Abstract—In patients with gastroesophageal reflux disease
(GERD), esophageal symptoms are traditionally diagnosed by
monitoring the contact time between the reflux content and the
esophagus using multichannel intraluminal impedance and pH
(MII-pH) catheters. However, esophageal catheter for
quantifying the volume of reflux content is still lacking. The
present work proposes an innovative method to develop a
longitudinal ultrasonic catheter and an information extraction
system for reflux event detection and reflux volume estimation.
Gastroesophageal model that mimics reflux events was
developed to test the proposed catheter. Ultrasonic sensing was
evaluated by simulating different volumes of reflux. The
obtained signals showed good consistency in detecting reflux
events and measuring reflux volume. During an in vivo human
testing, a MII-pH catheter was used simultaneously to compare
the ultrasonic output. Both in vitro and in vivo human testing
results demonstrated the feasibility of utilizing the proposed
method for gastroesophageal reflux (GER) detection and reflux
CCORDING to recent statistics provided by the
National Institutes of Health, symptoms consistent with
gastroesophageal reflux disease (GERD) resulted in 710,000
hospitalizations in the USA for 2002 . Gastroesophageal
reflux (GER) is characterized by abnormal retrograde flow of
gastric content into the esophagus resulting in various
symptoms and mucosal damage . It is one of the most
common conditions that affect the gastrointestinal tract and is
usually thought to be the cause of the majority of esophageal
symptoms . By estimating the GER volume, the total
amount of highly concentrated damaging substances can be
quantified, which can help to better understand the adverse
GERD phenomena and to evaluate the efficacy of various
antireflux treatments . Two ambulatory methods have been
attempted to estimate the GER volume: 1) multichannel
intraluminal impedance and pH (MII-pH) monitoring and 2)
Manuscript received April 20, 2010. This work was supported in part by
Sandhill Scientific Inc., by the Gastrointestinal Motility Laboratory
(Edmonton, AB, CDN), and by the China Scholarship Council.
X. Gao is with the Department of Electrical and Computer Engineering,
University of Calgary, Calgary, AB T2N 1N4, Canada (phone:
403-210-9553; fax: 403-282-6855; e-mail: email@example.com).
D. Sadowski is with the Division of Gastroenterology, University of
Alberta, Edmonton, AB T6G 2J1, Canada and the University of Alberta
Hospital, Edmonton, AB T6G
M. P. Mintchev is with the Department of Electrical and Computer
Engineering, University of Calgary, Calgary, AB T2N 1N4, Canada and also
with University of Alberta, Edmonton, AB T6G 2J1, Canada (e-mail:
2J1, Canada (e-mail:
intraluminal transversal ultrasonic imaging . MII-pH
monitoring has a high sensitivity in detecting the proximal
extent of GER, which is considered to be an indirect indicator
of reflux volume . However, the relationship between the
volume and the proximal extent of reflux has not been
reliably obtained . Recently, intraluminal transversal
ultrasonic imaging catheters have been studied for measuring
the cross-sectional area (CSA) of the distal esophagus during
distensions by spontaneous reflux events. Compared to the
distal distension in normal subjects, much larger distensions
of the esophagus were found in patients with GERD
symptoms , . However, in order to measure the volume
of reflux, cross-sectional images of esophagus at different
longitudinal levels are necessary. Due to cost and technical
limitations, multichannel transversal ultrasonic imaging
catheters for recording multiple cross-sectional images
simultaneously are still not available. In the present study we
suggest the limitations of MII-pH monitoring and transversal
ultrasonic imaging for reflux volume estimation can be
overcome using a longitudinal ultrasonic technique and
customized information extraction software.
The aim of this research is to develop an innovative
esophageal catheter with an embedded longitudinal ultrasonic
sensor capable not only of detecting spontaneous reflux
events, but also of dynamically estimating GER volumes at
5cm above the lower esophageal sphincter (LES).
Advantageously, this catheter-based design will provide a
miniature, simple and low-cost diagnostic tool in 24-hour
ambulatory studies, in which the patients could continue with
their typical daily activities.
II. METHODS AND MATERIALS
A. Principle of Operation
Longitudinal ultrasonic sensing is capable of detecting
distension of distal esophagus provoked by gastric content
present in the esophageal lumen. During the resting state, the
walls of the distal esophagus are collapsed and LES is closed.
