FU GEA
P.
J.
aSc
Sci
sity
bDe
Lon
cGr
of W
dDe
Me
tari
Ab
pla
po
im
of
lat
sti
ne
tha
me
are
su
era
gio
mo
sw
air-
tor
Co
All
Ke
gu
co
Th
be
al.
19
rip
be
wh
an
tat
cu
co
ou
bra
an
se
se
et
on
inv
oro
tio
tio
oro
ica
he
init
Lo
se
mo
oro
tat
me
ce
lat
wit
us
Cli
as
an
19
(H
19
stim
(e.
the
no
pulses delivered to the oropharynx, either unilaterally or,
particularly, bilaterally, increased the frequency of saliva
*Correspondence to: R. E. Martin, School of Communication Sciences &
Disorders, Faculty of Health Sciences, Elborn College, Room 2568, 1201
We
N6B
E-m
Abb
tion
ana
Mo
prim
ech
Ne
030
doiswallowing in healthy younger adults (Theurer et al.,stern Road, University of Western Ontario, London, Ontario, CanadaNCTIONAL MRI OF OROPHARYN
SÖRÖS,a E. LALONE,a R. SMITH,a T. STEVENS,b
THEURER,a,c R. S. MENONb,d AND R. E. MARTINa,c,d,*
hool of Communication Sciences & Disorders, Faculty of Health
ences, Elborn College, Room 2568, 1201 Western Road, Univer-
of Western Ontario, London, Ontario, Canada N6B 1H1
partment of Medical Biophysics, University of Western Ontario,
don, Ontario, Canada
aduate Program in Health and Rehabilitation Sciences, University
estern Ontario, London, Ontario, Canada
partment of Physiology and Pharmacology, Schulich School of
dicine and Dentistry, University of Western Ontario, London, On-
o, Canada
stract—Although the posterior oral cavity and oropharynx
y a major role in swallowing, their central representation is
orly understood. High-field functional magnetic resonance
aging of the brain was used to study the central processing
brief air-pulses, delivered to the peritonsillar region of the
eral oropharynx, in six healthy adults. Bilateral air-pulse
mulation was associated with the activation of a bilateral
twork including the primary somatosensory cortex and the
lamus, classic motor areas (primary motor cortex, supple-
ntary motor area, cingulate motor areas), and polymodal
as (including the insula and frontal cortex). These results
ggest that oropharyngeal stimulation can activate a bilat-
lly distributed cortical network that overlaps cortical re-
ns previously implicated in oral and pharyngeal sensori-
tor functions such as tongue movement, mastication, and
allowing. The present study also demonstrates the utility of
pulse stimulation in investigating oropharyngeal sensorimo-
processing in functional brain imaging experiments. Crown
pyright © 2008 Published by Elsevier Ltd on behalf of IBRO.
rights reserved.
y words: insula, supplementary motor area, anterior cin-
late gyrus, primary somatosensory cortex, primary motor
rtex, functional magnetic resonance imaging.
e posterior oral cavity, and adjacent oropharynx, are
lieved to play critical roles in bolus transport (Palmer et
, 1992), and triggering of pharyngeal swallowing (Doty,
68; Miller, 1982, 1998; Carpenter, 1989). While the pe-
heral and brainstem innervation of the oropharynx have
en described (Capra, 1995; Jean, 2001), the extent to
ich the oropharynx is represented within central cortical
uroscience 153 (2008) 1300–130820
mi
19
se
ph
po
1H1. Tel:
1-519-661-2111x88186; fax:
1-519-850-2369.
ail address: remartin@uwo.ca (R. E. Martin).
reviations: BOLD, blood-oxygenation-level-dependent; fMRI, func-
al magnetic resonance imaging; ICA, independent component
lysis; ISI, interstimulus interval; MI, primary motor cortex; MNI,
ntreal Neurological Institute; MRI, magnetic resonance imaging; SI,
ary somatosensory cortex; SMA, supplementary motor area; TE,
o time; TR, repetition time.
