI), a widely used method for dif-
fusion magnetic resonance imaging (dMRI), 7.1.2.4 Cerebellum
can be used to estimate the number of poten- In addition to the cerebral cortex, structural
tial trajectories between brain regions, indexed features of the cerebellum have also been
as the strength of structural connectivity (see widely investigated in recent years. Based on
Chapter 2). An age-related difference in struc- 210 older people (mean age = 70.7 years), the
tural connectivity has been identified, based estimated size of bilateral cerebellar grey and
on 47 healthy adults aged 18–82 years white matter was approximately 94 and 31 ml,
(Zimmermann et al. 2016). The research fur- respectively (Koppelmans et al. 2015). Age-
ther identified the association between age related changes in brain features of the cere-
and the coupling between structural connec- bellum were reported in a recent longitudinal
tivity and intrinsic functional connectivity. A study based on cognitively normal adults for a
recent dMRI study based on a large sample of period of 3.1 years. At the baseline session, the
elderly people (73–76 years), with a follow-up average (±standard deviation) cerebellar vol-
of three years, revealed a decreased FA and an ume adjusted for intracranial volume is
increased MD in multiple tracts (Alloza approximately 123 (±10) cm3 (Tabatabaei-
et al. 2018), which is similar to the previous Jafari et al. 2017). The study reported an
findings (Lövdén et al. 2014). Furthermore, annual shrinkage rate of 0.36% for the cerebel-
the study investigated the change in the lar volume (Tabatabaei-Jafari et al. 2017). It
whole-brain architecture of structural connec- should be noted that the cerebellum is a func-
tivity using a graph-based method (see tionally and structurally inhomogeneous
Section 2.4). The authors found a significant structure. Recent neuroimaging findings sug-
decline in the mean edge weight and strength gest that the anterior and posterior lobes of the
of structural connectivity. However, changes cerebellum may participate in different mental
in the global efficiency and the clustering functions, collaborating with different cortical
coefficient were not statistically significant regions (Stoodley and Schmahmann 2010). A
(Alloza et al. 2018). In general, both findings longitudinal study reveals a significant positive
from cross-sectional and longitudinal research correlation between the age and the annual
suggest that aging may be associated with a rate of change in the posterior lobe of the cer-
decreased WMV and FA. Moreover, the altera- ebellum, but a negative correlation in the ante-
tion in structural connectivity may occur not rior lobe of the cerebellum (Taki et al. 2013).
only in a specific region but also in the whole- Therefore, the association between aging and
brain architecture of structural connectivity. brain features may vary between different sub-
Additionally, the structural covariance of grey regions of the cerebellum.
graph-based methods, has gradually revealed a connectivity is consistently found in the DMN.
consistent pattern of ‘decreased segregation’ of However, the age-related alterations may be less
the functional connectome in older people. In a clear-cut for other networks, such as the soma-
study based on 210 healthy adults, Chan et al. tosensory network (Sala-Llonch et al. 2015).
(2014) reported that as age increased, the
connectivity within an individual functional
7.1.4 Clinical Implications
module of the brain decreased; in contrast,
the connectivity between different systems Neurodegenerative disorders, including
increased (Chan et al. 2014). In other words, in Alzheimer’s disease (AzD) and Parkinson’s
the older group compared to the younger group, disease (PD), have become the major challenge
the whole-brain connectome showed a reduc- in geriatric healthcare, and aging is one of the
tion in the degree of segregation. Moreover, the major risk factors of the diseases (Hou
degree of decreased segregation varied between et al. 2019). An earlier neuroimaging meta-
brain systems (networks): the association sys- analysis based on 429 patients of amnestic
tems showed a pronounced decrease in segre- mild cognitive impairment (MCI) revealed
gation, particularly in the oldest age group that grey matter atrophy at the left hippocam-
(Chan et al. 2014). Consistently, the degree of pus and parahippocampus. Notably, the reduc-
modularity, which signifies the segregation of tion of GMV of these regions, as a neuroimaging
communities of nodes, was significantly lower marker, was predictive of the conversion from
in the older group compared to the younger amnestic MCI to AzD (Ferreira et al. 2011). A
group (Cao et al. 2014; Geerligs et al. 2015; Song recent study found that patients with AzD, but
et al. 2014). The decrease in modularity was not MCI, showed a larger rate of atrophy of the
associated with a decrease in local efficiency cerebellar volume compared to the cognitively
(Cao et al. 2014, Geerligs et al. 2015, Song normal participants. The findings suggest the
et al. 2014), but the global efficiency of whole- cerebellar atrophy may be associated with the
brain connectome remained similar between progression of the late phase of cognitive
the age groups (Geerligs et al. 2015). Such age- impairment (Tabatabaei-Jafari et al. 2017). The
related differences in the whole-brain func- neuroimaging markers derived from structural
tional connectome may reflect a substantial brain features also help to predict individual
association between aging and information sensorimotor performance. For example, in
transmission between brain regions (Sala- older people, cerebellar WMV was found asso-
Llonch et al. 2015). On the one hand, the ciated with hand-tapping speed and manual
decreased segregation, as demonstrated by a dexterity (Koppelmans et al. 2015). Finally, the
lower degree of modularity of weaker intra- difference in functional features, such as
network connectivity, may signify a loss of intrinsic functional connectivity, also reflects
selectivity in recruiting neural substrates (King clinical conditions. As noted previously, the
et al. 2018). On the other hand, the correspond- DMN has been widely investigated for its asso-
ing increase in inter-network connectivity may ciation with aging. Patients with AzD showed
signify the compensation of function, i.e. to decreased activity in the posterior cingulate
maintain the function by recruiting a more and hippocampus during a resting state com-
extensive network (Cox et al. 2015). In sum, the pared to healthy controls (Greicius et al. 2004).
findings suggest that as age increases, a global In general, the findings suggest that in older
change in the functional architecture of the people, brain features derived from structural
brain may occur. Moreover, the alterations in or functional neuroimaging may provide
the overall connectome do not imply a homog- important clues about the progression of neu-
enous change in all networks. Decreased rodegenerative disorders.
7.1.6 Summary
7.2.2 General Health Conditions
●● As age increased, a global reduction in the
Associated with Aging
volume and cortical thickness would be
expected. This global change does not corre- According to the World Health Organization
spond with a homogenous change in all (WHO), aging is associated with ‘a gradual
regions. In contrast, the rates of regional decrease in physical and mental capacity, a
change differ substantially between brain growing risk of disease, and ultimately, death’
regions. (https://www.who.int/news-room/fact-sheets/
●● The region-wise diversity is also significant detail/ageing-and-health). At the biological
in terms of resting-state functional connec- level, aging is the result of ‘the accumulation
tivity, which showed consistent age-related of a wide variety of molecular and cellular
damage over time’, according to the WHO course of the decline may vary between differ-
(https://www.who.int/news-room/fact-sheets/ ent mental functions. The capacity of some
detail/ageing-and-health). Therefore, as a nat- mental functions may show a life-long decline,
ural process, aging relates to both physical i.e. a decline throughout one’s lifespan. In
and mental changes of individual health other mental functions, a decline of capacity
conditions. In the following sections, we focus may occur in one’s late life (Hedden and
on the general association between aging Gabrieli 2004). The capacity for information
and health. processing, such as the speed of perceptual
processing and working memory, presents a
7.2.2.1 General Issues of Age-related Effect trend of life-long decline. In contrast, the
on Physical and Mental Conditions capacity of semantic memory presents a trend
Instead of a discussion about all the physical of late-life decline. Finally, some capacity is
and mental changes related to aging, we focus relatively stable throughout one’s lifespan,
on the two common findings that have been such as autobiographical memory (i.e. the
repeatedly reported in geriatric research. memory of personal history and self-relevant
Firstly, the age-related effect varies between information) (Hedden and Gabrieli 2004). In
different parts of our body. This is exemplified general, the capacity of mental functions may
by the physical changes of the skeletal muscle. decline as age increases, but the time course of
As age increases, muscles become ‘atrophic’ in the declination may vary between different
general, such as a decreased cross-sectional mental functions.
area of muscles. The morphological change
may be associated with the reduction in the 7.2.2.2 Relationship Between Aging
number of muscle fibres or/and the atrophy of and Diseases
the fibre per se (McComas 1998). Notably, the Aging is the major risk factor for many disor-
aging-related effects on muscle tissues differ ders, including cardiovascular and neurode-
between the jaw and the limb muscles. As age generative diseases (McHugh and Gil 2018).
increases, the masseter showed a decreased Traditionally, people tend to link aging with
proportion of the type I fibres and an increased ‘losses’ in physical or mental capabilities. This
proportion of the type II fibres, while the belief is further reinforced by research findings
biceps showed no changes in the type I fibre from a comparison between different age
but a decreased proportion of the type IIB groups. The comparison usually reveals a
fibres (Monemi et al. 1998). In jaw muscles, decreased performance in physical or mental
the nerve terminals in muscle fibres maintain tasks in older subjects compared to younger
a higher capacity of regeneration compared to subjects (Rowe and Kahn 1987). However, the
the limb muscles (Avivi-Arber and Sessle 2018). results usually ignore the heterogeneity within
Clinically, jaw muscles are less susceptible an age group. Even though an older group
to certain neuromuscular diseases and age- shows an averagely worse performance com-
related atrophy compared to limb muscles pared to a younger group, some older individu-
(Avivi-Arber and Sessle 2018). The diversity of als may perform better than others. The
age-related effects between the jaw and the variability of performance within older people
limb muscles may underlie the different clini- can be interpreted by multiple factors, e.g.
cal symptoms and signs between the regions. individual habits or psychosocial factors,
The second common finding from the litera- which are extrinsic variables that modify the
ture is about the time course of the age-related effect of aging (Rowe and Kahn 1987). In suc-
effect. It is widely accepted that as age cessful aging, these extrinsic factors may play a
increases, the capacity of mental functions neutral or positive role that renders older peo-
may decline to some degree. However, the time ple with little or no loss of physiological
functions (Rowe and Kahn 1987). Based on the sharply during the younger stage (20–40
nation-representative longitudinal data of years old), not in the elderly stage of one’s
Taiwan, researchers categorized the older lifespan (Kassebaum et al. 2014a). In addi-
cohorts into four types, successful aging, usual tion to oral diseases, fear and anxiety may
aging, health declining and care demanding, mediate the association between aging and
according to six indicators: chronic diseases, tooth loss (Friedman and Lamster 2016;
physical function difficulties, depressive symp- Kossioni and Dontas 2007). According to a
toms, social support, social participation and nation-wide survey based on Finnish adults,
economic satisfaction (Hsu and Jones 2012). a strong fear of dental visiting is prevalent in
Based on the selected indicators, the cohort all age groups; even in the elderly group
with successful aging showed fewer problems (more than 75 years old), there was a preva-
in chronic diseases, physical function difficul- lence of 2.2% of being very much afraid of
ties and depressive symptoms compared to the visiting dentists (Liinavuori et al. 2016). In
health-declining and care-demanding cohorts general, both pathological and psychosocial
(Hsu and Jones 2012). Notably, different factors may potentiate the rate of tooth loss
cohorts showed different trajectories of aging as age increases.
(i.e. the progression of the age-related variable
over time), supporting the validity of assessing 7.2.3.2 Orofacial Muscles
aging using multiple indicators. As noted in the preceding section, the age-
related effect of skeletal muscles may vary
between the jaw and limb muscles. The jaw
7.2.3 The Age-related Differences in the
muscles may be innervated by more motor
Stomatognathic System
units, compared to the limb muscles
In the following section, we focus on the senso- (McComas 1998). At the histochemical level,
rimotor aspects regarding feeding behaviour, the jaw muscles are composed of type I fibres
including teeth and dentition, orofacial mus- as well as type II fibres. Therefore, the jaw mus-
cles and salivary secretion. We also focus on cles have a longer contraction time, and they
the differences between different age groups are less susceptible to fatigue (McComas 1998).
based on cross-sectional studies. Therefore, these muscles can perform the
chewing movement for a long period. While
7.2.3.1 Tooth Loss animal research focuses on histological con-
Tooth loss is commonly found in older peo- figurations of orofacial muscles, imaging
ple. Meta-analytic findings reveal that the methods make it feasible to assess the quality
prevalence of severe tooth loss (i.e. having of jaw muscles. Recent studies using ultra-
fewer than nine remaining permanent teeth) sound sonography revealed that the echo
increased gradually with age (Kassebaum intensity of masticatory muscles was not only
et al. 2014b). However, it should be noted associated with the displacement of the mus-
that the loss of natural teeth cannot be fully cles (Hara et al. 2020) but also systemic factors
attributed to aging since tooth loss is mark- of aging, including grip force and walking
edly associated with dental diseases, includ- speed (Yamaguchi et al. 2019). Maximal tongue
ing periodontitis and severe dental caries, pressure (MTP), another critical physiological
and the prevalence of periodontitis and den- factor associated with mastication and swal-
tal caries is high among older people (Lopez lowing, also declines as age increases (Liu
et al. 2017). As age increases, the multirooted et al. 2019; Ohno et al. 2020). Notably, in older
teeth are at a greater risk of periodontitis and subjects, a higher MTP was not associated with
tooth loss (Lamster et al. 2016). Interestingly, better echo intensity, an indicator of muscular
the incidence of severe periodontitis peaks quality, of the tongue muscle (Chantaramanee
et al. 2019). In contrast, oral diadochokinesis 7.2.4 Age-related Changes in Oral
was associated with the quality of the tongue Sensory Functions
muscle (Chantaramanee et al. 2019). The find-
In the preceding sections, we focused on the
ings suggest a complex relationship between
age-related effect on the physiological founda-
strength, structure and skill-related tongue
tions of the stomatognathic system, including
movement.
jaw muscles and saliva secretion. In recent
years, there has been an increased number of
7.2.3.3 Salivary Secretion
studies focusing on the age-related effect on
Both hyposalivation (i.e. the reduction of
oral functions. As summarized in Table 7.1, in
salivary flow rate, SFR) and xerostomia (i.e.
the following sections, we focus on the associa-
subjective perception of dry mouth) are com-
tion between aging and changes in the thresh-
monly seen in older dental patients (Kossioni
old or the intensity of oral sensorimotor
and Dontas 2007; Lamster et al. 2016).
functions.
However, the association between aging and
salivary secretion may be a more complicated
issue because reduced saliva secretion, which 7.2.4.1 Somatosensory Functions
can be quantified by SFR, is associated with Earlier studies revealed a mixed result of the
both intrinsic factors (e.g. changes in salivary association between aging and oral somatosen-
glands) and extrinsic factors (e.g. polyphar- sation. For example, by investigating 60 healthy
macy) (Villa et al. 2015). An early longitudi- adults of a broad age range (23–96 years),
nal study over three years on subjects with Calhoun et al. (1992) found no significant age-
a broad age range (26–90 years) revealed related differences in thermal sensitivity, som-
no significant difference in stimulated or esthetic sensitivity, proprioception, vibration
unstimulated SFR of major glands over a detection (on the soft palate) and two-point
three-year period (Ship et al. 1995). Findings discrimination (on the tongue and palate). In
from cross-sectional research revealed a sig- contrast, a declined ability of tactile, vibratory
nificantly lower unstimulated SFR in older and two-point discrimination was noted on the
groups (age > 40 years) compared to the lip (Calhoun et al. 1992). In general, oral sen-
youngest group (age ≦ 39 years). In contrast, sation remained good as age increases, with a
the stimulated SFR of parotid saliva was not slight decline after the age of 80 years (Calhoun
significantly different between the age groups et al. 1992). Another study based on 178
(Percival et al. 1994). In another cross- healthy adults aged 20–89 years revealed an
sectional study with a larger sample size elevation in sensory thresholds, including
(N = 540) of healthy adults, the researchers warming and cooling temperatures, pain,
found that the older group (≧70 years) touch, two-point discrimination and taste, for
showed a significantly lower stimulated SFR the older subjects (≧65 years) compared to the
compared to the other age groups (Smith younger subjects (<45 years) (Heft and
et al. 2013). Meta-analytic findings showed Robinson 2010). A main effect of sex was
that of the stimulated and unstimulated SFR noted, but an interaction between sex and age
of the submandibular/sublingual salivary was insignificant, except for the sensation of
glands were significantly lower in older sour (i.e. older men having the highest sour
adults compared to younger adults (Affoo threshold). Such an age-related difference
et al. 2015). The findings generally reveal a was also found for lingual tactile thresholds,
reduction in SFR in elderly people. However, which were higher in subjects aged >40 years
the association between the reduction and (Bangcuyo and Simons 2017). In general,
the intrinsic/extrinsic factors requires more the findings suggest that changes in oral
investigation. somatosensory functions, including elevation
Table 7.1 Recent findings on age-related changes in oral sensorimotor functions (since 2015).
Notes: BMS: burning mouth syndrome; GF: grip force; MBF: maximum biting force; MP: masticatory performance;
MTP: maximum tongue pressure.
of thresholds and declined abilities of discrim- et al. 2017), though the conclusions are limited
ination, may occur selectively in some regions by a smaller sample size. Notably, in older
(e.g. the lip and the chin). The diversity of adults, there was no significant difference in
results may be attributed to the variations of the pain detection threshold or the tactile
testing methodology (Heft and Robinson 2017). detection threshold between patients with
Furthermore, recent studies have presented a burning mouth syndrome (BMS) and healthy
potential link between oral somatosensory controls (Honda et al. 2019). The findings
functions and hearing threshold (Etter et al. reveal great individual variations in oral soma-
2016) as well as systemic factors (Silva tosensory functions in older adults.
not statistically significant within the older attributed to the alterations from multiple
group (Lin et al. 2017a). Notably, in the study, aspects, including dentition, masticatory mus-
the MP was quantified by assessing the oral cles, the sensory feedback of periodontal tis-
mixing ability (see Section 3.1) but not the cut- sues and salivation.
ting ability. It is possible that other factors,
such as individual variability in oral sensory 7.2.5.3 The Role of Motor Control
functions, may contribute to the association and Behavioural Adaptation
between the MMV and the MP in older During mastication, the tongue plays a key
subjects. One of the key factors for sensory role in mixing and transporting food bolus in
feedback during mastication is the mechano- the oral cavity. In a recent study, Sagawa
receptors in the periodontal ligament et al. (2019) investigated the cutting ability of
(Trulsson 2006). Because periodontal diseases 245 healthy older adults with 28 natural
are common in older people, the detrimental teeth. By controlling the number of existing
effect on altered mechanoperception from per- teeth, they found that lower MP was associ-
iodontal diseases should be noted in older peo- ated with a decreased speed of tongue move-
ple. Clinical research revealed that patients ment as well as tongue muscle force (Sagawa
with reduced periodontal support in their et al. 2019). The findings suggest that in
anterior teeth would apply a greater force to addition to tooth loss, a declined motor skill
hold and split a peanut held teeth (Johansson of the tongue may contribute to a lower
et al. 2006). An increased alveolar bone-height- MP. In addition, Ohno et al. (2020) investi-
to-tooth-length (AB/T) ratio is widely adopted gated 152 adults ranging from 20 to 79 years
as an index of periodontal health. The subjects old. They found that both decreased MP (i.e.
with a lower AB/T ratio (<50%), which may the cutting ability) and the threshold of swal-
indicate moderate-to-severe generalized perio- lowing (i.e. the chewing cycles before swal-
dontitis, showed a lower MP compared to lowing a jelly) were associated with aging
those with a higher AB/T ratio (De Freitas (Ohno et al. 2020). Moreover, decreased MP
Borges et al. 2013). A large-scale survey of was also associated with decreased tongue
1875 Japanese elderly adults also revealed that pressure and decreased grip force, consistent
periodontitis is a risk factor of a lower MP with previous findings (Ohno et al. 2020;
(Kosaka et al. 2016). Our masticatory function Sagawa et al. 2019).
is also associated with salivation. The associa- The cause–effect relationship between the
tion between aging and salivation has remained change of motor skill and masticatory function
unclear. Though hyposalivation and xerosto- is complicated. It is reasonable to conceive that
mia are common in older people, these symp- those who have a better ability of motor con-
toms are largely associated with polypharmacy. trol (on their tongue, for example) will show a
Notably, saliva secretion of the parotid gland, better MP. However, it is also possible that they
as the major salivary gland, is stimulated by develop a better skill of tongue movement to
mastication. Consistently, an increased MP compensate for the declined masticatory func-
was associated with an increased stimulated tion, which may derive from tooth loss. In
SFR (when subjects were chewing a gum) other words, older people may develop some
rather than unstimulated SFR (when subjects strategies for adapting their oral conditions to
were resting) (Lin et al. 2017a). Older people maintain masticatory functions (see Chapter 8
may increase the time for chewing to facilitate for a detailed discussion). For example, due to
salivary output, which in turn facilitates the a declined MP, older people may fail to reduce
formation of a bolus to swallow (Peyron food into smaller pieces before swallowing it.
et al. 2017). In general, the findings suggest However, as shown in Section 3.1, for a specific
that age-related differences in MP may be type of food, there exists small inter-individual
variability of the particle sizes that are ready to reduction in the tongue pressure (Cullins and
swallow (Prinz and Lucas 1995; Woda Connor 2017). Together, the findings highlight
et al. 2006). For older people, to achieve a suit- that physiological features of the tongue play a
able size of bolus to swallow, they would key role in swallowing performance.
increase the number of masticatory cycles As noted in Section 3.1, the repeated saliva
before swallowing it (Peyron et al. 2017). Such swallowing test (RSST) has been adopted as a
a behavioural adaptation may be related to non-invasive assessment of swallowing perfor-
both age and food factors. For example, the mance. A significant difference in the RSST
number of pre-swallow cycles of chewing soft score was identified between younger and older
elastic food is positively correlated with age groups (Lin et al. 2019; Oguchi et al. 2000a, b).