The ultrasonic waves sent by the sensor are reflected from the
nearby esophageal wall. The presence of gastric content in the
esophageal lumen prolongs the time for the ultrasonic waves
to be reflected from the esophageal wall. These changes of
the echoes could be extracted from the ultrasonic system as an
indicator of the presence of GER content.
Pilot Study of Longitudinal Ultrasonic Sensor for Dynamic
Volumetric Assessment of Gastroesophageal Reflux
Xuexin Gao, Daniel C. Sadowski and Martin P. Mintchev
32nd Annual International Conference of the IEEE EMBS
Buenos Aires, Argentina, August 31 - September 4, 2010
978-1-4244-4124-2/10/$25.00 ©2010 IEEE899
If one or more diameters of the distal esophagus during a
GER event are measured dynamically, an estimation of the
reflux volume can be obtained with some approximations and
assumptions. The first important assumption is that the reflux
volume from the tip of the catheter down to the LES is
correlated with the total amount of reflux. The second
assumption is to consider the distal esophagus shape to be
symmetrical about one of its axes. The third one is that reflux
dynamics can be modeled by a mathematic function.
Model of the esophageal lumen which has been frequently
used in esophageal bolus transport simulation is a cylinder
. This cylindrical model could also be used for modeling
reflux content dynamics when distal esophageal lumen is
distended during reflux episodes. The volume of this model
V can be calculated according to the formula:
where D is the transversal distal esophageal diameter and
h is the length between transducer and the LES.
The measurement of D could be obtained by using a
longitudinal ultrasonic sensor with an angle of divergence ϕ :
D is the distance between the transducer and the
esophageal wall (Fig. 1).
Fig. 1. Position of the ultrasonic transducer within the distal esophagus and
schematic of ultrasonic measurement.
B. System Overview
The proposed system consists of a custom-designed
longitudinal ultrasonic catheter (Valpey Fisher, Hopkinton,
MA), an ultrasonic signal conditioner (USUltratek, Concord,
CA), and a custom-made data acquisition and signal
processing software (Fig. 2). The outer diameter and the
height of the cylindrically shaped transducer are 2.54 mm
each. The ultrasonic signal conditioner excites the transducer
with a low-power electrical pulse of 200 Volts and duration of
200 ns. The crystal is resonating at 1 MHz and acting as both
transmitter and receiver. The echo signals are amplified 78
dB, and then sampled at 5 MS/s. The computer based
software is a graphical user interface (GUI) developed using
Matlab environment (MathWorks, Natick, MA) for data
storing and signal processing.
Fig. 2. Block Diagram of the system
C. Signal Processing Algorithm
The proposed signal processing algorithm consists of
reflux detection stage and reflux volume estimation stage.
1) Reflux Detection: The underlying structure of the reflux
detection stage calculates the mean value of the echo
amplitudes received from the nearby interfaces. This value
decreases as the reflux develops and can be calculated
is the output sequence extracted from multiple
ultrasonic waveform frames.
echo from a certain distance m in the current sampling
waveform. R determines the distance range of the nearby
interfaces for calculating the mean amplitude.
Determining R was achieved by placing the ultrasonic
catheter into a gastroesophageal model during simulated
reflux events and comparing their corresponding waveforms
with the waveforms acquired during non-reflux state. By
repeating this procedure 10 times, an optimal value of the
distance was found equal to 2 cm.
1- 10 )( , ..., N, nnx
is the amplitude of the
When the catheter is intubated into the esophagus of a
patient, it is estimated that the ultrasonic transducer may
respond not only to reflux events but to other events (i.e.
cough, belching and respiration). Therefore, the acquired
should be initially subjected to a filtering stage
which focused on amplifying the response characteristic of
the reflux events. The filter, which is a 10th order Butterworth
low-pass filter with a passband frequency of 0.2 Hz, is
designed with the assumption that the duration of one reflux
episode in the subject is normally longer than 2.5 s.
In the decision stage, reject control is added to extract the
time points at which the echoes are smaller than a threshold
value, which indicates the presence of reflux (Fig. 3).
Fig. 3. Structural overview of the reflux detection stage.
2) Reflux Volume Estimation: The reflux detection stage is
followed by the reflux volume estimation stage. After
recognizing the time points of reflux, a volume estimation
algorithm employs peak detection stage, which performs the
estimation of the diameter and the height of the reflux model.