6-4522/08$32.00
0.00 Crown Copyright © 2008 Published by Elsevier Ltd on beh
:10.1016/j.neuroscience.2008.02.079
1300L AIR-PULSE STIMULATION
d subcortical structures, and the nature of this represen-
ion, remain poorly understood. This is in contrast to
rrent knowledge of the oral cavity, jaw, and tongue (for a
mprehensive review see Sessle, 2006) for which previ-
s electrophysiologic studies in animal models, and
in-imaging studies in humans, have identified sensory
d motor representations within the lateral primary and
condary sensorimotor cortices (Penfield and Rasmus-
n, 1950; Murray et al., 1991; Martin et al., 1999; Servos
al., 1999). These cortical areas appear to mediate not
ly oral sensation and elemental movements, but also are
olved in the initiation and control of a variety of complex
pharyngeal sensorimotor behaviors including mastica-
n, swallowing, voluntary tongue movement, and saliva-
n (Sessle, 2006).
There is reason to suggest that, like the oral cavity, the
pharynx is represented within the cerebral cortex. Clin-
lly, patients who have swallowing impairment following
mispheric stroke often show a delay, or abnormality, in
iation of pharyngeal swallowing (Johnson et al., 1992;
gemann, 1998), a function ascribed to oropharyngeal
nsory mechanisms in the healthy population. Further-
re, the palate and larynx, structures that adjoin the
pharynx, have been shown to have cortical represen-
ions, suggesting the possibility of oropharyngeal central
chanisms.
The relative lack of knowledge of an oropharyngeal
ntral somatosensory representation appears to be re-
ed to the relative difficulty of accessing the oropharynx
h the variety of somatic stimuli that have typically been
ed in studies of oral sensation (Boliek et al., 2007).
nical testing of oral sensory functions has involved the
sessment of two-point discrimination (Aviv et al., 1992)
d of oral stereognosis (using oral forms) (Jacobs et al.,
98). Physiologic studies have employed a blunt probe
ayashi et al., 1984), electrical pulses (Hayashi et al.,
84) or distilled water (Zald and Pardo, 2000) as intraoral
uli. Many of these stimuli cannot reach the oropharynx
g. oral form, blunt probe). A further limitation of some of
se stimuli, particularly electrical stimulation, is that it is
t a natural stimulus and may evoke a gustatory percept.
In a previous study, we demonstrated that trains of air05). Air pulses have also been used clinically to deter-
ne sensory thresholds of the laryngopharynx (Aviv et al.,
93). The air pulse has certain advantages as a somato-
nsory stimulus in studies of oropharyngeal sensorimotor
ysiology. First, an air pulse can be directed toward the
sterior oral cavity and oropharynx, allowing examination
alf of IBRO. All rights reserved.

of
as
ele
pe
cu
slo
Th
plo
reg
pre
ing
Pa
Six
oro
par
yea
me
Th
son
(fM
sub
rela
our
air-
as
low
com
vol
fem
ava
ana
of
Re
Su
par
Sti
Fo
thr
ous
fun
of t
pha
righ
del
20
we
the
thic
two
infe
ma
edg
ter
late
ton
lati
oro
stim
alb
the
twe
con
wit
Y-c
exp
cue
vis
Re
Lar
(AD
sig
res
gen
car
afte
(me
fM
Bra
wh
wit
gra
fre
ber
pad
5 m
ant
fro
10
abo
ima
seg
rep
Fig
ryn
thro
P. Sörös et al. / Neuroscience 153 (2008) 1300–1308 1301areas that are not accessible with other methods such
a probe. Air pulses are natural stimuli, compared with
ctrical stimulation, that do not appear to evoke gustatory
rceptions. In addition, air pulses activate all four types of
taneous mechanoreceptors (fast adapting type 1 and 2,
wly adapting type 1 and 2) (Mizobuchi et al., 2000).
erefore, air pulses may have utility as a means of ex-
ring the cortical representation of the oropharynx, a
ion believed to be critical in swallowing. The aim of the
sent study, then, was to investigate the central process-
of oropharyngeal air-pulse stimulation.
EXPERIMENTAL PROCEDURES
rticipants
right-handed female volunteers with no history of swallowing,
facial, gastrointestinal, respiratory, or neurological problems
ticipated in the study (mean age: 27 years, age range: 21–45
rs). All subjects were non-smokers, and were not taking any
dications that may have affected their oropharyngeal function.
e sample was composed exclusively of females for two rea-
s. First, our previous functional magnetic resonance imaging
RI) experience has suggested that fMRI data from female
jects are relatively less affected by motion artifact, perhaps
ted to their smaller larynx and laryngeal movement. Second,
previous study demonstrating the effects of oropharyngeal
pulse application on saliva swallowing (Theurer et al., 2005),
well as previous physiologic and brain-imaging studies of swal-
ing, have focused on female subjects, thus providing a basis of
parison with the present study of females. In addition to the six
unteers mentioned above, fMRI was performed in a 25-year-old
ale participant whose recordings of laryngeal motion were not
ilable. This participant was excluded from the following data
lysis.