(Peyron et al. 2017; Woda et al. 2006). In older Because older people also show a lower SFR
people with a complete denture (CD), the num- compared to younger people, the age-related
ber of masticatory cycles increased as the hard- difference in swallowing performance may be
ness of food increased (Veyrune et al. 2007). attributed to an age-related difference in
Notably, the pre-swallow number of mastica- SFR. The hypothesis was, however, not sup-
tory cycles should not be confused with masti- ported by research evidence. In their original
catory frequency, i.e. ‘number of cycles per research of the RSST, Oguchi et al. (2000b) did
second with the sequence’ (of mastication) not find a significant difference in the RSST
(Peyron et al. 2017). Masticatory frequency scores between the condition of dry swallow and
generally keeps constant as age increases (Lin the condition when artificial saliva was supplied
et al. 2019; Peyron et al. 2004). Therefore, older in both younger and older subjects. In older sub-
people may adapt to tooth loss by increasing jects, the RSST score was not significantly cor-
the number of masticatory cycles for process- related with either stimulated or unstimulated
ing food (Peyron et al. 2017; Woda et al. 2006). SFR (Lin et al. 2019). Consistently, multiple
regression analyses revealed that both self-
assessed xerostomia and objective saliva secre-
7.2.6 Swallowing Functions
tion were not associated with the RSST score
Tongue movement plays a key role in swallow- when gender, age and the number of prescribed
ing as well as in mastication. In general, the medicines were controlled (Persson et al. 2019).
duration of swallowing may prolong in older How the performance of the swallowing test
people, partly because of the slowdown of the relates to individul differences in the stomatog-
major movement (e.g. tongue thrust) during nathic system would require further research.
swallowing (Peyron et al. 2017). A recent study
of 197 older subjects revealed that decreased
7.2.7 Summary
tongue pressure was associated with increased
age (in male subjects) and the occurrence of ●● The age-related effects vary between differ-
sarcopenia (in both female and male subjects). ent parts of our body. The capacity of mental
In contrast, the jaw-opening force was only functions may decline as age increases, but
associated with the occurrence of sarcopenia the time course of the declination may vary
in male subjects (Machida et al. 2017). In between different mental functions.
healthy older subjects, the tongue pressure was ●● Aging does not necessarily mean a disease
associated with the tongue muscle thickness condition of losses in physical or mental
(Yoshimi et al. 2018). The findings echoed the conditions. The practical goal of healthcare
results from animal research, which revealed for geriatric people is to promote healthy
that aging was associated with a decreased aging, i.e. to help older people to maintain
number and size of muscle fibre in the functional ability and well-being, according
tongue muscles, which may contribute to the to the WHO.
●● Both pathological factors (e.g. increased rates can be attributed to peripheral factors, such
of caries and periodontitis) and psychosocial as an increasing number of missing teeth and
factors (e.g. dental fear and anxiety) may worse periodontal status. However, the cen-
potentiate the rate of tooth loss as age increases. tral nervous system also plays a key role in
●● Changes in somatosensory functions, includ- the age-related effect. For example, jaw move-
ing elevation of thresholds and declined abili- ment is associated with mechanisms of
ties of discrimination, may occur selectively rhythmic movement and reflexes, which are
in some regions (e.g. the lip and the chin). dominated by the brainstem (Lund 1991).
The diversity of results may reflect the varia- The brain’s role in the link between aging and
tions of testing methodology. mastication is often overlooked by dentists.
●● Age-related differences in MP may be attrib- The next section outlines the neuroimaging
uted to the alterations from multiple aspects, findings of age-related brain mechanisms of
including dentition, masticatory muscles mastication.
and periodontal conditions.
●● In older people, behavioural modification, 7.3.2.1 Aging and the Brain Mechanisms
such as increasing duration of mastication, of Mastication
may play a key role in adaptively maintain- With the advent of neuroimaging methods,
ing masticatory function. researchers have started to investigate the role
of aging in the brain mechanisms of mastica-
tion. In a pioneering study of this topic,
7.3 Association Between Onozuka et al. (2003) compared the brain acti-
the Brain and Oral Functions vation during gum chewing between younger,
in Older People middle-aged and older (aged 65–73 years)
healthy subjects. The study provides three crit-
7.3.1 Introduction ical findings that have been identified in the
As shown in the preceding chapters, aging is following studies. Firstly, mastication was
associated with a substantial structural and associated with the activation of several brain
functional change in both the stomatognathic regions, including the primary motor cortex
system and the brain. Notably, such an age- (M1) and the primary somatosensory cortex
related effect does not necessarily mean a dis- (S1). The pattern of activation was consistently
ease condition – the age-related changes may identified in all age groups. In other words, the
occur under healthy aging when individuals cortical processing of sensory and motor infor-
are free of age-related diseases (e.g. neurode- mation of mastication is highly associated with
generative disorders). In the following sec- mastication, as identified in a recent meta-
tions, we review more evidence between the analysis (Lin 2018a). The activation of sensori-
brain and oral functions in elderly people, motor regions may represent a core mechanism
focusing on mastication and swallowing. We of mastication, which is common to all age
highlight the brain–stomatognathic associa- groups. Secondly, mastication was associated
tion primarily for people with healthy aging, with the activation of more widely distributed
while more issues of patients with age-related brain regions, not limited to the S1/M1. Such
diseases are discussed in the next chapter. an extensive activation was identified in the
PMC/supplementary motor area (SMA), the
PFC, the cerebellum, the thalamus and the
7.3.2 Association Between Mastication
insula (Onozuka et al. 2003). Consistently, a
and the Brain in Older People
recent fMRI study revealed that mastication
As discussed in the preceding sections, the was associated with only the S1/M1 in younger
age-related decline of masticatory function dentulous subjects, but extensive regions
(including in the PMC/SMA, the PFC, the volume and whole-brain GMV in subjects with
cerebellum, the thalamus and the insula) in complete or partial tooth loss compared to the
older dentulous subjects (aged 80–83 years) group without tooth loss (adjusted for total
(Kobayashi et al. 2020). The findings suggest intracranial volume and age). Kobayashi et al.
that older individuals compared to younger (2018) found a smaller GMV in multiple brain
individuals may engage with a more compli- regions, including the hippocampus, caudate
cated pattern of information processing dur- nucleus and temporal pole, in edentulous sub-
ing mastication. Thirdly, Onozuka et al. (2003) jects compared to dentulous older subjects. In
found that during mastication older subjects cognitively healthy older adults, their MP (i.e.
showed lower activation in the S1/M1 com- the mixing ability) was positively correlated
pared to younger subjects. Notably, they with the regional GMV of the M1, the PMC
showed higher activation predominantly in and the dorsolateral PFC (Lin et al. 2015). In
the PFC compared to younger subjects addition to grey matter, alterations in white
(Onozuka et al. 2003). In another fMRI study, matter may be associated with mastication. In
Fang et al. (2005) compared the brain activa- patients with stroke, the amount of white mat-
tion between different age groups for hand ter hyperintense (WMH) lesions is associated
and oral movement (including jaw opening/ with not only swallowing difficulty but also
closing and chewing). They also found inadequate mastication (Moon et al. 2017).
increased activation of the PFC during the The association between the presence of
chewing task. Notably, such a PFC activation moderate-to-severe WMH and edentulism was
was not found for hand movement (Fang found, though both factors covaried with age
et al. 2005). The PFC activation may highlight (Del Brutto et al. 2016). In addition to the find-
the role of cognitive processing in mastication, ings of structural features, findings from rs-
such as adaptation to a denture (Kamiya fMRI revealed that age may play a key role in
et al. 2016; Kimoto et al. 2011; Narita the pattern of intrinsic functional connectivity
et al. 2019). In general, the studies consistently related to mastication. For example, in cogni-
revealed both qualitative and quantitative tively healthy older adults, an increased MP
differences in the brain mechanisms of masti- was associated with a stronger connection
cation between younger and older adults. between the cerebellum and the PMC/dorso-
From the qualitative aspect, mastication was lateral PFC (Lin et al. 2015). Comparing to
engaged with activation in more extensive younger subjects, older subjects showed a dis-
brain regions in older adults. From the quanti- tinct functional network associated with mas-
tative aspect, older adults showed both lower tication. In older subjects, an increased MP is
and higher activation in multiple brain regions further associated with an increased connec-
compared to younger adults. tion between the sensorimotor regions and the
anterior insula, the cingulate cortex and the
7.3.2.2 Recent Neuroimaging Findings parietal lobe (Lin et al. 2017b). The findings
of Aging and Mastication echo the previous findings from task-based
Table 7.2 summarizes recent neuroimaging fMRI studies (Kobayashi et al. 2020; Onozuka
findings of the brain mechanisms associated et al. 2003), revealing a more extensive area
with mastication in older adults. Several stud- associated with mastication in older adults.
ies have been published for the association Several studies have focused on the associa-
between structural brain features and the indi- tion between oral rehabilitation (i.e. wearing a
vidual variability in masticatory function. denture) and changes in brain activity. Earlier
Based on a random sample of 394 dementia- studies reported a reduced activation of the
free older adults (aged ≧60 years) from Sweden, PFC in the condition of wearing an implant-
Dintica et al. (2018) found a smaller total brain supported denture (IOD) compared to the
Table 7.2 Neuroimaging research on brain mechanisms of aging and oral functions (since 2015).
Notes: EEG: electroencephalography; EMG: electromyography; fMRI: functional magnetic resonance imaging;
fNIRS: functional near-infrared spectroscopy; OD: oropharyngeal dysphagia; PET: positron emission tomography;
rs-fMRI: resting-state functional magnetic resonance imaging; rTMS: repetitive transcranial magnetic stimulation;
SCR: skin conductance response; sMRI: structural magnetic resonance imaging; TRP: transient receptor potential;
VBM: voxel-based morphometry.
to younger subjects. The laterality of brain acti- compensatory mechanisms of the brain
vation has been further investigated by (Malandraki et al. 2010), and such an ‘overacti-
Malandraki et al. (2010), who performed a vation’ may reflect an increased task demand
region-of-interest (ROI) analysis on the brain in older adults (Humbert et al. 2009). In gen-
activation of S1, the M1 and the PMC, associ- eral, the earlier findings revealed that swal-
ated with water swallowing. The authors found lowing, like mastication, was associated with
a greater activation at the right hemisphere in both qualitative and quantitative differences
younger subjects during swallowing. In con- between age groups.
trast, in older subjects (mean age 70.2 years),
there was no significant difference between 7.3.3.2 Recent Neuroimaging Findings
hemispheres (Malandraki et al. 2010). of Aging and Swallowing
Furthermore, the activation of the PMC was The neuroimaging findings published more
identified during the preparatory stage of swal- recently are summarized in Table 7.2. Using
lowing, which may be associated with motor fMRI, Windel et al. (2015) investigated the
planning of the movement (Malandraki brain activation associated with water swal-
et al. 2010). A subsequent ROI analysis reveals lowing in older adults. They investigated the
a generally decreased activation in the S1 and brain activation associated with the movement
the M1 during water swallowing in older sub- of the thyroid cartilage, a critical feature of
jects compared to younger subjects (Malandraki swallowing movement. Moreover, subjects’
et al. 2011). In contrast, during planning (i.e. skin conductance response (SCR), as an index
anticipating to swallow), decreased activation of emotional arousal during swallowing, was
was found only in part of the S1 (Malandraki recorded (Windel et al. 2015). In older sub-
et al. 2011). Together, the findings suggest that jects, there was higher activation in the PFC
different brain mechanisms in motor control (BA10) compared to younger subjects. Notably,
may underlie the age-related effect on older subjects also showed a higher SCR com-
swallowing. pared to younger subjects, which may reflect a
The results from the earlier studies, though higher level of autonomic arousal and cogni-
not revealing a consistent finding of the age- tive efforts during swallowing. In older sub-
related brain mechanisms of swallowing, sug- jects, a higher SCR was associated with higher
gest several important clues about aging and activation in the inferior frontal cortex (i.e.
swallowing. First compared to younger adults, BA45), the cingulate cortex, the insula and the
older adults did not show an overall ‘reduction’ hippocampus, as well as the S1/M1 area. The
of brain activation in all areas associated with strong association between the SCR and the
swallowing. They have distinct patterns of acti- extensive brain regions outside the S1 and the
vation with some regions (particularly the M1 may reflect increased recruitment of atten-
PFC) showing a greater activation than tional resources and an increased emotional
younger subjects. Secondly, in older subjects, valence of swallowing (Windel et al. 2015).
multiple regions (other than the S1/M1) would Using electroencephalography (EEG), Rofes
participate in swallowing, such as the PFC, the et al. (2017) investigated the event-related
parietal, the temporal and the cerebellar potential (ERP) associated with the electrical
regions. The behavioural meaning of such a stimulation at an intra-pharyngeal site. They
sophisticated pattern of activation remains found decreased activity in the parietal area
unclear. A potential explanation is that for (including the S1 and the M1) and frontal area
older adults more efforts (e.g. attention and in older (non-dysphagic) subjects compared to
cognitive control) are demanded to maintain younger subjects. In contrast, older patients
swallowing. Older adults would show a more with dysphagia show decreased activity in the
symmetrical activation of brain regions as parietal area but increased activity in the
frontal area compared to older non-dysphagia pharyngeal ERP (Tomsen et al. 2020). These
subjects (Rofes et al. 2017). In terms of the clinical findings, being limited in a smaller
structural features of swallowing, Lin et al. sample size and the heterogeneity of clinical
(2019) found that in cognitively healthy older conditions, highlight the role of the brain in
adults, the GMV and regional volume of the the rehabilitation of swallowing functions.
posterior cerebellum were positively correlated
with swallowing performance, as quantified by
7.3.4 Considerations of Research Design
the repetitive saliva swallowing test (RSST).
and Future Directions
Notably, the posterior cerebellum is associated
with not only motor control but also cognitive The current neuroimaging evidence, though
processing of movement (Stoodley and providing several important clues about the
Schmahmann 2010). The findings strengthen age-related brain mechanisms of oral func-
the connection between swallowing and the tions, has not formed a full-fledging theoretical
cognitive–affective processing in older adults. framework for explaining the association
There have been more studies that investigate between aging and oral functions. From a clin-
the interventional effect on improving swal- ical perspective, such a framework should be
lowing compared to mastication (Table 7.2). able to (i) form a better prediction of individual
For example, using positron emission tomog- differences in oral functions, (ii) help evaluat-
raphy (PET), Park et al. (2017) identified a cor- ing the oral functions of older people with neu-
tical activation of the bilateral sensorimotor rological deficits and (iii) help evaluating the
cortex during swallowing in non-dysphagic treatment effect on oral dysfunctions (e.g. dys-
older adults. For older patients with dysphagia, phagia). The current neuroimaging evidence
they found that repetitive transcranial mag- has identified several brain regions associated
netic stimulation (rTMS) for two weeks would with aging and oral functions, but the results
improve patients’ pharyngeal motor function have not been fully translated into clinical
and showed increased activation at the bilat- application. In order to bridge this gap, two
eral M1, PMC and the right PFC (Park general issues for future research are discussed
et al. 2019). In contrast, the effect of rTMS was in the following sections.
not significant in non-dysphagic older adults
(Park et al. 2017). Another strategy for improv- 7.3.4.1 The Definition and Selection of the
ing swallowing functions is to improve the sen- Study Group
sory processing of the pharynx, which is key to One of the major limitations of the current
the reflexive movement during the pharyngeal studies is the lack of subjects from a broader
phase of swallowing. Tomsen et al. (2019) age group. As shown in Table 7.2, most of the
manipulated the sensory processing using oral studies on mastication focused on patients
capsaicin treatment. They found that the aged from 55 to 70 years. Few studies focused
patients who received capsaicin treatment on an even elder group (e.g. over 85 years old)
showed an improved swallowing function and comparison between the ‘young–old’ vs.
compared to the placebo group. The improve- ‘old–old’ subgroup is still uncommon. It should
ment was associated with changes in the be noted that not all the oral functions ‘decline’
latency and amplitude of pharyngeal ERP with the same time course. For example, the
(Tomsen et al. 2019). Furthermore, the MP may not drop dramatically until 70–80 years
researchers found that in patients with dys- old (Lin et al. 2017a), while severe periodonti-
phagia, the stimulation of TRPA1 (an ion tis, a major cause of tooth loss, may start even
channel associated with airway irritation) in the middle age (Kassebaum et al. 2014b).
would improve swallowing function, which While most of the studies focused on a com-
was also associated with changes in parison between an ‘older’ (e.g. aged > 65 years)
characteristic proteins, which form insoluble reports of the topic, i.e. the 2007 research
aggregates and affect neurons and glial cells derived from the famous ‘Nun study’, a cohort
(Brettschneider et al. 2015). Both AzD and PD study for clinical risk factors of AzD. In this
are well known for their high prevalence world- study, Stein et al. (2007) analyzed 144 subjects
wide (Scheltens et al. 2016). Patients with neu- for their medical and dental records over
rodegenerative disorders would gradually lose approximately 12 years. They found that ‘a low
their abilities of self-care in daily life. For exam- number of teeth increased the risk of higher
ple, patients with moderate-to-severe AzD may prevalence and incidence of dementia’ (Stein
have difficulty to independently carry out daily et al. 2007). As usually cited in the following
activities, such as eating and cleaning. The studies (e.g. Cerutti-Kopplin et al. 2016), the
increased demand for caring leads to a great occurrence of new dementia cases (i.e. the inci-
economic burden on the medical system and dence) was higher in the group with fewer
society (Dorsey et al. 2013). One of the key (0–9) non-third molar teeth compared to the
issues of dental patients with neurological dis- group with more (≧10) non-third molar teeth
orders is their declined oral health (Cerutti- (Stein et al. 2007). However, in the original
Kopplin et al. 2016; Tada and Miura 2017; study by Stein et al. (2007), several key findings
Tonsekar et al. 2017; Wu et al. 2016). The sec- were less cited and largely neglected. Firstly, at
tion focuses on cognitive impairment and the initial stage of the longitudinal investiga-
AzD. The association between oral health and tion (i.e. the time when subjects received the
cognitive impairment, and the potential brain first cognitive examination), the association
mechanisms underlying the association, is out- between the number of tooth loss and demen-
lined in the following sections. tia was not significant after adjusting for
systemic factors. Secondly, they found an
association between the tooth–dementia asso-
7.4.2 The Oral–Cognitive Association
ciation and the polymorphism in the apolipo-
The association between neurodegenerative protein E4 allele (APOE4) allele: in the subjects
disorders and declined oral health has been without APOE4, a lower number of teeth were
reported in various clinical and epidemiologi- associated with dementia prevalence (Stein
cal studies, which have been summarized in et al. 2007). These critical findings suggest that
previous reviews (Cerutti-Kopplin et al. 2016; dementia was not only associated with tooth
Tada and Miura 2017; Tonsekar et al. 2017; Wu loss. There may exist an interaction between
et al. 2016). While some researchers confirmed oral factors and individual factors (i.e. the
the association between worse oral factors and APOE4 genotype). Cumulating evidence has
worse cognitive conditions (Cerutti-Kopplin suggested that AzD is associated with multiple
et al. 2016; Tada and Miura 2017), others con- factors that are interwoven in a complex rela-
sidered it inconclusive based on current evi- tionship (Scheltens et al. 2016). For example,
dence, which is mostly cross-sectional and another earlier longitudinal study over two to
correlational (Tonsekar et al. 2017; Wu three years revealed that changes in SFR, but
et al. 2016). In the following section, we briefly not the number of missing/decayed teeth, sig-
outline the status quo of the research and high- nificantly decreased in patients with AzD com-
light the reason why neuroimaging plays a key pared to healthy controls (Ship and
role in clarifying the oral–cognitive association. Puckett 1994). In the five-year longitudinal
study based on 1566 non-demented older sub-
7.4.2.1 Clinical Findings of Oral– jects (aged ≧60 years) in the Hisayama area,
Cognitive Associations Japan, the association between the number of
When it comes to the oral–cognition associa- remained teeth and the risk of AzD was only
tion, it is worthy of reviewing one of the earlier borderline (p = 0.08) (Takeuchi et al. 2017).
Another recent study from the Netherlands on animal research should not be over-interpreted
114 older subjects with dementia has revealed as ‘a decline of cognitive functions (of all
that their MP was negatively correlated with aspects)’ since spatial memory is part of the
the score of general cognition. However, the cognitive functions (Lin 2018b). A precise defi-
association was not established when the factor nition of the association also helps to avoid a
of independence of daily activities was included premature claim of the cause–effect relation-
(Weijenberg et al. 2015). Finally, a recent study ship between the variables. The term ‘declined
on 537 older subjects (aged ≧65 years) revealed oral health’ may represent multiple deficits or
that those who had at least one APOE4 allele suboptimal oral conditions. For example, a
and fewer than nine teeth would show an recent study in Spain reported that patients
increased risk of mild memory impairment with AzD showed poor periodontal health, a
(MMI) compared to those who had neither fac- lower SFR and a higher incidence of candida
tor. Notably, the risk did not show a significant infection compared to non-AzD controls
increase in the subjects with either one factor (Aragón et al. 2018). A longitudinal study in
alone (Okamoto et al. 2017). These findings did Japan with a larger sample size (N = 1566)
not totally falsify or support the oral–cognitive revealed that tooth loss is a risk factor for all-
hypothesis as recently debated. However, the cause dementia and AzD. However, the pro-
evidence suggests that tooth loss is not the only portion of tooth loss also covaried with other
(or even the major) factor contributing to cog- oral factors, such as the frequency of tooth
nitive impairment. Instead, oral factors can be brushing and attending dental care, as multi-
one of the multiple factors that are associated ple systemic factors (Takeuchi et al. 2017).
with cognitive impairment. AzD would be Therefore, it should be noted that there may
associated with the interaction between the exist multiple versions of the oral–cognitive
oral factors and other systemic factors rather association, from a narrower but clearly
than the oral factors per se. defined version (e.g. the association between
tooth loss and deficits in spatial memory) to a
7.4.2.2 Revisiting Oral–Cognitive Associations broader but vaguely defined one (e.g. the asso-
What needs to be clarified first is the question ciation between oral health and mental
itself: what are the objects to be studied in health). From the point of research design, the
terms of the oral–cognition association? While different aspects of the oral–cognitive associa-
some research focuses on the association tions should be investigated with different
between very specific conditions, such as the research approaches.
association between ‘tooth loss’ and ‘dementia’
(Thomson and Barak 2020), other research
7.4.3 The Role of the Brain in the
focuses on the association from a broader
Oral–Cognitive Associations
scope, such as the association between ‘oral
health’ and ‘cognitive status’ (e.g. Wu From the perspective of the brain–stomatog-
et al. 2016). The different associations (in plu- nathic axis (BSA), the oral–cognitive associa-
ral form) should be defined respectively. For tions can be examined from three aspects,
example, if ‘cognitive status’ is the condition which respectively correspond to some
one hopes to investigate, we may need to define testable hypotheses proposed previously
what mental functions are included in the cog- (Lin 2018b). As mentioned in Chapter 1,
nitive status. Some studies (predominantly under the framework of the BSA, the
animal research) may focus on the function of brain, the stomatognathic system and feed-
spatial memory and learning (Fukushima- ing behaviour are bidirectionally linked
Nakayama et al. 2017). A conclusion that (Figure 7.1a). The framework is based on the
‘tooth loss may impair spatial memory’ from oral-brain-behaviour (OBB) model, which
highlights both the CNS modulation on oral 7.4.3.1 An Emphasis on the Brain’s Role
functions and the sensory feedback from the in Behaviour Adaptation
stomatognathic system, which together main- Firstly, poor oral health may be attributed to
tains feeding behaviour. Moreover, the BSA more difficulty in conducting health-related
framework highlights the role of our brain in behaviour (e.g. being unable to brush teeth)
cognitive, affective and motivational process- (Figure 7.1b). The frequency to self-care one’s
ing of the adaptation related to feeding behav- oral hygiene decreased in patients with neuro-
iour. These mental functions are crucial for degenerative disorders (Aragón et al. 2018).
individuals to adapt to changes in oral condi- Notably, cognitive impairment may also relate
tions. In addition, the brain also plays a key to a declined ability of behavioural adaptation,
role in the behaviour of oral healthcare, such as learning a new motor skill. In older
which in turn influences the structural and people, sensorimotor adaptation and learning
functional integrity of the stomatognathic (see Box 7.1) demand a great cognitive effort,
system. In the following sections, we discuss such as an increased attention control on a sen-
how the BSA framework helps to clarify some sorimotor task (Seidler et al. 2010; Ward 2006).
major postulations regarding oral–cognitive For example, during eating, older individuals
associations. may spend more time chewing and pay more
Figure 7.1 Conceptual links between oral health and cognitive functions. (a) The framework of the
brain–stomatognathic axis (BSA). The framework highlights that the brain plays a key role in behavioural
adaptation in feeding and oral care behaviour, which further relates to oral health. (b) The potential role of
the brain in behavioural adaptation. Poor oral health may be attributed to increasing difficulty in
conducting health-related behaviour (e.g. being unable to brush teeth), which is derived from
neurodegenerative disorders. (c) The potential role of oral factors in brain pathologies. The ‘oral-on-brain
effect’ may be associated with multiple factors, such as reduced sensory feedback from the loss of occlusal
contact or increased periodontal inflammation/infection. Notably, when the brain is affected by poor oral
health, it may be followed by worse adaptation of feeding and oral care behaviour, which furthers
exacerbates one’s oral health. (d) The potential role of a common factor that affects both cognitive and oral
functions. For example, aging is associated with structural and functional alterations of the cerebellum,
which relates to not only oral sensorimotor functions but also cognition. The arrow filled with a slash
pattern denotes the potential causal links of the framework.
attention to their oral conditions during chew- dementia observed in older individuals may
ing (Peyron et al. 2004; Zelig et al. 2019). reflect an increment of worse oral hygiene and
Therefore, patients with neurodegenerative deterioration of cognitive functions, both asso-
disorders may be less able to adopt new skills ciated with the lower cognitive functions dur-
for maintaining oral hygiene, thus leading to ing one’s early life (Thomson and Barak 2020).
worse oral conditions. Back to the BSA framework, the postulations
Notably, from the perspective of individual stated above highlight that worse cognitive
life course, poorer oral care can be traced back functions are associated with worse oral
to one’s early life. The individual cognitive hygiene, and cognitive functions are associated
reserve, which is associated with one’s intelli- with behavioural adaptation to oral conditions.
gence, education and occupation, may play a If the general hypothesis holds, one may expect
key role in their susceptibility to diseases that individual variability in brain functions
(Barulli and Stern 2013; Cabeza et al. 2018). related to sensorimotor adaptation and learn-
According to Thomson and Barak, better cogni- ing should be associated with oral health. For
tive functions during childhood are associated example, patients with AzD may have more
with a better cogntive function in the adulthood difficulty in learning a new skill for maintain-
and better oral health. Therefore, better cogni- ing oral hygiene, and their worse performance
tive functions during early life are associated in learning is associated with the brain mecha-
with a fewer number of missing teeth in one’s nisms of cognitive, affective and motivational
late life (Thomson and Barak 2020). Therefore, processing. Neuroimaging will be a suitable
the association between tooth loss and tool for investigating such a hypothesis.
7.4.3.2 An Emphasis on Brain Pathologies oral factor, the declined brain functions, in
Derived from Oral Factors turn, may exacerbate individual oral health
Secondly, the declined oral functions and due to lower ability in behavioural adaptation.
poorer oral hygiene may influence the brain, a For example, periodontal inflammation may
hypothesis that has been partly supported by mediate the neuroinflammatory process of the
animal research (e.g. Fukushima-Nakayama brain, which relates to worse cognitive func-
et al. 2017). Cumulating evidence revealed that tions. Patients feel more difficulty in maintain-
when mastication by occluding posterior teeth ing their oral hygiene due to the decline in
is compromised (e.g. by experimental extrac- cognitive functions, and it exacerbates their
tion), animal subjects showed a declined abil- oral conditions (e.g. leading to more serious
ity of spatial memory. Moreover, the declined periodontal inflammation). In other words,
spatial memory is associated with deficits at different mechanisms may coexist and engage
the cellular levels, primarily in the hippocam- with each other, forming a vicious cycle that
pus (for a detailed review, see Lin (2018b). The worsens both oral and cognitive functions.
benefit of the animal research is to clarify the
cause–effect relationship, i.e. an ‘oral-on-brain 7.4.3.3 An Emphasis on a Common Factor That
effect’, which can be generally described as Influences Both Cognitive and Oral Conditions
‘reducing occlusal contact may lead to worse Thirdly, one cannot exclude the possibility that
spatial memory’, though the molecular mecha- the oral–cognitive associations are dominated
nisms underlying such an effect have not been by a common factor (Figure 7.1d). For exam-
fully elucidated (Figure 7.1c). Some potential ple, aging is associated with structural and
mechanisms have been proposed. For exam- functional alterations of the cerebellum, which
ple, using transcranial Doppler ultrasound, relates to not only oral sensorimotor functions
Hasegawa et al. (2007) found a significant (Lin et al. 2015, 2019) but also cognitive–affec-
increase in blood flow velocity of the middle tive functions. The cerebellar alterations are
cerebral artery when subjects (i.e. 12 younger found in patients with declined cognitive func-
adults) were performing a clenching and gum- tions (Koziol et al. 2014). It plays a critical role
chewing task compared to the resting condi- in sensorimotor adaptation and learning
tion. The finding implies that physiological (Della-Maggiore et al. 2015). Therefore, cere-
changes in the brain may be associated with bellar deficits may influence one’s ability to
oral motor function. Also, in recent years, keep or learn new behaviour for maintaining
there has been cumulating evidence showing oral hygiene. Meanwhile, cumulating evidence
that periodontal inflammation/infection may has suggested the role of cerebellar deficits
be associated with the mechanisms of neuro- in cognitive impairment (Schmahmann et
degeneration (Kamer et al. 2015). Other al. 2019; Tabatabaei-Jafari et al. 2017).
researchers suggested that masticatory defi- Therefore, both cognitive impairment and
ciency may be associated with the effect of poor oral health may be linked to a common
stress on cognitive functions (Chen et al. 2015). deficit, which may relate to the cerebellum.
All these postulations highlight the possibility
that deficits of the brain are associated with an
7.4.4 Animal Research
early deficit of the oral conditions (Figure 7.1c).
on Neurodegenerative Disorders
However, the specific mechanisms underlying
and Oral Functions
this association have remained unclear.
Notably, this oral-on-brain framework does Animal research shows a significant advantage
not exclude the brain’s role in behaviour adap- in clarifying the causal relationship between
tation, as discussed in the preceding section. the oral and the cognitive factors by manipu-
When brain functions are compromised by an lating experimental conditions and controlling
extended the traditional view of the pathology neurological disorders because both are closely
of dementia, highlighting the role of different associated with cognitive functions. In the fol-
brain regions/networks in cognitive and senso- lowing sections, we discuss the recent findings
rimotor aspects of neurodegeneration. of the association between neurodegenerative
disorders and oral functions (Table 7.3).
7.4.6 Association Between
7.4.6.1 Oral Functions and
Neurodegenerative Disorders
Cognitive Impairment
and Oral Functions
While previous studies have investigated the
While there has been an extensive investiga- association between brain volume and tooth
tion on the brain mechanisms of neurodegen- loss by excluding patients with dementia
erative disorders, at present, these studies (Dintica et al. 2018; Kobayashi et al. 2018), a
mainly focused on the cognitive changes of the recent study has compared the association
patients. The association between oral func- between 40 cognitively impaired and 30 cogni-
tions and the brain mechanisms of neurode- tively healthy older people (Lin et al. 2020a).
generative disorders has been seldom studied. Compared to the cognitively healthy subjects,
A potential reason for this neglect is that oral the cognitively impaired patients showed a
functions are, traditionally, considered merely negative correlation between GMV at the hip-
sensorimotor processing, which is not much pocampal area and the number of tooth loss
relevant to cognition. However, as shown in (Lin et al. 2020a). Notably, the negative associ-
Chapter 4, cumulating evidence from neuro- ation (with tooth loss) was also significant in
imaging research has suggested that both mas- the motor area, including the M1 and the
tication and swallowing are associated with PMC. The findings suggested that in the
complicated cognitive–affective processing. In patients with cognitive impairment, tooth loss
other words, there may exist an overlap may be associated with a widespread reduction
between the oral domain and the domain of in GMV rather than a region-specific effect.
Table 7.3 Neuroimaging research on brain mechanisms of oral functions and neurodegenerative disorders
(since 2010), see also Table 7.2.
Notes: AD: Alzheimer’s disease; EEG: electroencephalography; HD: Huntington’s disease; PD: Parkinson’s disease;
PET: positron emission tomography; sMRI: structural magnetic resonance imaging; VBM: voxel-based morphometry.
Likewise, a recent study of 15 patients with clinically useful model for predicting individ-
cognitive impairment revealed that atrophy of ual oral functions based on brain features (Lin
grey matter was negatively correlated with the et al. 2020b). Nevertheless, the findings reveal
number of present teeth (Egashira et al. 2021). that the association between the brain and oral
However, it should be noted that all the find- function may vary between different cognitive
ings were derived from a cross-sectional data- status in older people.
set. Therefore, whether the correlation signifies
an actual cause–effect relationship (between 7.4.6.2 Oral Status and Other
tooth loss and brain features) has remained Neurodegenerative Disorders
unclear. In the cognitively healthy subjects, As shown in Table 7.3, there have been studies
their MP (assessed using an oral mixing test) focusing on Huntington’s disease (HD),
showed a positive correlation with regional Creutzfeldt–Jakob disease (CJD) and PD. The
GMV, particularly at the right PMC (Lin studies reveal not only alterations in the brain
et al. 2020a, b). Moreover, regression analyses but also in altered functions of the tongue
revealed that in the cognitively healthy sub- (Gallagher et al. 2011; Hasegawa et al. 2011;
jects, in addition to tooth loss, the GMV of the Scahill et al. 2013). Compared to healthy con-
PMC was also a significant factor predictive of trols, PD patients showed a lower tongue
the individual variability in MP. In contrast, in strength. The tongue strength was a significant
cognitively impaired patients, the number of predictor of a greater degree of clinical disabil-
tooth loss but not the GMV would predict their ity (Gallagher et al. 2011). PD patients also
MP (Lin et al. 2020b). The findings suggest that showed altered features of somatosensory
in cognitively healthy subjects, not only the cortical-evoked potentials when they received
structural deficits (i.e. an increased number of oropharyngeal mechanical stimulation (Pitts
tooth loss) but also the difference in brain fea- et al. 2016). In patients with HD, a lower
tures associated with motor control would pre- tongue force was associated with a smaller
dict their MP (Lin et al. 2020b). It is noteworthy GMV of the striatum (Scahill et al. 2013). The
that in this study, due to the limited size of the association between oral functions and the
study sample, it is difficult to establish a CJD, a neurodegenerative disorder derived
from the pathogen prion, was reported in a functions will influence how patients adapt to
case study. The patient showed dysphagia and their new dentures. Notably, such a prediction
paralysis, which can be derived from central from cognitive status to oral health (and vice
but not peripheral deficits (Hasegawa versa) may only apply to a specific context or
et al. 2011). It should be noted that in these population. For example, individual variability
studies changes in motor performance were in brain features may predict MP in cognitively
not limited to the tongue or swallowing healthy subjects, and the number of tooth loss
(Gallagher et al. 2011, Hasegawa et al. 2011, will be a better predictor of MP in cognitively
Scahill et al. 2013). In other words, deficits in impaired patients (Lin et al. 2020b). It would
oral functions may be part of an overall reduc- be oversimplified to reason that there is only a
tion of sensorimotor performance in patients single oral–cognitive association. In other
with neurodegenerative disorders. Still, the words, multiple oral–cognitive associations
findings demonstrate that deficits in oral func- may coexist, and it is critical to delineate the
tions are associated with deficits of the brain, specific context or population that a predictive
highlighting the importance of brain mecha- model applies to.
nisms in geriatric oral health.
7.4.7.2 Can We Confirm the Cause–Effect
Relationship Between Oral
7.4.7 Translational Research for Dental
and Cognitive Aspects?
Patients with Neurodegenerative Disorders
A common myth for interpreting neuroimag-
One of the advantages of neuroimaging ing findings is to over-interpret the cause–
research, in comparison to animal research, is effect relationship between the brain and
to provide direct evidence to human subjects behaviour based on correlational findings. In
about the association between oral conditions some studies, a causal statement was expressed
and alterations of the brain. However, most of by a ‘softened’ tone, such as ‘poor oral health
the neuroimaging findings were based on a may lead to cognitive decline’. However, such a
cross-sectional design with a limited sample claim selectively highlights one direction of
size. In the following sections, we discuss sev- causal effect (i.e. from oral to cognitive), while
eral critical questions related to oral–cognitive discussion of the reverse direction (i.e. from
associations and their applications in clinical cognitive to oral) is selectively neglected. The
practice. Hill criteria of causal inference may help to
identify if there is consistent evidence of the
7.4.7.1 Can We Predict the Status of Oral causal effect (Van Reekum et al. 2001). For
Health Based on Cognitive Status (and example, if poor oral health is the cause of cog-
Vice Versa)? nitive decline, it would occur earlier than the
As noted in the preceding sections, there have cognitive deficits are identified. If severe dete-
been an increasing number of studies that sup- rioration of oral health is associated with only
ported a relationship between oral conditions a very trivial change in cognitive function, it
and cognitive functions. However, the results may not be a major contributor to cognitive
merely describe an association between the status. Another critical aspect of the Hill crite-
factors and provide little information for clini- ria is the specificity of causal inference. It is
cians to predict individual differences in their noteworthy that a lack of specificity may not
symptoms. For example, a periodontist may invalidate the causal inference because both
want to know if patients’ periodontal health oral health and cognitive status are influenced
will get worsen, given that they develop cogni- by multiple factors. The lack of specificity may
tive impairment. Likewise, a prosthodontist imply indirect causal effects. For example,
may want to know whether declined cognitive poor oral health may induce worse nutrition
intake and poorer social interaction, and it is clinical decision-making and trigger a fashion
the change in the nutritional and social condi- of unnecessary treatment for vulnerable older
tions that contribute to cognitive decline. people (Thomson and Barak 2020).
References
Affoo, R.H., Foley, N., Garrick, R. et al. (2015). and structural brain connectivity in older
Meta-analysis of salivary flow rates in young age: a longitudinal connectome and
and older adults. J. Am. Geriatr. Soc. 63: tractography study. Neuroimage 183:
2142–2151. 884–896.
Alloza, C., Cox, S.R., Blesa Cábez, M. et al. Aragón, F., Zea-Sevilla, M.A., Montero, J. et al.
(2018). Polygenic risk score for schizophrenia (2018). Oral health in Alzheimer’s disease: a
multicenter case-control study. Clin. Oral Proc. Natl. Acad. Sci. U. S. A. 111:
Investig. 22: 3061–3070. E4997–E5006.
Avivi-Arber, L. and Sessle, B.J. (2018). Jaw Chantaramanee, A., Tohara, H., Nakagawa,
sensorimotor control in healthy adults and K. et al. (2019). Association between echo
effects of ageing. J. Oral Rehabil. 45: 50–80. intensity of the tongue and its thickness and
Bangcuyo, R.G. and Simons, C.T. (2017). Lingual function in elderly subjects. J. Oral Rehabil.
tactile sensitivity: effect of age group, sex, and 46: 634–639.
fungiform papillae density. Exp. Brain Res. Chen, H., Iinuma, M., Onozuka, M., and Kubo,
235: 2679–2688. K.Y. (2015). Chewing maintains
Barulli, D. and Stern, Y. (2013). Efficiency, hippocampus-dependent cognitive function.
capacity, compensation, maintenance, Int. J. Med. Sci. 12: 502–509.
plasticity: emerging concepts in cognitive Cox, S.R., Bastin, M.E., Ferguson, K.J. et al.
reserve. Trends Cogn. Sci. 17: 502–509. (2015). Compensation or inhibitory failure?
Blumen, H.M., Allali, G., Beauchet, O. et al. Testing hypotheses of age-related right frontal
(2019). A gray matter volume covariance lobe involvement in verbal memory ability
network associated with the motoric cognitive using structural and diffusion MRI. Cortex
risk syndrome: a multicohort MRI study. 63: 4–15.
J. Gerontol. A Biol. Sci. Med. Sci. 74: 884–889. Cullins, M.J. and Connor, N.P. (2017).
Brettschneider, J., Del Tredici, K., Lee, V.M., and Alterations of intrinsic tongue muscle
Trojanowski, J.Q. (2015). Spreading of pathology properties with aging. Muscle Nerve 56:
in neurodegenerative diseases: a focus on E119–E125.
human studies. Nat. Rev. Neurosci. 16: 109–120. De Freitas Borges, T., Regalo, S.C., Taba, M. Jr.
Cabeza, R., Albert, M., Belleville, S. et al. (2018). et al. (2013). Changes in masticatory
Maintenance, reserve and compensation: the performance and quality of life in individuals
cognitive neuroscience of healthy ageing. Nat. with chronic periodontitis. J. Periodontol. 84:
Rev. Neurosci. 19: 701–710. 325–331.
Calhoun, K.H., Gibson, B., Hartley, L. et al. De Rui, M., Marini, I., Bartolucci, M.L. et al.
(1992). Age-related changes in oral sensation. (2015). Pressure pain threshold of the
Laryngoscope 102: 109–116. cervico-facial muscles in healthy elderly
Callaert, D.V., Ribbens, A., Maes, F. et al. (2014). people: the role of gender, age and
Assessing age-related gray matter decline with dominance. Gerodontology 32: 274–280.
voxel-based morphometry depends Del Brutto, O.H., Mera, R.M., and Zambrano,
significantly on segmentation and M. (2016). The influence of age in the
normalization procedures. Front. Aging relationship between cerebral small vessel
Neurosci. 6: 124. disease and edentulism. The Atahualpa
Cao, M., Wang, J.H., Dai, Z.J. et al. (2014). project. Eur. Neurol. 76: 112–116.
Topological organization of the human brain Della-Maggiore, V., Landi, S.M., and Villalta,
functional connectome across the lifespan. J.I. (2015). Sensorimotor adaptation: multiple
Dev. Cogn. Neurosci. 7: 76–93. forms of plasticity in motor circuits.
Cerutti-Kopplin, D., Feine, J., Padilha, D.M. et al. Neuroscientist 21: 109–125.
(2016). Tooth loss increases the risk of Dintica, C.S., Rizzuto, D., Marseglia, A. et al.
diminished cognitive function: a systematic (2018). Tooth loss is associated with
review and meta-analysis. JDR. Clin. Trans. accelerated cognitive decline and volumetric
Res. 1: 10–19. brain differences: a population-based study.
Chan, M.Y., Park, D.C., Savalia, N.K. et al. Neurobiol. Aging 67: 23–30.
(2014). Decreased segregation of brain Dorsey, E.R., George, B.P., Leff, B., and Willis,
systems across the healthy adult lifespan. A.W. (2013). The coming crisis: obtaining care
for the growing burden of neurodegenerative Gazzaniga, M.S., Ivry, R.B., and Mangun,
conditions. Neurology 80: 1989–1996. G.R. (2019). Cognitive Neuroscience: The
Egashira, R., Umezaki, Y., Mizutani, S. et al. Biology of the Mind. W. W. Norton &
(2021). Relationship between cerebral atrophy Company.
and number of present teeth in elderly Ge, Y., Grossman, R.I., Babb, J.S. et al. (2002).
individuals with cognitive decline. Exp. Age-related total gray matter and white matter
Gerontol. 144: 111189. changes in normal adult brain. Part I:
El Tumi, H., Johnson, M.I., Dantas, P.B.F. et al. volumetric MR imaging analysis. AJNR Am.
(2017). Age-related changes in pain sensitivity J. Neuroradiol. 23: 1327–1333.
in healthy humans: a systematic review with Geerligs, L., Renken, R.J., Saliasi, E. et al. (2015).
meta-analysis. Eur. J. Pain 21: 955–964. A brain-wide study of age-related changes in
Erickson, K.I., Leckie, R.L., and Weinstein, functional connectivity. Cereb. Cortex 25:
A.M. (2014). Physical activity, fitness, and gray 1987–1999.
matter volume. Neurobiol. Aging 35 (Suppl 2): Goodpaster, B.H., Park, S.W., Harris, T.B. et al.
S20–S28. (2006). The loss of skeletal muscle strength,
Etter, N.M., Dressler, E.V., and Andreatta, mass, and quality in older adults: the health,
R.D. (2016). The relationship between labial aging and body composition study. J. Gerontol.
vibrotactile detection and pure-tone hearing A Biol. Sci. Med. Sci. 61: 1059–1064.
thresholds in healthy, ageing adults. Int. Greicius, M.D., Srivastava, G., Reiss, A.L., and
J. Speech Lang. Pathol. 18: 89–96. Menon, V. (2004). Default-mode network
Fang, M., Li, J., Lu, G. et al. (2005). A fMRI study activity distinguishes Alzheimer’s disease
of age-related differential cortical patterns from healthy aging: evidence from functional
during cued motor movement. Brain Topogr. MRI. Proc. Natl. Acad. Sci. U. S. A. 101:
17: 127–137. 4637–4642.