The study was conducted in accordance with the Declaration
Helsinki after approval by the University of Western Ontario
view Board for Health Sciences Research Involving Human
bjects. All individuals gave written informed consent before
ticipating in the study.
mulation
r oropharyngeal stimulation, trains of air pulses were delivered
ough a custom-made lower dental splint as described previ-
ly (Theurer et al., 2005). Each subject participated in three
ctional imaging runs of 5 min duration. Each run contained one
he following conditions: left-sided, right-sided, or bilateral oro-
ryngeal air-pulse stimulation. The order of stimulation (left,
t, bilateral) was randomized. Within each run, stimuli were
ivered in six blocks of 10 s duration each. Each block contained
air pulses (stimulation frequency: 2 Hz). Successive blocks
re separated by a mean interstimulus interval (ISI) of 20 s, with
ISI varied randomly between 15 s and 25 s.
Splints were made of human-implant-grade silicone with a
kness of less than 2 mm (Fig. 1). For delivery of the air-pulses,
polyethylene tubes were embedded on the left and right
rior border of the splint, lateral to the alveolar ridge of the
ndible, that extended approximately 1 cm past the posterior
e. Tubes (inner diameter: 1.14 mm, outer diameter: 1.57 mm)
minated with a circular opening (diameter: 1.5 mm) on the
ral wall. This opening directed the air-pulse toward the peri-
sillar region of the lateral oropharynx. During unilateral stimu-
on, air pulses stimulated mainly the left and right part of the
pharynx. The air pulse, however, filled the entire cavity and
ulated the contralateral portion of the oropharynx as well,
eit less intensively. Anteriorly, the tubes from both sides exited
gle
Easplint approximately 1 cm to the left of midline, passed be-
en the subject’s lips, and extended 50 cm where they were
nected to a larger-diameter tube (i.e. 2 mm) that interfaced
h a Y-connector. Large-diameter tubing (i.e. 4 mm) ran from the
onnector to an air bulb that was manually operated by the
erimenter. Stimulation was performed in response to a visual
that was projected on a screen inside the magnet room,
ible to the experimenter but not to the participant.
cording of laryngeal movements
yngeal movements were recorded using PowerLab, v4.1.1
Instruments, Colorado Springs, CO, USA) from the output
nal of a pressure transducer driven from expanding magnetic
onance imaging (MRI)-compatible bellows (Siemens, Erlan-
, Germany) positioned comfortably over the subject’s thyroid
tilage. In the vast majority of events, participants swallowed
r air pulse stimulation. Swallows occurred on average 125.6 s
anstandard deviation) after the onset of the stimulation.
RI data acquisition
in imaging was performed on a Varian UNITY INOVA 4 T
ole-body MRI system (Varian, Palo Alto, CA, USA) equipped
h 40 mT/ms Siemens Sonata actively shielded whole-body
dients and amplifiers. A whole-head quadrature birdcage radio
quency (RF) coil transmitted and received the MR signal (Bar-
i et al., 2000). The subject’s head was restrained with foam
ding. Functional data were collected from 18 contiguous,
m-thick axial slices oriented approximately parallel to the
erior commissure–posterior commissure plane and extending
m the superior extent of the paracentral lobule to approximately
mm below this plane. During each functional task described
ve, blood-oxygenation-level-dependent (BOLD) T2*-weighted
ges were acquired continuously using an interleaved, four
ment, echo planar imaging sequence (6464 matrix size,
etition time (TR)2000 ms, echo time (TE)10 ms, flip an-
. 1. Custom-built, fMRI compatible lower dental splint for oropha-
geal air-pulse stimulation. The arrows indicate the openings
ugh which the air pulses exit the tubing on the sides of the splint.30°, field of view19.2 cm, volume collection time2 s).
ch image was corrected for physiologic fluctuations using a