Feng, P., Huang, L., and Wang, H. (2014). Taste Hara, K., Tohara, H., Namiki, C. et al. (2020).
bud homeostasis in health, disease, and aging. Relationship between displacement of the
Chem. Senses 39: 3–16. masseter muscle during biting and
Ferreira, L.K., Diniz, B.S., Forlenza, O.V. et al. masseter muscle quality and bite force in
(2011). Neurostructural predictors of healthy elderly persons. J. Oral Rehabil. 47:
Alzheimer’s disease: a meta-analysis of VBM 441–448.
studies. Neurobiol. Aging 32: 1733–1741. Hasegawa, Y., Ono, T., Hori, K., and Nokubi,
Friedman, P.K. and Lamster, I.B. (2016). Tooth loss T. (2007). Influence of human jaw
as a predictor of shortened longevity: exploring movement on cerebral blood flow. J. Dent. Res.
the hypothesis. Periodontol 2000. 72: 142–152. 86: 64–68.
Fukushima-Nakayama, Y., Ono, T., Hayashi, Hasegawa, J., Okumura, Y., Osumi, E. et al.
M. et al. (2017). Reduced mastication (2011). Creutzfeldt–Jakob disease with
impairs memory function. J. Dent. Res. 96: paralysis of the unilateral vocal cord and soft
1058–1066. palate. Tohoku J. Exp. Med. 225: 277–283.
Gallagher, C.L., Johnson, S.C., Bendlin, Hatch, J.P., Shinkai, R.S., Sakai, S. et al. (2001).
B.B. et al. (2011). A longitudinal study of Determinants of masticatory performance in
motor performance and striatal dentate adults. Arch. Oral Biol. 46: 641–648.
[18F]fluorodopa uptake in Parkinson’s Hedden, T. and Gabrieli, J.D. (2004). Insights
disease. Brain Imaging Behav. 5: 203–211. into the ageing mind: a view from cognitive
Gargouri, F., Kallel, F., Delphine, S. et al. (2018). neuroscience. Nat. Rev. Neurosci. 5: 87–96.
The influence of preprocessing steps on graph Heft, M.W. and Robinson, M.E. (2010). Age
theory measures derived from resting state differences in orofacial sensory thresholds.
fMRI. Front. Comput. Neurosci. 12: 8. J. Dent. Res. 89: 1102–1105.
Heft, M.W. and Robinson, M.E. (2017). Kimoto, K., Ono, Y., Tachibana, A. et al. (2011).
Somatosensory function in old age. J. Oral Chewing-induced regional brain activity in
Rehabil. 44: 327–332. edentulous patients who received mandibular
Honda, M., Iida, T., Kamiyama, H. et al. (2019). implant-supported overdentures: a preliminary
Mechanical sensitivity and psychological report. J. Prosthodont. Res. 55: 89–97.
factors in patients with burning mouth King, B.R., Van Ruitenbeek, P., Leunissen,
syndrome. Clin. Oral Investig. 23: 757–762. I. et al. (2018). Age-related declines in motor
Hou, Y., Dan, X., Babbar, M. et al. (2019). Ageing performance are associated with decreased
as a risk factor for neurodegenerative disease. segregation of large-scale resting state brain
Nat. Rev. Neurol. 15: 565–581. networks. Cereb. Cortex 28: 4390–4402.
Hsu, H.C. and Jones, B.L. (2012). Multiple Kobayashi, T., Kubota, M., Takahashi, T. et al.
trajectories of successful aging of older and (2018). Effects of tooth loss on brain structure:
younger cohorts. Gerontologist 52: 843–856. a voxel-based morphometry study.
Huang, C.C., Hsieh, W.J., Lee, P.L. et al. (2015). J. Prosthodont. Res. 62: 337–341.
Age-related changes in resting-state networks Kobayashi, T., Fukami, H., Ishikawa, E. et al.
of a large sample size of healthy elderly. CNS (2020). An fMRI study of the brain network
Neurosci. Ther. 21: 817–825. involved in teeth tapping in elderly adults.
Humbert, I.A., Fitzgerald, M.E., McLaren, Front. Aging Neurosci. 12: 32.
D.G. et al. (2009). Neurophysiology of Koini, M., Duering, M., Gesierich, B.G. et al.
swallowing: effects of age and bolus type. (2018). Grey-matter network disintegration as
Neuroimage 44: 982–991. predictor of cognitive and motor function
Ikebe, K., Matsuda, K., Kagawa, R. et al. (2012). with aging. Brain Struct. Funct. 223:
Masticatory performance in older subjects 2475–2487.
with varying degrees of tooth loss. J. Dent. Koppelmans, V., Hirsiger, S., Mérillat, S. et al.
40: 71–76. (2015). Cerebellar gray and white matter
Johansson, A.S., Svensson, K.G., and Trulsson, volume and their relation with age and
M. (2006). Impaired masticatory behavior in manual motor performance in healthy older
subjects with reduced periodontal tissue adults. Hum. Brain Mapp. 36: 2352–2363.
support. J. Periodontol. 77: 1491–1497. Kosaka, T., Ono, T., Kida, M. et al. (2016).
Kamer, A.R., Pirraglia, E., Tsui, W. et al. (2015). A multifactorial model of masticatory
Periodontal disease associates with higher performance: the Suita study. J. Oral Rehabil.
brain amyloid load in normal elderly. 43: 340–347.
Neurobiol. Aging 36: 627–633. Kossioni, A.E. and Dontas, A.S. (2007). The
Kamiya, K., Narita, N., and Iwaki, S. (2016). stomatognathic system in the elderly. Useful
Improved prefrontal activity and chewing information for the medical practitioner. Clin.
performance as function of wearing denture Interv. Aging 2: 591–597.
in partially edentulous elderly individuals: Koziol, L.F., Budding, D., Andreasen, N. et al.
functional near-infrared spectroscopy study. (2014). Consensus paper: the cerebellum’s role
PLoS One 11: e0158070. in movement and cognition. Cerebellum 13:
Kassebaum, N.J., Bernabe, E., Dahiya, M. et al. 151–177.
(2014a). Global burden of severe periodontitis Kubo, K.Y., Murabayashi, C., Kotachi, M. et al.
in 1990–2010: a systematic review and (2017). Tooth loss early in life suppresses
meta-regression. J. Dent. Res. 93: 1045–1053. neurogenesis and synaptophysin expression in
Kassebaum, N.J., Bernabe, E., Dahiya, M. et al. the hippocampus and impairs learning in
(2014b). Global burden of severe tooth loss: a mice. Arch. Oral Biol. 74: 21–27.
systematic review and meta-analysis. J. Dent. Lamster, I.B., Asadourian, L., Del Carmen, T.,
Res. 93: 20S–28S. and Friedman, P.K. (2016). The aging mouth:
differentiating normal aging from disease. brain volume and masticatory performance
Periodontol. 2000 72: 96–107. differed in cognitively impaired and non-
Lemaitre, H., Goldman, A.L., Sambataro, F. et al. impaired older people. Exp. Gerontol.
(2012). Normal age-related brain morphometric 137: 110942.
changes: nonuniformity across cortical Lindeboom, J. and Weinstein, H. (2004).
thickness, surface area and gray matter Neuropsychology of cognitive ageing,
volume? Neurobiol. Aging 33 (617): e1–e9. minimal cognitive impairment, Alzheimer’s
Liinavuori, A., Tolvanen, M., Pohjola, V., and disease, and vascular cognitive impairment.
Lahti, S. (2016). Changes in dental fear among Eur. J. Pharmacol. 490: 83–86.
Finnish adults: a national survey. Community Liu, H., Qin, L., Wu, Y. et al. (2019). Oral
Dent. Oral Epidemiol. 44: 128–134. physiological characteristics among Chinese
Lin, C.S. (2018a). Meta-analysis of brain subjects in the eastern region of China. Arch.
mechanisms of chewing and clenching Oral Biol. 108: 104539.
movements. J. Oral Rehabil. 45: 627–639. Lopez, R., Smith, P.C., Gostemeyer, G., and
Lin, C.S. (2018b). Revisiting the link between Schwendicke, F. (2017). Ageing, dental caries
cognitive decline and masticatory dysfunction. and periodontal diseases. J. Clin. Periodontol.
BMC Geriatr. 18: 5. 44 (Suppl 18): S145–S152.
Lin, C.S. (2019). Functional adaptation of Lövdén, M., Köhncke, Y., Laukka, E.J. et al.
Oromotor functions and aging: a focused (2014). Changes in perceptual speed and
review of the evidence from brain white matter microstructure in the
neuroimaging research. Front. Aging Neurosci. corticospinal tract are associated in very old
11: 354. age. Neuroimage 2 (102 Pt): 520–530.
Lin, C.S., Wu, S.Y., Wu, C.Y. et al. (2015). Lund, J.P. (1991). Mastication and its control by
Gray matter volume and resting-state the brain stem. Crit Rev Oral Biol Med. 2: 33–64.
functional connectivity of the motor Ly, M., Canu, E., Xu, G. et al. (2014). Midlife
cortex-cerebellum network reflect the measurements of white matter microstructure
individual variation in masticatory predict subsequent regional white matter
performance in healthy elderly people. atrophy in healthy adults. Hum. Brain Mapp.
Front. Aging Neurosci. 7: 247. 35: 2044–2054.
Lin, C.S., Wu, C.Y., Wu, S.Y. et al. (2017a). Machida, N., Tohara, H., Hara, K. et al. (2017).
Age-and sex-related differences in masseter Effects of aging and sarcopenia on tongue
size and its role in oral functions. J. Am. Dent. pressure and jaw-opening force. Geriatr.
Assoc. 148: 644–653. Gerontol. Int. 17: 295–301.
Lin, C.S., Wu, C.Y., Wu, S.Y. et al. (2017b). Malandraki, G.A., Sutton, B.P., Perlman, A.L.,
Age-related difference in functional brain and Karampinos, D.C. (2010). Age-related
connectivity of mastication. Front. Aging differences in laterality of cortical activations
Neurosci. 9: 82. in swallowing. Dysphagia 25: 238–249.
Lin, C.S., Wu, C.Y., Wang, D.H. et al. (2019). Malandraki, G.A., Perlman, A.L., Karampinos,
Brain signatures associated with swallowing D.C., and Sutton, B.P. (2011). Reduced
efficiency in older people. Exp. Gerontol. somatosensory activations in swallowing with
115: 1–8. age. Hum. Brain Mapp. 32: 730–743.
Lin, C.S., Lin, H.H., Fann, S.W. et al. (2020a). Martin, R., Barr, A., Macintosh, B. et al. (2007).
Association between tooth loss and gray Cerebral cortical processing of swallowing in
matter volume in cognitive impairment. Brain older adults. Exp. Brain Res. 176: 12–22.
Imaging Behav. 14: 396–407. McComas, A.J. (1998). Oro-facial muscles:
Lin, C.S., Lin, H.H., Wang, S.J., and Fuh, internal structure, function and ageing.
J.L. (2020b). Association between regional Gerodontology 15: 3–14.
McHugh, D. and Gil, J. (2018). Senescence and function until swallowing using gummy jelly
aging: causes, consequences, and therapeutic in subjects aged 20–79 years. J. Oral Rehabil.
avenues. J. Cell Biol. 217: 65–77. 47: 851–861.
Monemi, M., Eriksson, P.O., Eriksson, A., and Okamoto, N., Morikawa, M., Amano, N. et al.
Thornell, L.E. (1998). Adverse changes in fibre (2017). Effects of tooth loss and the
type composition of the human masseter apolipoprotein E varepsilon4 allele on mild
versus biceps brachii muscle during aging. memory impairment in the Fujiwara-kyo
J. Neurol. Sci. 154: 35–48. study of Japan: a nested case–control study.
Moon, H.I., Nam, J.S., Leem, M.J., and Kim, J. Alzheimers Dis. 55: 575–583.
K.H. (2017). Periventricular white matter Onozuka, M., Fujita, M., Watanabe, K. et al.
lesions as a prognostic factor of swallowing (2003). Age-related changes in brain regional
function in older patients with mild stroke. activity during chewing: a functional
Dysphagia 32: 480–486. magnetic resonance imaging study. J. Dent.
Nakashima, Y., Kimoto, S., Ogawa, T. et al. Res. 82: 657–660.
(2015). Characteristics of the pain tolerance Oue, H., Miyamoto, Y., Okada, S. et al. (2013).
threshold induced by electrical stimulation of Tooth loss induces memory impairment and
the alveolar ridge. Clin. Exp. Dent. Res. neuronal cell loss in APP transgenic mice.
1: 80–86. Behav. Brain Res. 252: 318–325.
Nagashima, S., Kimoto, K., Ono, Y. et al. (2020). Pang, Q., Hu, X., Li, X. et al. (2015). Behavioral
The effect of masticatory behaviour on impairments and changes of nitric oxide and
generalized attention in heathy volunteers. inducible nitric oxide synthase in the brains of
Psychogeriatrics 20: 254–261. molarless Km mice. Behav. Brain Res. 278:
Narita, N., Ishii, T., Iwaki, S. et al. (2019). 411–416.
Prefrontal consolidation and compensation as Park, J.W., Sim, G.J., Kim, H.J. et al. (2017).
a function of wearing denture in partially Changes of cortical activation in swallowing
edentulous elderly patients. Front. Aging following high frequency repetitive
Neurosci. 11: 375. transcranial magnetic stimulation in older
Ogawa, T., Annear, M.J., Ikebe, K., and Maeda, adults. Neurogastroenterol. Motil. 29: e13123.
Y. (2017a). Taste-related sensations in old age. Park, J.W., Kim, H., Park, T. et al. (2019). A pilot
J. Oral Rehabil. 44: 626–635. study of the effects of high-frequency
Ogawa, T., Uota, M., Ikebe, K. et al. (2017b). repetitive transcranial magnetic stimulation
Longitudinal study of factors affecting taste on dysphagia in the elderly.
sense decline in old–old individuals. J. Oral Neurogastroenterol. Motil. 31: e13561.
Rehabil. 44: 22–29. Peelle, J.E., Cusack, R., and Henson, R.N. (2012).
Oguchi, K., Saitoh, E., Baba, M. et al. (2000a). Adjusting for global effects in voxel-based
The repetitive saliva swallowing test (RSST) as morphometry: gray matter decline in normal
a screening test of functional dysphagia (2) aging. Neuroimage 60: 1503–1516.
validity of RSST. Jpn. J. Rehabil. Med. 37: Percival, R.S., Challacombe, S.J., and Marsh,
383–388. P.D. (1994). Flow rates of resting whole and
Oguchi, K., Saitoh, E., Mizuno, M. et al. (2000b). stimulated parotid saliva in relation to age and
The repetitive saliva swallowing test (RSST) as gender. J. Dent. Res. 73: 1416–1420.
a screening test of functional dysphagia (1) Persson, E., Wårdh, I., and Östberg, P. (2019).
Normal values of RSST. Jpn. J. Rehabil. Med. Repetitive saliva swallowing test: norms,
37: 375–382. clinical relevance and the impact of saliva
Ohno, K., Fujita, Y., Ohno, Y. et al. (2020). secretion. Dysphagia 34: 271–278.
The factors related to decreases in Peyron, M.A., Blanc, O., Lund, J.P., and Woda,
masticatory performance and masticatory A. (2004). Influence of age on adaptability of
human brain networks. Brain Connect. 4: associated with normal aging. Neuroimage 52:
662–676. 1215–1223.
Stein, P.S., Desrosiers, M., Donegan, S.J. et al. Thomson, W.M. and Barak, Y. (2020). Tooth loss
(2007). Tooth loss, dementia and and dementia: a critical examination. J. Dent.
neuropathology in the Nun study. J. Am. Dent. Res. 100: 226–331.
Assoc. 138: 1314–1322. quiz 1381–2. Tomasi, D. and Volkow, N.D. (2012). Aging and
Stoodley, C.J. and Schmahmann, J.D. (2010). functional brain networks. Mol. Psychiatry 17
Evidence for topographic organization in (471): 549–558.
the cerebellum of motor control versus Tomsen, N., Ortega, O., Rofes, L. et al. (2019).
cognitive and affective processing. Cortex Acute and subacute effects of oropharyngeal
46: 831–844. sensory stimulation with TRPV1 agonists in
Storsve, A.B., Fjell, A.M., Tamnes, C.K. et al. older patients with oropharyngeal dysphagia:
(2014). Differential longitudinal changes in a biomechanical and neurophysiological
cortical thickness, surface area and volume randomized pilot study. Therap. Adv.
across the adult life span: regions of Gastroenterol. 12: 1756284819842043.
accelerating and decelerating change. Tomsen, N., Alvarez-Berdugo, D., Rofes, L. et al.
J. Neurosci. 34: 8488–8498. (2020). A randomized clinical trial on the
Tabatabaei-Jafari, H., Walsh, E., Shaw, M.E. et al. acute therapeutic effect of TRPA1 and TRPM8
(2017). The cerebellum shrinks faster than agonists in patients with oropharyngeal
normal ageing in Alzheimer’s disease but not dysphagia. Neurogastroenterol. Motil.
in mild cognitive impairment. Hum. Brain 32: e13821.
Mapp. 38: 3141–3150. Tonsekar, P.P., Jiang, S.S., and Yue, G. (2017).
Tada, A. and Miura, H. (2017). Association Periodontal disease, tooth loss and dementia:
between mastication and cognitive status: a is there a link? A systematic review.
systematic review. Arch. Gerontol. Geriatr. Gerodontology 34: 151–163.
70: 44–53. Trulsson, M. (2006). Sensory-motor function of
Takeda, Y., Oue, H., Okada, S. et al. (2016). human periodontal mechanoreceptors. J. Oral
Molar loss and powder diet leads to memory Rehabil. 33: 262–273.
deficit and modifies the mrna expression of Van Reekum, R., Streiner, D.L., and Conn,
brain-derived neurotrophic factor in the D.K. (2001). Applying Bradford Hill’s criteria
hippocampus of adult mice. BMC for causation to neuropsychiatry: challenges
Neurosci. 17: 81. and opportunities. J. Neuropsychiatry Clin.
Takeuchi, K., Ohara, T., Furuta, M. et al. (2017). Neurosci. 13: 318–325.
Tooth loss and risk of dementia in the Veyrune, J.L., Lassauzay, C., Nicolas, E. et al.
community: the Hisayama study. J. Am. (2007). Mastication of model products in
Geriatr. Soc. 65: e95–e100. complete denture wearers. Arch. Oral Biol. 52:
Taki, Y., Thyreau, B., Kinomura, S. et al. (2013). 1180–1185.
A longitudinal study of age-and gender- Villa, A., Connell, C.L., and Abati, S. (2015).
related annual rate of volume changes in Diagnosis and management of xerostomia and
regional gray matter in healthy adults. Hum. hyposalivation. Ther. Clin. Risk Manag.
Brain Mapp. 34: 2292–2301. 11: 45–51.
Teismann, I.K., Steinstraeter, O., Schwindt, Ward, N.S. (2006). Compensatory mechanisms
W. et al. (2010). Age-related changes in in the aging motor system. Ageing Res. Rev. 5:
cortical swallowing processing. Neurobiol. 239–254.
Aging 31: 1044–1050. Weijenberg, R.A., Lobbezoo, F., Visscher, C.M.,
Thambisetty, M., Wan, J., Carass, A. et al. (2010). and Scherder, E.J. (2015). Oral mixing ability
Longitudinal changes in cortical thickness and cognition in elderly persons with
Dental Neuroimaging: The Role of the Brain in Oral Functions, First Edition. Chia-Shu Lin.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
Companion website: www.wiley.com/go/lin/dental-neuroimaging
Figure 8.1 The concept of neuroplasticity and functional adaptation. (a) With a ‘pre-programmed’ nervous
system, our behaviours responding to environmental stimuli are determined by a fixed set of stimulus–
response links. For example, a great danger will elicit a stronger emotional response compared to a mild
danger (the upper panel). However, our nervous system is modifiable and can be tuned according to
environmental changes. Long-term experience may sculpt the brain at the structural and functional levels,
leading to different behaviours responding to environmental stimuli. For example, past experience may
predispose the brain to be more sensitive to danger and make a stronger emotional response (the lower
panel). (b) Functional adaptation is associated with the improvement of one’s functional performance under
environmental challenges. For example, individuals learn to run faster (i.e. increased performance) when
they are threatened by a greater danger (the upper panel). Compensation, in contrast, highlights the
restoration of performance from a worse status back to a normal status. For example, individuals with a
disability can run as quickly as normal individuals when facing danger by compensating their mobility with
the help from tools and rehabilitative therapy (the lower panel).
chronic pain, their behaviour towards stimuli several misconceptions regarding plasticity.
may vary depending on their past experience Firstly, brain plasticity is often linked to the
of pain. In other words, our behaviour is modi- recovery of functions in patients with brain
fiable rather than pre-programmed, and the disorders (e.g. stroke or trauma). Therefore,
neural system is flexible to be tuned according brain plasticity is sometimes conceived as an
to environmental changes rather than hard improvement responding to the malfunction
wired. For example, long-term experience of caused by pathological conditions. This is not
chronic pain may sculpt the brain at the struc- entirely correct because plasticity is ‘an intrin-
tural and functional levels, leading to different sic property of the nervous system retained
behavioural responses to environmental stim- throughout a lifespan’ (Pascual-Leone
uli (Figure 8.1a). et al. 2005). The brain is continuously being
moulded as long as it interacts with the envi-
8.1.2.2 Misconceptions About Brain Plasticity ronment, as a normal ongoing state (Pascual-
Brain plasticity is often linked to the concept of Leone et al. 2005). Therefore, it is not a
functional adaptation, which is associated phenomenon specific to a disease. Secondly,
with the improvement of one’s functional per- plasticity is associated with behavioural modi-
formance under environmental challenges fication, which can be acquired by learning
(Frisancho 1993) (Figure 8.1b). There are (e.g. to develop a skill of using a high-speed
dental handpiece). Therefore, plasticity is often term ‘compensation’ may refer to a process
considered for the gain of new abilities. associated with the damage of the CNS and
However, the behavioural change may not deficits in functions. The compensating brain
necessarily be an increase in capability. It may should ‘assist in maintaining a normal level of
be loss of an already gained function, the behaviour’ (Gregory et al. 2017). Therefore, a
release of a previously suppressed function or prerequisite condition for functional compen-
expressing a similar function by different sation is that one’s performance returns to a
neural substrates (Pascual-Leone et al. 2005). normal level. From the perspective of ageing
Moreover, even if the behavioural change is a neuroscience, the term also refers to the situa-
positive one (e.g. acquiring expertise in danc- tion ‘in which age-related increases in brain
ing), such a change does not necessarily activity are directly correlated with better per-
correspond to a ‘positive’ change of the brain formance in older adults’ (Cabeza et al. 2018).
(e.g. an increased amount of a brain feature). Both the perspectives from neurology and
For example, a previous study of structural ageing neuroscience consistently highlight
brain features found that ballet dancers, com- that compensation relates to the change of
pared to nondancers, showed a lower (but not performance from a sub-normal level (due to
higher) grey matter volume (GMV) in multiple diseases, trauma or ageing) to a normal level
regions of sensorimotor control and learning (Figure 8.1b).
(Hänggi et al. 2010). One should be careful
when making inferences between the modifi-
8.1.3 Neural Mechanisms of Plasticity
cation at behavioural and brain aspects, which
can be far complicated at the cellular level The concept of neuroplasticity has evoked
(Zatorre et al. 2012). great interest in neuroscience. However, the
mechanisms underlying neuroplasticity are
8.1.2.3 Recovery, Reorganization complicated because of the complexity of the
and Compensation nervous system. The nervous system is organ-
From the neurological perspective, the term ized with a hierarchical architecture, from the
‘recovery’ mostly applies to conditions when cellular level (e.g. a neuron) to the brain level
patients suffer brain damage. For example, (e.g. the whole-brain connectome). Therefore,
post-traumatic patients may initially lose some researchers have used different approaches to
functions corresponding to the regions of brain investigate neuroplasticity, from the changes
damage. Recovery is identified when patients of synaptic activity to the changes of func-
regain their performance to the pre-traumatic tional/structural brain connectome. Many
level. Such a recovery is more commonly seen neuroimaging research focuses on neuroplasti-
for the functions of cognitive domains city at the whole-brain scale. The major aim of
(Mogensen 2011). Functional recovery is asso- these studies is to identify a structural or func-
ciated with neural reorganization, which is tional change corresponding to the effect of
usually promoted by training programmes for long-term experience, which can be the acqui-
rehabilitative purposes. Notably, even without sition of a new skill or the suffering from
explicit training, patients are continuously chronic pain. Because the nervous system is
challenged by daily life activities as an ‘infor- organized in a hierarchical architecture, the
mal’ training and reorganization may still alterations at the whole-brain scale reflect the
undergo (Mogensen 2011). Therefore, from the changes in the synapse, which is the major
behavioural perspective, the term ‘reorganiza- ‘site’ where plasticity occurs. Synaptic plastic-
tion’ is often associated with brain plasticity, ity can be altered in several ways, including
especially in the literature on the acquisition of changing synaptic structures, such as the
new skills and expertise (Chang 2014). The dendritic spines and filopodia, adding new
synapses and removing pre-existed synapses neuroplasticity beyond the synaptic level.
(Fu and Zuo 2011). For example, during motor Neuroimaging contributes to our understand-
training, elimination of the pre-existed spines ing of brain plasticity by associating the struc-
and generation of new spines were identified tural and functional changes of neuroimaging
in mice (Xu et al. 2009). When adult animals findings with changes in individual experience
received lesions and sensory manipulations, or learning process (Zatorre et al. 2012). In the
the dynamics of axonal arborization were following section, we present several notable
modified (Fu and Zuo 2011). At the synaptic examples of neuroimaging research on brain
level, both de novo growth and removal of den- plasticity, focusing on the structural and func-
dritic spines are associated with neuroplasti- tional features identified via magnetic reso-
city (Fu and Zuo 2011). nance imaging (MRI).
In addition to the structural changes at the
synaptic level, neuroplasticity can be investi- 8.1.4.1 Plasticity Identified by Structural
gated by assessing the functional changes of Brain Features
synaptic activity. Traditionally, the association In a pioneering study, Maguire et al. (2000)
between structural and functional changes is compared GMV of licensed London taxi driv-
interpreted under the framework of Hebbian ers, who were experts in spatial navigation,
plasticity (Hebb 1949), which is generally sum- with control subjects (non-taxi drivers). They
marized as ‘units that fire together, wire found that an increased GMV in the posterior
together’ (Munakata and Pfaffly 2004). In hippocampus of the taxi-driver group vs. con-
other words, structural changes (i.e. how neu- trols and, importantly, the posterior hip-
rons are wired) are associated with the neural pocampal volume was positively correlated
activity (i.e. firing a neuronal signal) related to with the duration they serve as taxi drivers
functional changes. The activity-dependent (Maguire et al. 2000) (Figure 8.2a). The study
alteration of synaptic plasticity includes the showed that the individual variability in struc-
long-lasting strengthening (i.e. long-term tural features may be associated with the effect
potentiation, LTP) and the long-lasting reduc- of the plasticity of experience/learning. In
tion of the synaptic strength (i.e. long-term another landmark study, Draganski et al.
depression, LTD). In terms of Hebbian plastic- (2004) identified the structural alteration
ity, the strength of synaptic plasticity is associ- related to a training course of juggling. Using
ated with the correlation between presynaptic a longitudinal design, they identified changes
and postsynaptic activities. Animal research in regional GMV before vs. after the juggling
has revealed that LTP occurs when presynaptic exercise. Moreover, the regional GMV change
and postsynaptic activities are correlated, and was associated with the performance of a jug-
the activation of the postsynaptic N-methyl-d- gling task (Draganski et al. 2004). Subsequently,
aspartate receptor (NMDAR) plays a key role Scholz et al. (2009) investigated the plasticity
in the plasticity (Smith et al. 2009). In general, of white matter associated with juggling train-
the structural and functional changes are ing using diffusion tensor imaging (DTI). They
closely associated at the synaptic level, and found an increased fractional anisotropy (FA)
both are fundamental for the neural mecha- in the intraparietal sulcus during the post-
nisms of plasticity. training scan compared to the pre-training
scan. Furthermore, the increased FA sus-
tained for four weeks after the subjects ceased
8.1.4 Neuroimaging as a Tool
to practice juggling, suggesting a long-lasting
for Studying Brain Plasticity
effect (Scholz et al. 2009) (Figure 8.2b).
The advent of neuroimaging has offered a Together, the findings reveal that alterations
great opportunity for researchers to investigate of grey and white matter may reflect the
Figure 8.2 Experimental design of neuroimaging research on brain plasticity. (a) A cross-sectional study
reveals a significant difference in structural brain features between subjects with and without a
professional skill (e.g. driving). Individual variations in brain features further relate to the duration of
skill training (the left panel). Results from the cross-sectional design only suggest, but not confirm, the
causal direction of a plastic effect. The difference in brain features (as observed by neuroimaging) may be
attributed to prior experience of training (the middle panel). However, it is also possible that the
individual differences in the brain features predispose their performance of a skill (the right panel). (b) To
better elucidate the plastic effect of training, a longitudinal design is used to assess the performance and
brain features at different stages of training. Importantly, one can assess whether the plastic effect, i.e.
changes in brain features during the training session, can last for a period or vanish right after the
termination of training. (c) The same longitudinal design also helps to elucidate individual differences in
their performance. For example, the variation in brain features may account for the difference in
learning speed.
plasticity related to sensorimotor training in improved from Day 2 to Day 5, the activation
human subjects. of the motor-related regions (including the
PMC and the cerebellum) decreased. Across
8.1.4.2 Plasticity Identified by Functional individual subjects, a greater performance
Brain Features was associated with lower activation of
In addition to the structural features, func- these regions (Steele and Penhune 2010).
tional magnetic resonance imaging (fMRI) Lewis et al. (2009) investigated the changes
has been widely used to assess the func- in the intrinsic functional network before
tional features associated with learning. For and after subjects received training on a
example, Steele and Penhune (2010) investi- shape discrimination task. They found an
gated the brain activation associated with activation at the visual cortex associated
motor sequence learning in younger adults. with the modulation of visual stimuli. In
They identified brain activation at the addition, they identified more negative con-
regions related to motor learning (including nectivity between the dorsal attention net-
the premotor cortex [PMC], the supplemen- work and the modulated area of the visual
tary motor area [SMA], the basal ganglia cortex after training (vs. before training). In
and the cerebellum) during the early stage contrast, less negative connectivity between
of learning. Notably, as the performance the default mode network (DMN) and the
unmodulated area of the visual was identi- revealed changes in regional activation asso-
fied after training (vs. before training) ciated with the performance of learning (e.g.
(Lewis et al. 2009). The findings suggest that Steele and Penhune 2010), the association
brain plasticity may be associated with between the functional features of neuroim-
changes in the functional connectivity aging and the underlying synaptic changes
between brain regions, in addition to the has remained unclear (Poldrack 2000). Using
activation of a single region. In another lon- DTI, researchers have identified the altera-
gitudinal study with a longer training tions of white matter in rats trained with a
period, Herholz et al. (2016) investigated the spatial navigation task for five days.
changes in brain activation after six-week Furthermore, they found that at the brain
piano training and the association between level, the alterations observed via DTI were
the activation and the learning rate of sub- associated with an increase in glial fibrillary
jects (Figure 8.2c). They found that the acti- acidic protein (GFAP) and synaptophysin,
vation of the PMC and the cerebellum was suggesting an effect of synaptic plasticity at
associated with the ‘learning effect’ (i.e. the cellular level (Blumenfeld-Katzir
after > before piano training). Moreover, the et al. 2011). Another study has identified that
brain activity before training (including the in human subjects, learning-induced plastic-
primary auditory cortex, the hippocampus ity on the white matter can be identified just
and the caudate nucleus) was predictive to after two hours of training (Sagi et al. 2012).
the learning rate that subjects demonstrated In this study, DTI was performed before and
during training (Herholz et al. 2016). In gen- after subjects played a computer game for
eral, the findings suggest that functional two hours. They identified a decrease in
features, including task-related activation mean diffusivity in the hippocampus and the
and the pattern of intrinsic functional fea- parahippocampus, the regions associated
tures, can be sculpted by motor or percep- with spatial navigation. Moreover, they had
tual learning. the animal subjects perform a water maze
task and found the same plasticity effect on
the hippocampus, with an increased GFAP
8.1.5 Limitations of Neuroimaging
and brain-derived neurotrophic factor
Research on Neuroplasticity
(BDNF), a factor for neuronal growth. In gen-
The above-mentioned studies have demon- eral, the findings from combined animal and
strated neuroimaging as a powerful tool to neuroimaging research suggest that brain
investigate the brain plasticity associated with plasticity identified using neuroimaging
learning. However, there remain several criti- methods may be associated with the cellular
cal questions unanswered. In the following mechanisms of synaptic plasticity.
sections, we focus on the association between
neuroimaging features and synaptic mecha- 8.1.5.2 Association Between Plasticity Induced
nisms and the association between plasticity by Training and Experience
induced by training and experience. While neuroimaging studies focus on the neu-
roplasticity of learning, such ‘learning’ is
8.1.5.1 Association Between Neuroimaging mostly induced by an explicit and well-defined
Features and Synaptic Mechanisms training protocol (e.g. piano training [Herholz
Synaptic plasticity plays a key role in neuro- et al. 2016] and juggling [Scholz et al. 2009]).
plasticity. However, the association between In contrast, much of our experience was
the plasticity at the synaptic level and the shaped implicitly by long-term interaction
plasticity at the brain level has not yet been with the environment. For example, the brain
clarified. For example, even though fMRI plasticity of chronic pain (see Chapter 6) and
chronic stress has been widely reported 8.2 Adaptation of Pain and Oral
(Radley et al. 2015). In most of these studies, Sensory Functions
changes in brain features are associated with
individual differences of the cognitive–affec- 8.2.1 Introduction
tive aspects (e.g. pain intensity, perceived stress
and disability) rather than the performance In the preceding chapters, we have discussed
quantified by a standardized task. However, the mechanisms of the human oral motor
one should keep in mind that the long-term (Chapter 4) and sensory functions (Chapters 5
experience of pain or stress is modulated by and 6), with the main focus on the interaction
the interaction with environmental stimuli, between the brain and the stomatognathic sys-
which are presented as a ‘background’ stimu- tem. In general, we interpreted the findings
lus. For example, chronic pain usually relates based on the oral–brain-behaviour (OBB)
to avoidance behaviour. Therefore, how indi- framework (see Chapter 1), highlighting the
viduals actively adapt to the environment plays exchange of sensory and motor information
a key role in the shaping of one’s long-term between the brain and the oral apparatus.
experience. It is also noteworthy that during However, the brain–stomatognathic associa-
training individual performance is associated tion is not a static one. According to the brain–
with their past experience. In other words, stomatognathic axis (BSA) framework, both
there exists a great overlap between the plastic- the brain and the stomatognathic system
ity shaped by long-term experience and a spe- respond to various environmental and bodily
cific training protocol. It requires further changes, which include the challenge of differ-
investigation of how these factors interactively ent foods, deficits of oral structure (e.g. tooth
form the basis of neuroplasticity. loss) and ageing. In addition to the essential
sensorimotor functions, the ability to maintain
feeding behaviour by adapting to the changing
8.1.6 Summary environmental and bodily conditions should
●● Brain plasticity refers to the capacity for the be considered as one of the primary functions
nervous system to be reorganized function- of the stomatognathic system. In this chapter,
ally and structurally. Such a reorganization we look into how individuals adapt to oral sen-
can be moulded by environmental changes sory stimuli. The role of the brain in such an
and experiences. adaptive process is highlighted.
●● Neuroplasticity may be represented at a dif-
ferent scale of the neural system. At the
8.2.2 The Manifolds of Adaptation
whole-brain level, experience and training
may mould the large-scale functional and Before the following discussion, first, we may
structural networks of the brain. At the syn- need to define the concepts of adaptation,
aptic level, both de novo growth and removal which plays a key role in the understanding of
of dendritic spines are associated with human behaviour. In a very broad sense, adap-
neuroplasticity. tation can be defined as ‘a process whereby the
●● Alterations of grey and white matter, as neu- organism has attained a beneficial adjustment
roimaging features, may reflect the plasticity to the environment’ (Frisancho 1993). To high-
related to sensorimotor training in human light the brain–stomatognathic association in
subjects. adaptation, we clarify the following concepts:
●● Functional brain features, including task- sensory adaptation (for oral sensory stimuli
related activation and the pattern of intrinsic and pain), behavioural adaptation (for main-
functional features, can be sculpted by motor taining feeding behaviour) and the difference
and perceptual learning. between functional and structural adaptation.
activation of the pain-adaptation paradigm, with the ACC and the vmPFC, may play a key
which demonstrates the decline of pain ratings role in pain adaptation.
after days of pain stimulation. The brain mech-
anisms of adaptation at the cognitive–affective 8.2.4.2 Cognitive Appraisal of Pain
level, mainly from cognitive appraisal, are also When individuals are appraising a stimulus,
discussed. they need to actively engage with it. Therefore,
attentional processing and working memory
8.2.4.1 Sensory Adaptation of Pain may play a key role for them to focus on
In an earlier study, Bingel et al. (2007) investi- the stimuli. Such a ‘high-level appraisal’,
gated the change of brain activation towards which taps more cognitive efforts, is different
heat pain stimuli delivered on eight consecu- from the ‘low-level appraisal’, which is usu-
tive days in healthy adults. Behaviourally, they ally automatic and pre-attentive (Kalisch
found a nonlinear decline of pain intensity et al. 2006). In one study, Kalisch et al. (2006)
during this period, similar to the findings from investigated the brain mechanisms of high-
Coghill et al. (1993). fMRI was performed on level appraisal of painful stimuli. In the con-
Day 1, Day 8 and two weeks after the end of trol condition, when subjects anticpated
pain stimulation (i.e. Day 22). They found a painful electrical stimuli, they reported
decline of brain activation of the pain-related increased anxiety. In contrast, when the sub-
regions, including the thalamus, the anterior jects needed to perceive pain and concur-
insula, the secondary somatosensory cortex rently perform a challenging cognitive task
(S2) and the putamen, corresponding to the (i.e. high cognitive-load condition), anxiety-
decline of perceived pain. Notably, comparing related activity in the PFC was attenuated.
the brain activation of Day 8 with Day 1, they (Kalisch et al. 2006). A potential explanation
found an increase in brain activation in the is that the attentional and working memory
sub-genual anterior cingulate cortex (ACC), resources have been diverted to the task in the
which may suggest a role of anti-nociceptive high cognitive-load condition, and the PFC
control of pain (Bingel et al. 2007). plays a key role in high-level appraisal.
Subsequently, Riedl et al. (2011) investigated Consistent with the behavioural findings, the
the intrinsic connectivity network (ICN) of 13 researchers found an attenuated activation at
healthy adults who received heat pain stimula- the vmPFC in the high cognitive-load condi-
tion for 11 consecutive days. In order to explore tion. The finding is consistent with the results
the ICN, the subjects received resting-state from meta-analyses, which revealed the
fMRI before and after pain stimulation on Day association between vmPFC activation
1 and Day 11. Pain adaptation was identified and reduced negative emotions (Diekhof
by a nonlinear decline of pain ratings (Riedl et al. 2011). In the meta-analysis, Diekhof
et al. 2011). They found that in Day 11, com- et al. (2011) investigated the pattern of brain
pared to Day 1 (i.e. after adaptation of pain), activation of fear extinction, placebo control
there was higher functional connectivity in the and cognitive appraisal. In the studies of cog-
sensorimotor network and the ventromedial nitive appraisal, subjects reported decreased
prefrontal cortex (vmPFC). After adaptation negative affect when they watched fear-
(i.e. Day 11), the connectivity of the vmPFC eliciting scenes and showed consistent activa-
became correlated with both pain ratings and tion at the lateral prefrontal cortex (PFC) and
the connectivity of the sensorimotor network the vmPFC. Notably, the vmPFC activation
(Riedl et al. 2011). The findings support was also identified in the studies of fear
the preliminary conclusion that cognitive– extinction and placebo control (Diekhof
affective processing of pain, which is associated et al. 2011). In another neuroimaging
meta-analysis, Buhle et al. (2014) reported to recognize the shape of an intraoral test piece
that the cognitive appraisal of emotional (for assessing oral stereognostic ability, OSA)
stimuli was associated with a consistent acti- and detect the presence of a thin foil in the
vation at the lateral PFC, the dorsomedial premolar area (for assessing tactile sensibil-
PFC and the superior parietal cortex. The ity). The experience of adaptation was assessed
study also showed the potential role of the using a questionnaire. Subjects were asked to
PFC in the cognitive appraisal of pain (Buhle rate their agreement to questions such as ‘I can
et al. 2014). speak well with the new dentures’ and ‘I have
difficulties in chewing’ (Müller et al. 1995).
The authors found a significant correlation
8.2.5 Oral Sensory Adaptation
between the subjective experience of adapta-
and Brain Plasticity
tion and tactile sensibility, but not for OSA
As noted in Chapter 5, feeding behaviour is (Müller et al. 1995). In a series of studies,
associated with the integration of information Ikebe et al. (2007b) found that older CD wear-
from multiple sensory modalities, including ers showed a decreased OSA when they
somatosensation and gustation. In the follow- removed their CDs. However, the difference in
ing section, we summarize some key findings OSA was not significant between older dentate
of the adaptation of the oral sensation. subjects and subjects wearing a CD (Ikebe
et al., 2007a). The findings suggest that the
8.2.5.1 Clinical Research on Oral patients had adapted to the CD so that their
Sensory Adaptation OSA was maintained. Notably, in the patients
The sensation of the force applied on teeth via with a CD, a better OSA was associated
mechanoreceptors is key for mastication. In a with a better masticatory performance (Ikebe
series of studies, Trulsson (2006) recorded the et al. 2007b). The findings suggest that indi-
nerve signals derived from a single tooth where vidual variability in the OSA may be critical to
force was applied. They found different pat- mastication.
terns of adaptation between the force applied In comparison to oral somatosensation,
on anterior teeth and the force applied on pos- more evidence has been reported for the phe-
terior teeth. In general, both anterior and nomenon of taste adaptation, i.e. a gradual
posterior teeth show a strong saturating decline of sensation during prolonged stimula-
stimulus-response relationship (Trulsson 2006). tion of gustation. Notably, while taste adapta-
The anterior teeth showed a higher sensitivity tion has been widely reported in laboratory
to the change of force when the force was research, in our daily life, we seldom ‘lose our
applied at a lower level (e.g. <1 N). In contrast, taste’ for food when we keep on consuming it.
the posterior teeth showed a higher sensitivity Previous studies revealed that the properties of
when the force applied was at a higher level taste stimuli, such as the viscosity of the test
(e.g. ~4 N) (Trulsson 2006). The different pro- solution, were not associated with the pattern
files in the adaptation of mechanical force of taste adaptation (Theunissen et al. 2000).
between anterior and posterior teeth may Interestingly, the conditions to deliver the test
reflect their different functional roles. In con- solution have a significant effect on taste adap-
trast to mechanosensation, there has been little tation. Compared to sense the taste via filter
evidence regarding the adaptation of oral stere- paper, sipping the solution or having it flow
ognosis. In a cross-sectional study, Müller et al. over the tongue will induce less taste adapta-
(1995) investigated the experience of adapta- tion (Theunissen et al. 2000). Notably, the lat-
tion of wearing a new complete denture (CD) ter conditions are similar to our natural
in edentulous patients. The subjects were asked behaviour of eating, where subtle mouth
movement occurs. The results of taste adapta- measured by cortical responses in the sensori-
tion echoed the findings of oral mechanosen- motor cortex is associated with the change in
sation and stereognosis, showing that oral various oral conditions, including orthodontic
sensory and motor functions are both critical movement, tooth extraction, implant installa-
to adaptation to oral conditions. tion and changes in occlusion (Avivi-Arber
et al. 2015a, 2015b; Sood et al. 2015). However,
8.2.5.2 Animal Research on Brain Plasticity it should be noted that the observed brain plas-
of Orofacial Sensorimotor Functions ticity may comprise multiple aspects of infor-
In contrast to clinical research, there has been mation processing. It may not be a pure ‘sensory
cumulating evidence from animal studies that adaptation’ because orthodontic treatment and
demonstrated the potential brain mechanisms extraction will induce changes not only in sen-
underlying the plasticity of oral sensorimotor sation but also in the pattern of mastication.
functions. For example, Sood et al. (2015) used Therefore, the brain plasticity demonstrated by
intracortical microstimulation (ICMS) to meas- animal studies may be related to the adaptation
ure the features of cortical response in the sen- from both sensory and motor aspects.
sorimotor cortex of animal subjects. The
features include the number of ICMS sites and 8.2.5.3 Neuroimaging Research on Brain
the onset latencies of cortical responses. They Mechanisms of Oral Sensory Adaptation
found that orthodontic tooth movement would In contrast to animal research, at present, the
induce changes in cortical response in the face brain plasticity associated with adaptation in
area of the sensorimotor cortex when the human subjects has been less investigated.
orthodontic force was continuously applied. Functional adaptation is associated with
The effect of brain plasticity may reflect adap- behavioural alterations in facing environmen-
tive sensorimotor changes derived from altered tal challenges (Frisancho 1993). Therefore, a
oral conditions, as induced by orthodontic neuroimaging study of longitudinal design
tooth movement (Sood et al. 2015). In a series may better reflect the effect of behavioural
of studies, Avivi-Arber et al. (2015a, 2015b) alterations and brain plasticity corresponding
used the ICMS method and investigated the to the alterations. Ideally, the study should
effect of dental modification on the brain of compare the brain features before and after an
animal subjects. They found that (i) tooth adaptive behaviour, and the association
extraction induced a sustained (one to two between behavioural and brain alterations can
months) change of cortical response in the be elucidated. In oral neuroscience, several
facial area of the sensorimotor cortex (Avivi- studies have focused on the plasticity associ-
Arber et al. 2015b), and (ii) the trimming of an ated with having a new dental device
incisor (which reduced occlusal contact) (Inamochi et al. 2017; Luraschi et al. 2013).
induced cortical plasticity in the facial area of Because oral rehabilitation is associated with
the primary motor cortex (Avivi-Arber the adaptation of not only sensory but also
et al. 2015a). Furthermore, (iii) when the den- motor aspects (e.g. the pattern of mastication),
tal implant was installed one month after we discuss the neuroimaging findings of the
extraction, the above-mentioned cortical studies in Section 8.3.
changes induced by extraction were reversed At present, there has been little evidence
(Avivi-Arber et al. 2015b) and (iv) cortical plas- about the brain mechanisms of adaptation of
ticity were observed when the subjects with oral somatosensation. The lack of evidence may
trimmed incisors were restored for incisal reflect several challenges of research design for
occlusion (Avivi-Arber et al. 2015a). Together, oral neuroscience. Firstly, to identify sensory
the findings suggest that brain plasticity adaptation, a subject should receive constant
stimuli continuously. Therefore, multiple neu- ●● Animal studies demonstrated that brain
roimaging scans are needed for tracing the brain plasticity measured by cortical responses in
plasticity during consecutive days of stimula- the sensorimotor cortex is associated with
tion (e.g. the pain adaptation paradigm (Bingel the change of various oral conditions,
et al. 2007)). Such a paradigm would require including orthodontic movement, tooth
more research resources to conduct. Secondly, extraction, implant installation and changes
most oral somatosensation (e.g. mechanosensa- in occlusion.
tion and stereognosis) is perceived synergisti-
cally with tongue and jaw movement (Haggard
and de Boer 2014). As shown in the studies of 8.3 Functional Adaptation
taste adaptation, the sensation of taste also of Mastication and Swallowing
relates to tongue movement (Theunissen
et al. 2000). Therefore, it may be difficult to 8.3.1 Introduction
disentangle between these motor factors and
sensory factors. Thirdly, cumulating evidence The preceding chapter focuses on the adapta-
suggests that anxiety related to pain is closely tion of pain and oral sensory functions. Both
associated with cognitive appraisal of a noxious clinical and neuroimaging findings suggest
stimulus (e.g. Kalisch et al. 2006). The associa- that the adaptation may be associated with
tion between oral sensory adaptation and complicated cognitive–affective processing of
the relevant cognitive–affective factors (e.g. the sensory information. Moreover, evidence from
pleasantness perceived from food) has remained animal research suggests that changes in oral
unclear. conditions may be associated with brain plas-
ticity, primarily in the sensorimotor cortex. As
the major steps of feeding, mastication and
8.2.6 Summary swallowing are associated with complicated
●● The ability to maintain feeding behaviour by sensory and motor processing (see Chapter 4).
adapting to the changing environmental and Therefore, the functional adaptation of masti-
bodily conditions should be considered as cation and swallowing, which aim to improve
one of the primary functions of the stoma- the performance of feeding under environ-
tognathic system. mental challenges (e.g. tooth loss), may be
●● Functional adaptation is associated with the associated with complicated mechanisms of
adaptation from structural to behavioural sensorimotor, cognitive and affective process-
aspects. The structural and behavioural ing of oral functions. In this chapter, we dis-
changes improve one’s functional perfor- cuss recent neuroimaging evidence of brain
mance under environmental challenges. plasticity associated with the adaptation of oral
●● Sensory adaptation does not mean a ‘slug- functions. Research design of this topic is
gish’ status of our perception, but a status for discussed.
individuals to get ready for the changing
environmental conditions. In human sub-
8.3.2 Adaptation of Mastication
jects, pain adaptation to constant noxious
and Swallowing
stimuli shows a nonlinear gradual decline in
perceived intensity. As noted in Section 8.2, functional adaptation
●● Neuroimaging evidence supports that pain aims to improve functional performance in fac-
adaptation is associated with the brain ing environmental challenges (Frisancho 1993).
mechanisms of cognitive–affective process- Importantly, functional adaptation may be
ing of pain, such as the cognitive appraisal of associated with structural changes or behav-
noxious stimuli. ioural changes. In the following sections, we
mainly focus on the adaptation at the behav- swallowing (Woda et al. 2006). In other words,
ioural level, including modification of food- the behavioural adaptation of mastication is
intake behaviour and changes in masticatory represented by increasing the number of chew-
and swallowing movement. Finally, the associ- ing cycles (Peyron et al. 2017). Other common
ation between mastication and masticatory adaptive behaviours for mastication include
muscle pain, in which adaptation may play a changing different sides for chewing, chewing
key role, is discussed. with the incisor, and chewing on the gums
(Zelig et al. 2019). Notably, all the behavioural
8.3.2.1 Adaptation by Modification modifications are associated with not only
of Food Intake changes in the motor aspects (e.g. jaw move-
To cope with changes in oral conditions, such ment), but also changes in the sensory and
as tooth loss, individuals may develop different cognitive–affective aspects. For example, when
ways to modify their food-intake behaviour. chewing with a different side of teeth, one
For example, older people with a lower num- needs to pay attention to the chewing move-
ber of teeth may selectively avoid eating some ment and carefully evaluate the size of a bolus.
food. An investigation on 2224 older people in Therefore, perceptual processing of texture
Taiwan revealed that in the subjects with fewer and stereognosis would also be part of the
than 20 nature teeth or 8 functional teeth units, adaptative processing.
more than 60% of them had difficulty eating
pork and chicken (Hsu et al. 2012). Notably, 8.3.2.3 Adaptation of Swallowing Functions
selectively avoiding some foods is not an adap- Until now, most studies focus on the use of
tive strategy for improving feeding functions behavioural strategies of adaptation, such as
because it may be detrimental to nutritional food modification and postural adjustment
intake (Hung et al. 2003). To improve the eat- (Loret 2015). A recent meta-analysis revealed
ing experience, individuals may try to modify that oldest–old people presented physiological
the properties of food by making it more suit- changes in swallowing movement. Most
able to eat. For example, the subjects with aspects of the changes, such as alterations in
fewer teeth or functional units would prefer the duration and pressure of the oesophageal
steamed or sliced food to grilled or fried food sphincter, are involuntary. Alterations of the
(Hsu et al. 2012). The same strategy is applied pattern of swallowing movement, such as the
for swallowing. To reduce the risk of aspira- delay in swallowing onset, have been found in
tion, patients with dysphagia would modify the subjects (Jardine et al. 2020).
food for improving taste, olfactory, thermal,
and texture perception, which may help to
8.3.3 Pain, Adaptation and Mastication
evoke a reflex during swallowing (Loret 2015).
It is very intuitive to think about pain as some-
8.3.2.2 Adaptation of Masticatory Functions thing disruptive to normal functions. For
In addition to modifying the physical proper- example, patients with masticatory muscle
ties of food, some may modify the pattern of pain may feel difficulty in moving their jaw.
masticatory ad swallowing movement to They may adapt to such sub-optimal condi-
improve feeding. For example, while older tions to maintain normal feeding. It has
people may have a lower masticatory perfor- remained unclear how pain, adaptation and
mance, the size of bolus ready for swallowing mastication interact with each other. According
does not increase as age increases (Peyron to the vicious cycle theory, pain may modulate
et al. 2017). To compensate for the reduced (either inhibit or excite) the activity of motor
masticatory performance due to healthy age- neurons at the CNS level and thus leads to
ing, older people may chew more times before muscle hyperactivity (Peck et al. 2008). The
theory predicts the existence of a pain–spasm– cerebellum and the trigeminal motor nucleus
pain cycle, in which ‘pain may further cause was associated with a higher pain catastro-
muscle spasm and that muscular activity can phizing score. In contrast, changes in the
be painful’ (Roland 1986). According to the brain activation of the dorsolateral PFC
Pain Adaptation Model, pain may lead to decreased as the pain catastrophizing score
changes in muscle activity. However, this increased, and the association was not signifi-
change has a protective role in movement: it cant when subjects were not in pain
restricts movement and therefore avoids fur- (Henderson et al. 2016). The findings suggest
ther injury from movement (Lund et al. 1991). that catastrophic thoughts may play a key role
Therefore, the framework predicts that corre- in the integration between pain and sensori-
sponding to the restriction of a movement, the motor processing.
agonist muscles of a movement would show
decreased activity, and the antagonist muscles
8.3.4 Neuroimaging Findings
would show increased activity (Peck
of Adaptation of Mastication
et al. 2008). By revising the original pain adap-
and Swallowing
tation model, Murray and Peck (2007) pro-
posed the Integrated Pain Adaptation Model Table 8.1 summarizes the recent neuroimaging
(IPAM), which highlights the complexity of findings of adaptation of mastication and swal-
individual variations in pain symptoms. In lowing. The studies can be broadly classified
general, pain will interact with the sensori- into two sets. The first set includes clinical
motor system, which plays a key role in the research, which focuses on the improvement
individual differences in pain experience. of performance in patients, such as changes in
Specifically, among the sensorimotor system, swallowing functions in patients with dyspha-
the primary motor area may play a critical gia. The second set includes studies on healthy
role since its activity can be depressed by nox- subjects, which focuses on the effect of train-
ious stimuli, as revealed by animal research ing on oral sensorimotor functions. The results
(Murray and Peck 2007). In human subjects, from recent neuroimaging studies are dis-
when being injected with nerve growth factor cussed in the following sections.
into the masseter, subjects showed an
increased jaw pain and jaw functional disabil- 8.3.4.1 Brain Mechanisms of Adaptation
ity as well as a decreased motor-evoked poten- of Mastication
tial amplitude and reduced corticomotor map As noted in the preceding section, the study
area and volume (Costa et al. 2019). The IPAM with a longitudinal design will better reflect
can be deemed as a generalized framework of the behavioural and brain alterations associ-
the original pain adaptation model, which ated with adaptation. However, at present,
focuses on not only the relationship between only a few neuroimaging studies have provided
pain and motor neurons but also the control longitudinal evidence for brain plasticity of
of the sensorimotor system. Notably, the asso- oral functions. In an earlier study, Luraschi
ciation between pain and mastication may be et al. (2013) investigated functional adaptation
modulated by cognitive–affective factors. As to wearing of a new CD in edentulous patients.
mentioned previously, individuals with Firstly, they found an improvement in mastica-
greater pain catastrophizing tend to form tory performance and biting force three
more catastrophic thoughts about their pain. months after subjects wore the denture. The
In a recent fMRI study, Henderson et al. (2016) improved masticatory ability suggests func-
found that when healthy pain-free adults were tional adaptation to the denture. Secondly, by
performing jaw movement in the presence of comparing the brain activation during oral
pain, increased brain activation of the M1, the tasks, they identified increased activation in
Table 8.1 Neuroimaging research on brain mechanisms of brain plasticity and oral functions (since 2010).
Huang et al. (2018) 17 patients with subacute rs-fMRI (before and after swallowing therapies)
stroke and dysphagia
Inamochi et al. (2017) 28 healthy adults fMRI (pre-insertion, Day 1, Day 3 and Day 7)/chewing
with experimental denture-based palatal plates
Avivi-Arber et al. (2017) 67 adult mice sMRI (post-mortem)/extraction of molar teeth
Lee et al. (2016) 17 patients of Bell’s palsy with fMRI/finger and orofacial movements
complete clinical recovery
Michou et al. (2015) 11 healthy adults receiving fMRI and MRS/saliva and water swallowing tasks
paired associative stimulation
Luraschi et al. (2013) 10 complete denture wearers fMRI (pre-insertion [T0], insertion [T1], 1 wk [T2]
and 3 mo postinsertion [T3])/wearing a complete
denture
Suntrup et al. (2013) 21 healthy adults MEG (before and after treatment)/anodal
transcranial direct current stimulation over the left
or right swallowing motor cortex
Garmi et al. (2013) 10 patients who were operated fMRI before and after surgery/face smile and
on lengthening temporalis clenching tasks
myoplasty
Arima et al. (2011) 13 healthy adults fMRI (before and 1 h, 1 d and 1 wk after tongue-
protrusion training)/tongue protrusion
Notes: fMRI: functional magnetic resonance imaging; MEG: magnetoencephalography; MRS: magnetic
resonance spectroscopy; rs-fMRI: resting-state functional magnetic resonance imaging; sMRI: structural
magnetic resonance imaging.
the bilateral S1 and M1 (Luraschi et al. 2013). with facial palsy. In a task-based fMRI study,
The findings correspond to the results of ani- Lee et al. (2016) reported that patients who
mal research, which suggest that adaptation recovered from Bell’s palsy showed a lower
to oral conditions is associated with the plastic- activation when they performed lip pursing
ity of the sensorimotor cortex (Avivi-Arber movements, compared to healthy controls, in
et al. 2011). Notably, during a jaw-clenching the S1 and the cingulate motor area. In a longi-
task, increased S1 activation was only found tudinal study, Garmi et al. (2013) reported an
one week after denture installation. Three altered activation in the sensorimotor area,
months after installation, the blood‑oxygen- before and after surgical intervention, in
level-dependent (BOLD) activation was nor- patients with facial paralysis.
malized to the condition before the installation In a recent study, Inamochi et al. (2017)
of a new denture. In contrast, behavioural investigated 28 healthy adults who had a
alterations, i.e. improvement of masticatory denture-base plate installed in the palate for
performance, showed a gradual increase from seven days. It is noteworthy that the base plate
immediate post-installation to three months was designed, for an experimental purpose, to
after installation. The study provides prelimi- disturb sensory feedback from the oral cavity.
nary findings of the brain mechanisms of Right after installing the plate, subjects showed
functional adaptation of wearing a dental pros- decreased masticatory performance (oral mix-
thesis. The plasticity at the sensorimotor area ing ability). The ability gradually recovered by
was also reported in clinical studies of patients Day 7. During a chewing task, subjects showed
brain activation of the right sensorimotor area, electrical stimuli and transcranial magnetic
and the activation gradually returned to the stimulation (TMS), on healthy adults. TMS
initial level, a pattern similar to the change of was applied to the pharyngeal representation
masticatory performance. In contrast, the brain of the motor cortex. fMRI investigation
activation of the ACC and the mPFC showed a revealed that after PAS intervention, subjects
reduction, but did not return to the initial level showed increased activation during water and
(Inamochi et al. 2017). The findings suggest saliva swallowing. Magnetic resonance spec-
that beyond the S1 and M1, other brain regions troscopy (MRS) investigation revealed that
associated with cognitive–affective processing the change in brain activation was associated
may show altered activation corresponding to with the change in the level of the inhibitory
the adaptation of oral functions. Notably, dif- neurotransmitter γ-aminobutyric acid (GABA;
ferent regions presented different patterns of Michou et al. 2015). Suntrup et al. (2013)
activation. Consistently, recent animal research adopted transcranial direct current stimula-
has demonstrated that after tooth extraction, tion (TDCS) to stimulate the motor cortex
there is a widespread change in GMV in brain associated with swallowing in healthy adults.
regions associated with cognitive–affective They found that after TDCS, subjects showed
processing, not confined to the sensorimotor an improvement in swallow task performance.
cortex (Avivi-Arber et al. 2017). The findings Magnetoencephalography (MEG) investiga-
support the preliminary conclusion that adap- tion also revealed increased activation of the
tation may be associated with plasticity beyond cortical network of swallowing (Suntrup
the sensorimotor cortex, and the different spa- et al. 2013). Because only healthy adults were
tial and temporal patterns of brain plasticity investigated in the studies (Michou et al. 2015;
require more investigation. Arima et al. 2011; Suntrup et al. 2013), the
results did not provide direct evidence about
8.3.4.2 Brain Mechanisms of Adaptation the brain mechanisms of functional adapta-
of Swallowing tion of swallowing dysfunction. Nevertheless,
In an earlier study, Arima et al. (2011) investi- the findings consistently suggest that the mod-
gated brain plasticity associated with the train- ification of swallowing performance can be
ing of tongue movement in healthy younger associated with the plasticity of functional
adults. fMRI was performed before and one brain features. In a recent study, Huang et al.
hour, one day and one week after training. (2018) found that in patients with dysphagia
They found increased activation in the M1 one due to stroke, swallowing therapies would
hour after training, and a more extensive acti- improve their performance in food intake.
vation in the M1, the SMA, the putamen and Notably, the change in intrinsic functional
the cerebellum one day after the training connectivity was also identified after vs. before
(Arima et al. 2011). In contrast, decreased acti- the therapies (Huang et al. 2018). Together, the
vation was found in the S1 and middle frontal findings suggest that improvement in swallow-
gyri, both one hour and one week after train- ing performance may be associated with vari-
ing (Arima et al. 2011). The findings suggest ous aspects of functional features, including
that alterations in the pattern of tongue move- brain activation, intrinsic functional connec-
ment, which is pivotal to both mastication and tivity and the level of neurotransmitters.
swallowing, are associated with changes in
brain activation with complicated patterns at
8.3.5 Summary
different temporal scales.
In another study, Michou et al. (2015) inves- ●● To cope with changes in oral conditions,
tigated the effect of paired associative stimula- such as tooth loss, individuals may develop
tion (PAS), which combined pharyngeal different ways to modify their food-intake
behaviour. Individuals may avoid the food approach’ for improving oral functions. In
that they feel difficult to eat. Also, they may some studies, improvement of oral functions
try to modify the properties of some foods by was achieved via direct stimulation of the
different ways of cooking to make them brain. Both TMS and TDCS are used to stimu-
more suitable for eating. late the motor cortex associated with mastica-
●● Adaptation of oral functions may be associ- tory and swallowing movements. We
ated with changes in masticatory and considered these methods as the ‘brain
swallowing movement. For example, to com- approach’ for improving oral functions. In the
pensate for the reduced masticatory perfor- following sections, we discuss recent neuroim-
mance due to healthy ageing, older people aging findings of the mechanisms of both oral
may chew more times before swallow food. and brain approaches for improving oral
●● Adaptation may play a key role in the asso- functions.
ciation between mastication and mastica-
tory muscle pain. Pain may lead to changes
8.4.2 Oral Approaches for Improving Oral
in muscle activity, which plays a protective
Functions in Older People
role in restricting jaw movement and avoid
further injury from the movement. Improving masticatory functions of older peo-
●● Recent neuroimaging findings support the ple has been an important clinical issue in
preliminary conclusion that adaptation may recent years due to the challenge of popula-
be associated with the plasticity beyond the tional ageing (Baram et al. 2020; Kim
sensorimotor cortex, and the different tem- et al. 2019). In gerodontology, there is a strong
poral patterns of the regional plasticity demand for developing oral exercises that are
require more investigation. simple and effective for older people to main-
●● Recent neuroimaging findings suggest that tain oral functions. The exercises may com-
improvement in swallowing performance prise multiple procedures of manipulating the
may be associated with various aspects of stomatognathic apparatus (e.g. the tongue, the
functional features, including brain activa- jaw and the lips) (e.g. Kim et al. 2019).
tion, intrinsic functional connectivity and However, until now, no consensus has been
the level of neurotransmitters. reached for a standard protocol for these exer-
cises. In the meantime, the effectiveness of the
training protocols on masticatory and swallow-
8.4 Brain Plasticity Associated ing functions is still inconclusive due to a great
methodological variability across different
with Oral Functional Training
exercises. For example, the protocol adopted
by Hakuta et al. (2009) consists of (i) an exer-
8.4.1 Introduction
cise of facial muscles by pronouncing vowels,
In this section, we focus on the association (ii) a tongue exercise by moving the tongue
between brain plasticity, functional adaptation directionally and (iii) the massage of major
and the approaches to improve oral functions. salivary glands. Subjects also checked for their
Clinically, a variety of exercises have been oral functions, e.g. to distinguish the taste with
adopted for improving oral functions, such as the tongue and to check the change in saliva
swallowing manoeuvres that alter the sensory secretion (Hakuta et al. 2009). In another pro-
experience and movement of swallowing tocol, subjects perform deep breathing, neck
(Bahia and Lowell 2020). Some methods focus stretching and exercises on lips/cheeks/tongue
on changing the properties of food for better with specific movements, such as inflation,
mastication and swallowing. In this chapter, protrusion and closure. Massage of salivary
we considered these methods as the ‘oral glands was also included in this protocol
(Ohara et al. 2015). In some protocols, oral manoeuvre, the tongue-holding exercise
exercises are performed between a warming up (Masako manoeuvre), and the head-lift
and a relaxing session, including deep respira- (Shaker) exercise. A recent meta-analysis
tion and neck and shoulder exercise. The oral revealed that effortful swallowing would gen-
exercise consists of the stretching movement erate greater tongue-to-palate pressure and
of the lips and the tongue, the exercise of mas- also pressures of the pharynx, upper oesopha-
ticatory muscles, the speaking exercise of sim- geal sphincter and oesophageal regions (Bahia
ple syllabi (e.g. ‘pa’, ‘ta’ and ‘ka’), and and Lowell 2020). However, the protocols for
swallowing exercise. The whole programme this exercise have not been standardized
lasts for approximately two hours per visit (Bahia and Lowell 2020). Meta-analytic find-
(Cho et al. 2012). The number of exercise items ings have revealed that the Mendelsohn
was further reduced in a simplified protocol, manoeuvre, which facilitates swallowing by
which includes only the exercises of the lips, prolonging the duration of upper oesophageal
the tongue, the cheek, masticatory muscles sphincter opening, was associated with
and salivation (Kim et al. 2019). The simplified increased duration and amplitude of pressure,
protocol improved oral functions for elderly displacements and muscle activity of the
people with poor mastication, swallowing and sphincter (Wheeler-Hegland et al. 2009). The
salivary functions, and the improvement Masako manoeuvre, which requires subjects to
would maintain for one week after training practice holding their tongues between teeth
(Kim et al. 2019). Notably, the exercises can be during swallowing, improved swallowing by
assisted by newly developed instruments so increasing the function of the superior pharyn-
that the training will be standardized. For geal constrictor. The head-lift technique,
example, when subjects practiced jaw opening, which requires subjects to rest in a supine posi-
a pre-formed intraoral device can be held in tion with their heads lifting to watch their toes,
between their front teeth with constant pres- was associated with an increased anteroposte-
sure for training the activity of lip and cheek rior opening of the upper oesophageal sphinc-
muscles (Baram et al. 2020). Chewing gum, ter (Antunes and Lunet 2012).
which can be standardized in shape and tex-
ture, is used for the exercise of masticatory
8.4.4 Neuroimaging Research on Brain
muscles (Baram et al. 2020). However, until
Plasticity of Improving Oral Functions
now, a report of systematic comparison of the
treatment effects between different exercises is Table 8.2 summarizes recent neuroimaging
still lacking. findings of brain plasticity associated with
changes in swallowing via oral approaches and
brain approaches, in which oral functions are
8.4.3 Oral Approaches for Functional
affected by direct stimulation of the brain.
Adaptation in Dysphagia
Some of the studies include healthy subjects
There has been a large body of publications only. Therefore, the findings may not reflect
about improving swallowing function because the brain mechanisms of functional adapta-
dysphagia is commonly co-morbid with neu- tion to diseases. Still, the findings provide a
rological disorders, including stroke and general picture of the brain mechanisms asso-
Parkinson’s disease, and head and neck cancer ciated with the improvement of oral functions.
(Foley et al. 2008). Several physical manoeu-
vres for improving swallowing have been 8.4.4.1 The Effect of Oral Approaches
developed and widely used for years, primar- on Swallowing
ily for patients with dysphagia. These tech- Suntrup et al. (2015) investigated the use of
niques include effortful swallow, Mendelsohn pharyngeal electrical stimulation (PES). They
Notes: EEG: electroencephalography; fMRI: functional magnetic resonance imaging; fNIRS: functional near-
infrared spectroscopy; MEG: magnetoencephalography; MRS: magnetic resonance spectroscopy; PET: positron
emission tomography; rs-fMRI: resting-state functional magnetic resonance imaging; rTMS: repetitive transcranial
magnetic stimulation; TBS: theta burst stimulation; TDCS: transcranial direct current stimulation.
Source: See also Table 4.3, a part of its data is summarized here.
found that in healthy subjects PES increased 8.4.4.2 The Effect of Brain Stimulation
swallowing capacity. Moreover, MEG revealed on Swallowing
reduced event-related desynchronization of A series of studies investigated the effect of
cortical oscillatory activity in the sensorimotor theta burst stimulation (TBS), an approach of
area (Suntrup et al. 2015). Using electroen- repetitive transcranial magnetic stimulation
cephalography (EEG), Jestrović et al. (2015) (rTMS), applied over the motor cortex related
investigated the brain mechanisms associated to the movement of suprahyoid muscles (Lin
with fluid viscosity and head posture in healthy et al. 2017; Ruan et al. 2017, 2019; Zhang
adults. They found that altered viscosity (e.g. et al. 2019, 2020). The authors found that TBS
water vs. honey-thick fluid) and posture (e.g. induced suprahyoid motor cortex excitability
chin-tuck position) was associated with the dif- that lasted at least 30 minutes (Lin et al. 2017).
ference of brain network at multiple frequency Moreover, the effect was associated with
bands (Jestrović et al. 2015, 2016). The changes changes in intrinsic functional connectivity,
in EEG activity were also associated with atten- including the regional homogeneity and the
tion deployment to swallowing and the volume amplitude of low-frequency fluctuation of
of bolus (Jestrović et al. 2018). However, the brain regions (Ruan et al. 2017, 2019), and
EEG findings do not reveal an effect of brain functional connectivity and degree centrality
plasticity owing to the cross-sectional design of of the functional network (Zhang
the studies. Using a neurofeedback approach, et al. 2019, 2020). In addition to the motor
Kober et al. (2015) investigated if training of cortex, rTMS applied over the cerebellum
motor imagery, i.e. to imagine swallowing also induced an increased cortico-pharyngeal
without an overt movement, would alter corti- motor evoked potential. The effect lasted for
cal activation associated with the execution 30 minutes after the intervention of rTMS
and imagery of swallowing. During training, (Vasant et al. 2015). In addition to TBS, a
healthy subjects practised increasing the recent randomized control trial revealed that
hemodynamic response associated with swal- in patients with dysphagia following stroke,
lowing via motor imagery. They found that the tDCS led to a significant improvement in
cortical activation of motor execution and their swallowing function compared to a
imagery was more pronounced after the train- sham treatment. MEG investigation revealed
ing (Kober et al. 2015). In a subsequent real- that the improvement was associated with
time fMRI study, they found that during alterations in the swallowing network
neurofeedback training, subjects increased the (Suntrup-Krueger et al. 2018). Furthermore,
activity in extensive regions associated with a recent MEG study investigated the effect of
swallowing, not confined to the motor cortex the PES (an oral approach of peripheral stim-
(Kober et al. 2019). The findings suggest that ulation) and the TDCS (a brain approach of
either physical methods (e.g. electrical stimu- central stimulation) on healthy adults with
lation) or behavioural methods (e.g. postural pharyngolaryngeal hypesthesia, which was
adjustment and practice of motor imagery) are experimentally induced by injection of lido-
associated with alteration in functional brain caine. The PES, rather than the TDCS, led to
features. Notably, a recent MEG study revealed improved swallowing function and associ-
that, in healthy subjects, stimulating the oro- ated alterations in brain activity (Muhle
pharynx with capsaicin improved the perfor- et al. 2020). Together, the findings suggest
mance of a swallowing task, but no significant that the brain approach may be a potential
change in brain activity was found with the method for improving swallowing function.
intervention (Suntrup-Krueger et al. 2021). A However, the association between behav-
consistent pattern of brain plasticity induced ioural improvement and brain plasticity has
by these methods has remained unclear. remained unclear.
References
Antunes, E.B. and Lunet, N. (2012). Effects of the following tooth loss in female mice. Front.
head lift exercise on the swallow function: a Neuroanat. 10: 121.
systematic review. Gerodontology 29: 247–257. Bahia, M.M. and Lowell, S.Y. (2020). A
Arima, T., Yanagi, Y., Niddam, D.M. et al. (2011). systematic review of the physiological effects
Corticomotor plasticity induced by tongue- of the effortful swallow maneuver in adults
task training in humans: a longitudinal fMRI with Normal and disordered swallowing. Am.
study. Exp. Brain Res. 212: 199–212. J. Speech Lang. Pathol. 29: 1655–1673.
Avivi-Arber, L., Martin, R., Lee, J.C., and Sessle, Baram, S., Karlsborg, M., and Bakke, M. (2020).
B.J. (2011). Face sensorimotor cortex and its Improvement of oral function and hygiene in
neuroplasticity related to orofacial sensorimotor Parkinson’s disease: a randomised controlled
functions. Arch. Oral Biol. 56: 1440–1465. clinical trial. J. Oral Rehabil. 47: 370–376.
Avivi-Arber, L., Lee, J.C., and Sessle, B.J. (2015a). Bingel, U., Schoell, E., Herken, W. et al. (2007).
Dental occlusal changes induce motor cortex Habituation to painful stimulation involves
neuroplasticity. J. Dent. Res. 94: 1757–1764. the antinociceptive system. Pain 131: 21–30.
Avivi-Arber, L., Lee, J.C., Sood, M. et al. (2015b). Blumenfeld-Katzir, T., Pasternak, O., Dagan, M.,
Long-term neuroplasticity of the face primary and Assaf, Y. (2011). Diffusion MRI of
motor cortex and adjacent somatosensory structural brain plasticity induced by a learning
cortex induced by tooth loss can be reversed and memory task. PLoS One 6: e20678.
following dental implant replacement in rats. Buhle, J.T., Silvers, J.A., Wager, T.D. et al. (2014).
J. Comp. Neurol. 523: 2372–2389. Cognitive reappraisal of emotion: a meta-
Avivi-Arber, L., Seltzer, Z., Friedel, M. et al. analysis of human neuroimaging studies.
(2017). Widespread volumetric brain changes Cereb. Cortex 24: 2981–2990.
Cabeza, R., Albert, M., Belleville, S. et al. (2018). Gazzaniga, M.S., Ivry, R.B., and Mangun,
Maintenance, reserve and compensation: the G.R. (2019). Cognitive Neuroscience: The
cognitive neuroscience of healthy ageing. Nat. Biology of the Mind. W. W. Norton & Company.
Rev. Neurosci. 19: 701–710. Gregory, S., Long, J.D., Tabrizi, S.J., and Rees,
Chang, Y. (2014). Reorganization and plastic G. (2017). Measuring compensation in
changes of the human brain associated with neurodegeneration using MRI. Curr. Opin.
skill learning and expertise. Front. Hum. Neurol. 30: 380–387.
Neurosci. 8: 35. Haggard, P. and De Boer, L. (2014). Oral
Cho, E.P., Hwang, S.J., Clovis, J.B. et al. (2012). somatosensory awareness. Neurosci. Biobehav.
Enhancing the quality of life in elderly Rev. 47: 469–484.
women through a programme to improve the Hakuta, C., Mori, C., Ueno, M. et al. (2009).
condition of salivary hypofunction. Evaluation of an oral function promotion
Gerodontology 29: e972–e980. programme for the independent elderly in
Clark, C.W. (1991). Modeling behavioral Japan. Gerodontology 26: 250–258.
adaptations. Behav. Brain Sci. 14: 85–93. Hänggi, J., Koeneke, S., Bezzola, L., and Jäncke,
Coghill, R.C., Mayer, D.J., and Price, D.D. (1993). L. (2010). Structural neuroplasticity in the
Wide dynamic range but not nociceptive- sensorimotor network of professional female
specific neurons encode multidimensional ballet dancers. Hum. Brain Mapp. 31:
features of prolonged repetitive heat pain. 1196–1206.
J Neurophysiol 69: 703–716. Hebb, D.O. (1949). The Organization of Behavior:
Costa, Y.M., Exposto, F.G., Kothari, M. et al. A Neuropsychological Theory. J. Wiley;
(2019). Masseter corticomotor excitability is Chapman & Hall.
decreased after intramuscular administration Henderson, L.A., Akhter, R., Youssef, A.M. et al.
of nerve growth factor. Eur. J. Pain 23: (2016). The effects of catastrophizing on
1619–1630. central motor activity. Eur. J. Pain 20: 639–651.
Diekhof, E.K., Geier, K., Falkai, P., and Gruber, Herholz, S.C., Coffey, E.B., Pantev, C., and
O. (2011). Fear is only as deep as the mind Zatorre, R.J. (2016). Dissociation of neural
allows: a coordinate-based meta-analysis of networks for predisposition and for training-
neuroimaging studies on the regulation of related plasticity in auditory-motor learning.
negative affect. NeuroImage 58: 275–285. Cereb. Cortex 26: 3125–3134.
Draganski, B., Gaser, C., Busch, V. et al. (2004). Hsu, K.J., Lee, H.E., Lan, S.J. et al. (2012).
Neuroplasticity: changes in grey matter Evaluation of a self-assessed screening test for
induced by training. Nature 427: 311–312. masticatory ability of Taiwanese older adults.
Foley, N., Teasell, R., Salter, K. et al. (2008). Gerodontology 29: e1113–e1120.
Dysphagia treatment post stroke: a systematic Huang, Y.C., Hsu, T.W., Leong, C.P. et al. (2018).
review of randomised controlled trials. Age Clinical effects and differences in neural
Ageing 37: 258–264. function connectivity revealed by MRI in
Frisancho, A.R. (1993). Human Adaptation and subacute hemispheric and brainstem
Accommodation. University of Michigan Press. infarction patients with dysphagia after
Fu, M. and Zuo, Y. (2011). Experience-dependent swallowing therapy. Front. Neurosci. 12: 488.
structural plasticity in the cortex. Trends Hung, H.C., Willett, W., Ascherio, A. et al.
Neurosci. 34: 177–187. (2003). Tooth loss and dietary intake. J. Am.
Garmi, R., Labbé, D., Coskun, O. et al. (2013). Dent. Assoc. 134: 1185–1192.
Lengthening temporalis myoplasty and brain IASP. 2020. IASP Terminology [Online].
plasticity: a functional magnetic resonance Available: https://www.iasp-pain.org/
imaging study. Ann. Chir. Plast. Esthet. 58: terminology?navItemNumber=576
271–276. [Accessed].
Ikebe, K., Amemiya, M., Morii, K. et al. (2007a). Kober, S.E., Grossinger, D., and Wood, G. (2019).
Comparison of oral stereognosis in relation to Effects of motor imagery and visual
age and the use of complete dentures. J. Oral neurofeedback on activation in the
Rehabil. 34: 345–350. swallowing network: a real-time fMRI study.
Ikebe, K., Amemiya, M., Morii, K. et al. (2007b). Dysphagia 34: 879–895.
Association between oral stereognostic ability Lee, J., Yang, J., Li, C. et al. (2016). Cortical
and masticatory performance in aged reorganization in patients recovered from
complete denture wearers. Int. J. Prosthodont. Bell’s palsy: an orofacial and finger
20: 245–250. movements task-state fMRI study. Neural
Inamochi, Y., Fueki, K., Usui, N. et al. (2017). Plast. 2016: 8231726.
Adaptive change in chewing-related brain Lewis, C.M., Baldassarre, A., Committeri, G. et al.
activity while wearing a palatal plate: an (2009). Learning sculpts the spontaneous
functional magnetic resonance imaging study. activity of the resting human brain. Proc. Natl.
J. Oral Rehabil. 44: 770–778. Acad. Sci. U. S. A. 106: 17558–17563.
Jardine, M., Miles, A., and Allen, J. (2020). A Lin, C.S. (2013). Pain catastrophizing in dental
systematic review of physiological changes in patients: implications for treatment
swallowing in the oldest old. Dysphagia 35: management. J. Am. Dent. Assoc. 144:
509–532. 1244–1251.
Jestrović, I., Coyle, J.L., and Sejdić, E. (2015). Lin, C.S., Niddam, D.M., Hsu, M.L., and Hsieh,
Characterizing functional connectivity J.C. (2013). Pain catastrophizing is associated
patterns during saliva swallows in different with dental pain in a stressful context. J. Dent.
head positions. J. Neuroeng. Rehabil. 12: 61. Res. 92: 130–135.
Jestrović, I., Coyle, J.L., Perera, S., and Sejdić, Lin, T., Jiang, L., Dou, Z. et al. (2017). Effects of
E. (2016). Functional connectivity patterns of theta burst stimulation on suprahyoid motor
normal human swallowing: difference among cortex excitability in healthy subjects. Brain
various viscosity swallows in normal and Stimul. 10: 91–98.
chin-tuck head positions. Brain Res. 1652: Loret, C. (2015). Using sensory properties of food
158–169. to trigger swallowing: a review. Crit. Rev. Food
Jestrović, I., Coyle, J.L., Perera, S., and Sejdić, Sci. Nutr. 55: 140–145.
E. (2018). Influence of attention and bolus Lund, J.P., Donga, R., Widmer, C.G., and Stohler,
volume on brain organization during C.S. (1991). The pain-adaptation model: a
swallowing. Brain Struct. Funct. 223: 955–964. discussion of the relationship between
Kalisch, R., Wiech, K., Critchley, H.D., and chronic musculoskeletal pain and motor
Dolan, R.J. (2006). Levels of appraisal: a activity. Can. J. Physiol. Pharmacol. 69:
medial prefrontal role in high-level appraisal 683–694.
of emotional material. NeuroImage 30: Luraschi, J., Korgaonkar, M.S., Whittle, T. et al.
1458–1466. (2013). Neuroplasticity in the adaptation to
Kim, H.J., Lee, J.Y., Lee, E.S. et al. (2019). prosthodontic treatment. J. Orofac. Pain 27:
Improvements in oral functions of elderly 206–216.
after simple oral exercise. Clin. Interv. Aging Maguire, E.A., Gadian, D.G., Johnsrude,
14: 915–924. I.S. et al. (2000). Navigation-related structural
Kober, S.E., Gressenberger, B., Kurzmann, change in the hippocampi of taxi drivers. Proc.
J. et al. (2015). Voluntary modulation of Natl. Acad. Sci. U. S. A. 97: 4398–4403.
hemodynamic responses in swallowing Michou, E., Williams, S., Vidyasagar, R. et al.
related motor areas: a near-infrared (2015). fMRI and MRS measures of
spectroscopy-based neurofeedback study. neuroplasticity in the pharyngeal motor
PLoS One 10: e0143314. cortex. NeuroImage 117: 1–10.
Zhang, G., Ruan, X., Li, Y. et al. (2019). Zhang, G., Gao, C., Ruan, X. et al. (2020).
Intermittent theta-burst stimulation reverses Intermittent theta-burst stimulation over the
the after-effects of contralateral virtual lesion suprahyoid muscles motor cortex facilitates
on the suprahyoid muscle cortex: evidence increased degree centrality in healthy subjects.
from dynamic functional connectivity Front. Hum. Neurosci. 14: 200.
analysis. Front. Neurosci. 13: 309.
Dental Neuroimaging: The Role of the Brain in Oral Functions, First Edition. Chia-Shu Lin.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
Companion website: www.wiley.com/go/lin/dental-neuroimaging
Patient B and Patient C are both office workers, be made according to the current status of
which demand a great degree of mental effort. individual health, and this ‘current status’
However, Patient C suffers from MCI at a includes not only the oral conditions but also
younger age. The effect of such an early decline the general physical and mental conditions of
in cognitive abilities on oral health and feed patients. The role of physical and mental
behaviour should be carefully monitored in the conditions is highlighted especially for older
long run. In sum, to predict the trajectory of people because better physical and mental
changes in individual oral functions, one needs conditions suggest a better ‘reserve’ that helps
to consider not only the current oral status but them to cope with the challenges caused by
also the brain-behavioural factors related to ageing (or age-related diseases) (Cabeza
feeding behaviour. Both the aspects are consid- et al. 2018).
ered as the critical elements in the BSIA. In the Individual differences in brain reserve and
following sections, we further elaborate on two cognitive reserve may play a key role in their
major aims for the BSIA in clinical oral health- susceptibility to diseases (Barulli and
care: prediction and classification. Stern 2013) (Figure 9.1a). For example, Patient
C in our previous case has been diagnosed
9.1.2.1 Prediction of Long-term Changes with MCI. Even the patient has good oral
in Oral Functions health at present, the deficits in brain and
Dentists need to evaluate not only patients’ behaviour, which may be associated with MCI,
current intraoral conditions, but also how their may impair the patient’s ability to maintain
oral conditions may change in the near future. oral health in the future. For example, if MCI
Predicting treatment outcomes is important progresses into dementia, chronic periodonti-
for periodontal, orthodontic and prosthetic tis may be more likely to develop because the
therapies, which need to be followed up and patient becomes gradually unable to maintain
evaluated for months or years. Therefore, one oral care. Therefore, the BSIA assesses indi-
of the primary aims of the BSIA is to predict vidual brain-behaviour factors that reflect the
how individual oral function and feeding status of brain and cognitive reserve, which
behaviour will change in the long run. would be helpful for predicting the change of
According to the BSIA, the prediction should individual oral function and feeding behaviour.
Figure 9.1 Major aims of brain–stomatognathic integrative assessment (BSIA). (a) Prediction of long-term
changes in oral functions. Individual differences in brain reserve and cognitive reserve may play a key role
in their susceptibility to diseases. The BSIA, which includes the assessment of cognitive functions of older
people, would help to predict the future condition of individual oral health. (b) Classification of patients
with different risks of oral diseases. The BSIA helps to classify the patients for their risk of oral diseases and
the prognosis of treatment, based on a full-scale assessment of general physical and mental status.
9.1.2.2 Classification of Patients with Different strength) and uses to taking a variety of foods.
Risks of Oral Diseases The focus of the BSIA also echoes recent
In addition to predicting the progression of indi- advancements in precision medicine, which
vidual oral health, classifying oral health condi- highlights that the strategies of disease preven-
tions between patients is another key domain in tion and treatment should take individual varia-
clinical practice. From the point of preventive bility into account (Collins and Varmus 2015). In
dentistry, it is important to classify patients with order to achieve the ‘precision’ in diagnosis and
a high risk of some diseases before serious treatment, the psychosocial aspects of patient
symptoms develop (Figure 9.1b). As noted in the behaviour, such as the interaction between life-
previous section, in older people, the suscepti- style and environment would play a key role.
bility of functional impairment is associated This is especially critical to feeding behaviour,
with their reserve in physical and mental condi- which is highly associated with both the envi-
tions (Barulli and Stern 2013; Cabeza et al. 2018). ronmental (e.g. the way of food processing) and
Therefore, one of the major aims of the BSIA is personal (e.g. masticatory functions) factors.
to assess these physical and mental factors, and According to the BSIA framework, whether or
the individual risk of having poor oral health can not patients can adapt themselves to environ-
be estimated according to these factors. As mental and bodily changes and maintain the
shown in our hypothetical cases, the BSIA may adequate ability of feeding is a critical factor for
help prosthodontists to classify Patient A and evaluating their susceptibility to oral diseases.
Patient B for their ability of food intake after
installing a denture. Upon missing teeth, Patient
9.1.3 Components of the BSIA
B, who is weaker in physical fitness and eats
energy bars only, may have more difficulty in The BSIA framework proposed here consists of
maintaining normal feeding compared to assessments for three domains (Table 9.1). (i)
Patient A, who has a better physical reserve (e.g. Assessment of the structural integrity of the
Venue
stomatognathic system, which includes the 9.1.4 How Does the BSIA Work? Some
‘conventional’ items of an oral examination, Practical Issues for Consideration
such as examination of teeth, periodon-
A critical challenge for the BSIA framework is its
tal conditions, oral mucosa and the tempo-
multidisciplinary nature. Because it consists of
romandibular joint. (ii) Assessment of
assessments from different fields, patients may
stomatognathic functions, which includes
need to visit different doctors for different assess-
‘functional tests’ that evaluate the perfor-
ments, which becomes very time-consuming
mance of individual oral functions, including
and increases patients’ financial burden.
swallowing, mastication and speech. Notably,
Moreover, for the residents living in long-term
nutritional status should also be included.
care institutes, particularly older people and
(iii) Assessment of the ‘reserves’ for patients
patients with special needs, it would be practi-
to functionally adapt to oral diseases. The
cally impossible for them to have multiple medi-
assessment of ‘reserves’ focuses on the brain-
cal/dental visits just for the assessments. In the
behavioural factors, which reflect patients’
following sections, we discuss these challenges
abilities to cope with the challenge from dis-
and propose potential solutions to the challenges.
eases (e.g. tooth loss or periodontitis) so that
normal feeding can be maintained. The 9.1.4.1 Integration of Assessment Outcomes
‘reserves’ to be assessed include biological It should be noted that the BSIA is a multidis-
features, such as the cortical thickness of the ciplinary approach. Ideally, the multidiscipli-
brain and the general status of body composi- nary investigation can be undertaken in a
tion (e.g. muscle mass and strength), and hospital, where different departments are in
behavioural features, such as one’s general charge of different assessment items. For
cognitive abilities (Barulli and Stern 2013; example, dentists and neurologists are in
Cabeza et al. 2018). charge of the assessment of oral functions and
Figure 9.2 Key elements for implementing the brain–stomatognathic integrative assessment. (a) The
multidisciplinary investigation can be undertaken in a hospital, where different departments are in charge
of different assessments. Critically, the results of an assessment from one discipline (e.g. the score of
cognitive tests from neurologists) are distributed for the use of other disciplines (e.g. dentistry). For
example, the cognitive performance of older patients, as assessed by neurologists, is available for
prosthodontists to evaluate if patients can adapt well to their new dentures. (b) In contrast to the hospital
setting, a home-based assessment can be facilitated by the use of teledentistry approaches and digital
technology. For example, in long-term care institutes or at home, patients can record their own oral status
by photographing (via a smartphone). The image record may include a photo about their intraoral
conditions (e.g. bleeding gum) or performance of oral functions (e.g. the food bolus after chewing). The
images are sent to the cloud storage service for further analysis. A preliminary assessment is performed
automatically by machine-learning-based algorithms, and critical problems (e.g. a poor oral mixing ability)
are screened and forwarded to dental professionals for further evaluation.
cognitive functions, respectively (Figure 9.2a). again send the image for evaluation. The
More sophisticated methods can be performed same approach applies to other domains of
in hospitals or clinics. For example, in dental the BSIA. For example, nutritional and cogni-
clinics, the pattern of tooth loss and the pattern tive status can be assessed by online question-
of jaw movement can be recorded using an naires, and the information is integrated for
intraoral scanner and a jaw-tracking device, dentists to make clinical advice, such as the
respectively (Table 9.1). Notably, the assess- need for a new denture or oral exercises for
ments can also be conducted outside hospitals improving oral functions (Figure 9.2b).
or clinics, such as long-term care (LTC) insti-
tutes, with some modifications (Table 9.1) and 9.1.4.3 Application of Teledentistry
integration of assessment results (Figure 9.2b). The concept of the integration of results from
For example, either in LTC institutes or in a home-based assessment fits into the trend of
patient’s home, intraoral conditions can be teledentistry, which has been widely dis-
recorded by photographing (via a mobile cussed and put into practice, especially dur-
phone) and sent to the cloud storage service. ing the period of COVID-19 lockdowns
The images can be proceeded by machine- (Rahman et al. 2020). Two approaches from
learning methods for an initial screen and eval- telehealth have been adopted for years.
uation. The results may be sent back to patients Firstly, patient conditions can be directly
or sent forward to dental professionals for fur- evaluated via a real-time consultation, which
ther investigation (Figure 9.2b). For example, can be undertaken distantly using video-
machine learning has been used to quantify mediated communication. Another approach
one’s mixing ability according to the AI-based is the ‘store-and-forward’ method, which can
analysis of the pattern of gum mixing (Vaccaro be implemented using cloud storage services.
et al. 2018). According to the BSIA, the pattern In terms of teledentistry, oral assessments are
of food mixing can be photographed by patients performed by a teledental assistant, who
and sent to the cloud service for further pro- helps to collect data (e.g. photographing the
cessing, and useful information can be sent to oral images). Subsequently, the data are
dentists for classifying and predicting patients’ uploaded and stored in a remote server for
feeding behaviour in the near future. dentists to evaluate. The store-and-forward
provided good sensitivity and specificity
9.1.4.2 A Home-based Assessment Based when dentists graded patients’ oral health
on Digital Technology from the stored images (Estai et al. 2016).
As noted in the preceding section, the key to Notably, such a method only utilized simple
run the BSIA model successfully is to increase devices, such as the camera from a smart-
its feasibility. Therefore, it is critical to devel- phone, for capturing pictures of oral condi-
oping a simple method of assessment that can tions (Estai et al. 2016). Therefore, the method
be performed in LTC institutes or one’s home is suitable for a home-based assessment of the
by caretakers, even without help from dental BSIA. The reliability of the assessment of oral
professionals (Figure 9.2b). As shown in conditions via smartphones has recently been
Table 9.1, the assessment of tooth loss and reported. Two examiners evaluated the condi-
masticatory performance can be simplified tion of oral caries via the images recorded
using digital methods. Intraoral conditions from dental patients using smartphones. The
can be photographed with the built-in camera results were matched to the results from a
from a mobile phone, and the images are sent face-to-face examination for the same patient.
for further evaluation. In terms of mastica- They found the sensitivity ranging from 0.60
tory performance, one can also photograph to 0.63, with good intra-rater reliability
the shape and colour of the chewed bolus and (weighted kappa: 0.84) (Estai et al. 2017). Six
examiners evaluated the oral health condition matter volume (GMV) of the premotor cortex
of children via the images taken using smart- would predict the individual difference in mas-
phones. The results showed averagely good ticatory performance (Lin et al. 2020). The find-
sensitivity and specificity (approximately 0.8) ings suggest that in addition to intraoral
and good inter-rater reliability (intra-class conditions, additional information of structural
correlation: 0.96 for average measures) brain features contributes to the prediction of
(Alshaya et al. 2020). Results from the studies individual oral functions. Furthermore, such a
have revealed that teledentistry using the model does not apply to patients with cognitive
intraoral images from smartphones is a prac- impairment. In patients with MCI or
tical method for an ‘offline’ assessment and Alzheimer’s disease (AzD), the number of
consultation for patients’ oral condition. tooth loss, rather than structural brain features,
However, it should be noted that most of the is the dominant factor associated with individ-
assessments relate to the examination of gen- ual differences in masticatory performance
eral oral conditions and caries, which would (Lin et al. 2020). The outcome highlights the
be relatively easier to be spotted via oral strength of the BSIA framework, i.e. by concur-
images. Other intraoral conditions, e.g. peri- rently assessing and analyzing the data of oral,
odontal health, may require more sophisti- physical and mental domains, one can better
cated assessment, such as dental probing, capture individual differences in feeding behav-
which cannot be easily replaced by images. iour for people with a more diverse background.
Moreover, the protocol to take an oral image
may require further standardization: while
9.1.6 Summary
using a mobile phone is an easy approach to
take intraoral images for healthy adults, it ●● The framework of BSIA aims to fully assess
may be relatively difficult to take a clear individual profiles of oral structure and func-
image for the very frail old patients or patients tions. The adoption of an integrative assess-
with cognitive impairment. ment, which consists of both the assessment
of intraoral conditions and the brain-
behavioural factors related to feeding behav-
9.1.5 Research on the BSIA
iour, should be considered in clinical practice.
A preliminary investigation has been con- ●● The BSIA assesses individual brain-behaviour
ducted at National Yang-Ming University, factors that reflect the status of brain and cog-
Taiwan, for testing the concept of the BSIA. The nitive reserve, which would be helpful for
main research focus is the association between predicting the change of individual oral func-
the number of tooth loss (i.e. the stomatog- tion and feeding behaviour and classifying
nathic structure), masticatory performance their susceptibility to oral diseases.
(i.e. the stomatognathic function) and individ- ●● The BSIA framework proposed here con-
ual variability in brain features (i.e. the indi- sists of assessments for three dimensions:
vidual reserve) (Table 9.1). Moreover, because assessment of the structural integrity of
brain structure and function show a substan- the stomatognathic system, assessment of
tial difference between cognitively healthy and stomatognathic functions and assessment
impaired subjects, the study also focused on of the ‘reserves’ for patients to functionally
the difference of the associations between adapt to oral diseases.
these two clinical groups (i.e. cognitively ●● Practically, the BSIA framework requires an
healthy vs. cognitively impaired). architecture of information exchange
In cognitively healthy subjects, multidiscipli- between the data collected from different
nary assessments revealed that both the num- venues. The development of home-based
ber of missing teeth and the regional grey assessment, based on digital technology,
may provide a solution for the BSIA of the 9.2.2 Cross-talk Between Neuroimaging
residents of long-term care institutions. and Animal Research
Throughout this book, we have seen a large
9.2 Future Direction body of evidence in oral neuroscience derived
of Neuroimaging from animal research. Many key findings of
the stomatognathic system are based on ani-
in Oral Neuroscience
mal research, which is able to delve into the
cause–effect relationship of neural mecha-
9.2.1 Introduction
nisms from molecular to behavioural levels
As noted in Chapter 1, neuroimaging has been (Iwata and Sessle 2019). An important task of
adopted extensively in research on clinical oral neuroscience is to associate the findings
neurology and psychiatry for more than from animal research with those from human
30 years. However, its application in oral neu- research. Neuroimaging may play a pivotal
roscience and dentistry has just started. As a role in bridging these two fields because it can
new topic with a strong multidisciplinary be applied to both animal and human sub-
nature, neuroimaging should not be consid- jects. In recent years, there is an increasing
ered as an approach irrelevant to other trend to investigate the brain mechanisms of
approaches. In the final chapter, we focus on animal subjects using neuroimaging methods
the importance (and necessities) of a collabo- (Table 9.2).
ration between neuroimaging and other The combined use of animal models and
approaches and discuss how this alliance revo- neuroimaging methods helps to elucidate the
lutionizes oral neuroscience and dentistry. causal mechanisms of oral functions. For
Table 9.2 Recent findings on neuroimaging research on oral functions based on animal models
(since 2015).
Notes: ASL: arterial spin labelling magnetic resonance imaging; sMRI: structural magnetic resonance imaging.
example, Nakamura et al. (2015) used optical cerebral blood flow (CBF) accompanied by
imaging to investigate the neural excitation in cellular changes in the hippocampus. Using
the sensorimotor cortex associated with pulpal structural MRI, Avivi-Arber et al. (2017) com-
stimulation. They found that electrical stimu- pared the post-mortem brain volume of mice
lation of dental pulp in maxillary and mandib- that received molar extraction and those who
ular teeth was associated with the activity of received sham operation. They found that
the primary somatosensory cortex (S1) and the tooth extraction was associated with a wide-
secondary somatosensory cortex (S2)/insula spread reduction in volumes of brain regions
(Nakamura et al. 2015). Kaneko et al. (2017) of sensorimotor and cognitive–affective pro-
found that in rats the activity of the S1 and the cessing. Moreover, increased volumes were
S2/insula was associated with mechanical and found in subcortical regions, such as the amyg-
electrical stimulation on the periodontal liga- dala (Avivi-Arber et al. 2017) (Figure 9.3). The
ment (PDL) of the maxillary first molar, neuroimaging findings of animal subjects,
respectively. Furthermore, the region that together, confirm the role of the sensorimotor
responded to mechanical stimuli on the incisor cortex in oral functions. Moreover, the findings
and molar would shift after receiving experi- help to elucidate the potential brain mecha-
mental tooth movement that ligated the inci- nisms of brain plasticity of altered oral
sor and molar (Kaneko et al. 2018). Using a conditions.
similar method, Horinuki et al. (2015) found
that one day after experimental tooth move-
9.2.3 Artificial Intelligence
ment, there was a greater and broader excita-
tion in the S2/insula in response to electrical The cross-talk between medicine and artificial
stimulation of the PDL. Using transgenic mice, intelligence (AI) has become a major field in
Mayrhofer et al. (2019) demonstrated a shift of healthcare. Particularly, the use of machine-
the activity from primary whisker sensory cor- learning methods has contributed to disease
tex (during stimulus detection) to primary diagnosis and clinical assessment. In terms of
tongue–jaw motor cortex (during a licking neuroimaging, methods based on machine
task) in goal-directed sensorimotor transfor- learning, such as multivariate pattern analy-
mations. Clemens et al. (2018) found that in sis, have been widely used for identifying the
rats the activity of the neurons of the oral complex pattern of brain activation associated
sensorimotor cortex may reflect multisensory with mental functions. For example, by train-
integration during feeding. The activity was ing an algorithm to map the association
associated with jaw and tongue movement between individual experience (e.g. painful or
related to feeding behaviour and the percep- not) and fMRI data collected during stimula-
tion of temperature (but not taste) of oral tion, one can decode pain experience from
stimuli (Clemens et al. 2018). brain activation with good accuracy (for
Animal research also helps to elucidate the example, see Brodersen et al. 2012; Wager
association between the stomatognathic sys- et al. 2013). In the following sections, we focus
tem and brain functions other than sensorimo- on the application of AI, especially the
tor processing. For example, Xu et al. (2015) machine-learning approach, on assessing the
reported that in rats tooth loss suppressed stomatognathic system.
the synchronization of local field potentials
between the amygdala and the anterior cingu- 9.2.3.1 Machine-learning-based Assessment
late cortex, which would be associated with the of Oral Structure and Function
deficits in decision-making. Luo et al. (2019) Using the convolutional neural network
found that rats with tooth loss showed not only approach, Xia et al. (2019) conducted a fully
deficits in spatial cognition but also decreased automatic segmentation of the mandible and
Figure 9.3 An example of neuroimaging investigation combined with animal research. Using structural
MRI, Avivi-Arber et al. (2017) compared the post-mortem brain volume between the mice that received
molar extraction and those who received sham operation. Tooth extraction was associated with a
widespread reduction in volumes of brain regions of sensorimotor and cognitive–affective processing.
Source: Avivi-Arber et al. (2017) with permission of Frontiers Media S.A. under the terms of the Creative
Commons Attribution 4.0 License.
the masseter based on computed tomography clinical features of the masseter (e.g. its size
(CT) images from 53 patients. By measuring and shape) can be assessed immediately when
the thickness of the masseter, the system the CT images are acquired and used for fur-
would help to locate the point for injecting the ther functional analysis. In terms of oral func-
muscle (Xia et al. 2019). Using the deep neural tion, Vaccaro et al. (2018) have analyzed the
network approach, Qin et al. (2018) delineated image of two-colour mixing chewing gum.
the masseter from cone-beam CT images by They developed a machine-learning algorithm
automatic segmentation. The segmented area that can estimate the masticatory performance
was highly similar with the masseter region based on the pattern of gum mixing, and the
labelled by two practitioners (Qin et al. 2018). results showed a significant difference between
Segmentation was also performed using other edentate and edentulous patients (Vaccaro
algorithms, such as a generative adversarial et al. 2018). One of the biggest advantages of
network (GAN)-based framework (Zhang AI-based methods is that a large body of data
et al. 2019). Because the segmentation is con- can be screened efficiently using the algo-
ducted automatically by computer algorithms, rithm, and dentists may focus more on the
cases that are picked by the algorithm for fur- the current data (Figure 9.1a). The application
ther (manual) investigation. Based on the AI- of AI may help to classify patients’ status based
based approach, it is feasible to establish a on the data collected from multidisciplinary
system of automatic and continuous assess- assessments. Such a method has been adopted
ment of oral conditions, which would be in many medical fields, such as the diagnosis
important for monitoring the oral health of of neurological and psychiatric disorders based
elderly and special needs patients. on the data from neuroimaging (Arbabshirani
et al. 2017; Calhoun et al. 2017).
9.2.3.2 Classification vs. Prediction
Notably, the AI-based approach is also useful 9.2.4 Summary
for generating a predictive model that forecasts
●● An important task of oral neuroscience is to
the change of patients’ oral conditions.
associate the findings from animal research
Traditionally, the research on the association
with those from human research.
between oral and systemic health is largely
Neuroimaging may play a pivotal role in
based on a design of group comparison. For
bridging these two fields because it can be
example, one would identify the risk of devel-
applied to both animal and human subjects.
oping periodontitis by comparing a group of
●● In terms of oral assessment, machine-
smokers with a group of non-smokers. The
learning-based methods have been used for
design and analyses clearly demonstrate the
segmenting the masseter from CT/MRI
difference at the group level. However, it fails
images and for assessing the pattern of mix-
to capture the complexity of individual status.
ing of chewing gum (i.e. oral mixing ability).
While group comparison reveals that a certain
factor (e.g. smoking) is associated with disease,
the use of machine-learning methods can Further Readings
reveal how multiple factors together relate to
disease. Another application of AI is to predict Please see the Companion Website for
the patients’ status in the long run, i.e. to fore- Suggested Readings and Tables with updated
cast the change of their future oral status from information.
References
Alshaya, M.S., Assery, M.K., and Pani, emerging concepts in cognitive reserve. Trends
S.C. (2020). Reliability of mobile phone in Cognitive Sciences 17: 502–509.
teledentistry in dental diagnosis and Brodersen, K.H., Wiech, K., Lomakina, E.I. et al.
treatment planning in mixed dentition. (2012). Decoding the perception of pain from
Journal of Telemedicine and Telecare fMRI using multivariate pattern analysis.
26: 45–52. NeuroImage 63: 1162–1170.
Arbabshirani, M.R., Plis, S., Sui, J., and Calhoun, Cabeza, R., Albert, M., Belleville, S. et al. (2018).
V.D. (2017). Single subject prediction of brain Maintenance, reserve and compensation: the
disorders in neuroimaging: promises and cognitive neuroscience of healthy ageing.
pitfalls. NeuroImage 145: 137–165. Nature Reviews. Neuroscience 19: 701–710.
Avivi-Arber, L., Seltzer, Z., Friedel, M. et al. Calhoun, V.D., Lawrie, S.M., Mourao-Miranda,
(2017). Widespread volumetric brain changes J., and Stephan, K.E. (2017). Prediction of
following tooth loss in female mice. Frontiers individual differences from neuroimaging
in Neuroanatomy 10: 121. data. NeuroImage 145: 135–136.
Barulli, D. and Stern, Y. (2013). Efficiency, Clemens, A.M., Fernandez Delgado, Y.,
capacity, compensation, maintenance, plasticity: Mehlman, M.L. et al. (2018). Multisensory
and motor representations in rat oral Mayrhofer, J.M., El-Boustani, S., Foustoukos,
somatosensory cortex. Scientific Reports G. et al. (2019). Distinct contributions of
8: 13556. whisker sensory cortex and tongue-jaw motor
Collins, F.S. and Varmus, H. (2015). A new cortex in a goal-directed sensorimotor
initiative on precision medicine. The New transformation. Neuron 103: 1034–1043. e5.
England Journal of Medicine 372: 793–795. Nakamura, H., Kato, R., Shirakawa, T. et al.
Estai, M., Kanagasingam, Y., Xiao, D. et al. (2015). Spatiotemporal profiles of dental pulp
(2016). A proof-of-concept evaluation of a nociception in rat cerebral cortex: an optical
cloud-based store-and-forward telemedicine imaging study. The Journal of Comparative
app for screening for oral diseases. Journal of Neurology 523: 1162–1174.
Telemedicine and Telecare 22: 319–325. Qin, H., Pei, Y., Guo, Y., et al. (2018). Masseter
Estai, M., Kanagasingam, Y., Huang, B. et al. segmentation from computed tomography
(2017). Comparison of a smartphone-based using feature-enhanced nested residual neural
photographic method with face-to-face caries network, 355–362. International Workshop on
assessment: a mobile teledentistry model. Machine Learning in Medical Imaging.
Telemedicine Journal and E-Health 23: Springer.
435–440. Rahman, N., Nathwani, S., and Kandiah,
Horinuki, E., Shinoda, M., Shimizu, N. et al. T. (2020). Teledentistry from a patient
(2015). Orthodontic force facilitates cortical perspective during the coronavirus pandemic.
responses to periodontal stimulation. Journal British Dental Journal 229: 1–4.
of Dental Research 94: 1158–1166. Vaccaro, G., Pelaez, J.I., and Gil-Montoya,
Iwata, K. and Sessle, B.J. (2019). The evolution of J.A. (2018). A novel expert system for objective
neuroscience as a research field relevant to masticatory efficiency assessment. PLoS One
dentistry. Journal of Dental Research 98: 13: e0190386.
1407–1417. Wager, T.D., Atlas, L.Y., Lindquist, M.A. et al.
Kaneko, M., Horinuki, E., Shimizu, N., and (2013). An fMRI-based neurologic signature
Kobayashi, M. (2017). Physiological profiles of of physical pain. The New England Journal of
cortical responses to mechanical stimulation Medicine 368: 1388–1397.
of the tooth in the rat: an optical imaging Xia, W.J., Han, W.Q., Zhang, Y., et al. (2019).
study. Neuroscience 358: 170–180. Automatic Masseter Thickness Measurement
Kaneko, M., Fujita, S., Shimizu, N. et al. (2018). and Ideal Point Localization for Botulinum
Experimental tooth movement temporally Toxin Injection, 2801-2804. Annu Int Conf
changes neural excitation and topographical IEEE Eng Med Biol Soc, 2019.
map in rat somatosensory cortex. Brain Xu, X., Cao, B., Wang, J. et al. (2015). Decision-
Research 1698: 62–69. making deficits associated with disrupted
Lin, C.S., Lin, H.H., Wang, S.J., and Fuh, synchronization between basolateral
J.L. (2020). Association between regional amygdala and anterior cingulate cortex in rats
brain volume and masticatory performance after tooth loss. Progress in Neuro-
differed in cognitively impaired and non- Psychopharmacology & Biological Psychiatry
impaired older people. Experimental 60: 26–35.
Gerontology 137: 110942. Zhang, Y., Pei, Y., Qin, H., et al. (2019). Masseter
Luo, B., Pang, Q., and Jiang, Q. (2019). Tooth loss Muscle Segmentation from Cone-Beam CT
causes spatial cognitive impairment in rats Images using Generative Adversarial
through decreased cerebral blood flow and Network, 1188–1192. 2019 IEEE 16th
increased glutamate. Archives of Oral Biology International Symposium on Biomedical
102: 225–230. Imaging (ISBI 2019). IEEE.
Index
Dental Neuroimaging: The Role of the Brain in Oral Functions, First Edition. Chia-Shu Lin.
© 2022 John Wiley & Sons Ltd. Published 2022 by John Wiley & Sons Ltd.
Companion website: www.wiley.com/go/lin/dental-neuroimaging