D. Experimental Setup
1) In Vitro Testing: Reflux periods were simulated using a
computer-controlled mechanical esophageal model (Fig. 4).
The stomach model was initially filled of colored solution.
For creating reflux episodes with repeatable strength and
duration, motor controlled clamps around the stomach part
were used to produce controllable pressure. The ultrasonic
catheter was positioned about 4 cm above Gripper1, which is
designed to have the similar function as LES.
Controlled by the clamps, the gastric content in the
stomach will flow into the esophagus if the LES doesn’t close
properly. The diameter of the esophageal model could be
varied from 0.5 to 2 cm to simulate different CSA of the distal
esophagus during reflux episodes. Using this setup, the
measurements from the ultrasonic output could be related to
the presence of reflux and the volume of that reflux.
Fig. 4. Experimental gastroesophageal model for testing the feasibility of the
reflux volume measurement design.
2) In Vivo Human Testing: In vivo human testing was
performed on a volunteer over a 1 h period to obtain initial
results of the sensor response to reflux dynamics. First, high
resolution esophageal manometry probe (Sierra Scientific
Instruments, Los Angeles, CA) was intubated transnasally to
locate the position of the LES. After the manometry probe
was pulled out, an MII-pH probe (Sandhill Scientific,
Highland, CO) was inserted transnasally, along with the
ultrasonic catheter inserted through the other nostril.
According to the manometrically determined position of the
LES, the tip of the ultrasonic catheter was adjusted to about
5cm above the LES. Then, the two probes were secured to the
subject at the point of entry.
During the tests, both liquid and gas reflux episodes were
indicated from the MII-pH system, which were used to
calibrate the ultrasonic system and to evaluate its responses to
the MII-pH registered reflux events. Sensor responses were
monitored as the subject experienced a number of events in
random order: heartburn, belching and swallowing.
A. In Vitro Testing
Different volumes of reflux were simulated to test the
proposed ultrasonic catheter. The LES was partially open
with the transducer positioned about 4 cm above it. The
diameter of the esophageal body between the transducer and
the LES was between 2 cm and 1 cm for two experimental
trials, respectively. A reflux event was defined within the
period when the ultrasonic transducer was submerged by the
simulated reflux material in the model. Then, the volume of
the simulated reflux was measured by a measuring cup. The
ultrasonic system showed good repeatability and accuracy in
detecting the simulated reflux events and in estimating the
reflux volume from the transducer down to the LES (Table I).
B. In Vivo Human Testing
During the test, reflux events were successfully identified
by the longitudinal ultrasonic channel compared to the
impedance recording (Table II). Dynamic ultrasonic changes Download full-text
associated with reflux presence were observed (Fig. 5).
SUMMARY OF LABORATORY BENCH RESULTS
Trial length (minutes)
Total simulated reflux events
Reflux events detected
Reflux volume (ml)
(Using a measuring cup)
Measured diameter (cm)
Measured height (cm)
Measured volume (ml)
SUMMARY OF IN VIVO HUMAN TESTING RESULTS
Total impedance changes
Total reflux events (MII)
Number of reflux events
Reflux event false negative
Reflux event false positive
Trial 1 Trial 2
Fig. 5. Ultrasonic reflux detection associated with impedance recording
during in vivo human testing. The shadow region in impedance recordings
indicates a liquid reflux event detected. FP indicates false positive.
Referring to Table II, 8/10 spontaneous reflux events were
identified correctly as referred to the MII method. However,
there were 6 false positives detected from the overall 16
impedance changing events. There were 2 false negatives
likely due to the movement of the MII-pH catheter, since the
ultrasonic probe may point at the impedance catheter
randomly. The volume of reflux from the transducer down to
the LES during the testing calculated by the proposed method
was 4.12±0.82 ml.
The goal of the proposed technique was to detect GER
events and to estimate the volume during a certain GER
period. In the described experiments, the GER events were
identified by a reduction of the ultrasonic signal amplitude,
which normally appeared when the ultrasonic transducer
became in contact with the reflux content. Estimation of the
reflux volume was triggered by the output of the reflux
identification algorithm. The described ultrasonic catheter
showed a good correlation between the ultrasonic output and
the reflux presence during in vitro and in vivo testing.
However, relative position of the impedance catheter affected
the echoes resulting in some false negative results.
Encapsulation techniques should be utilized to combine the
ultrasonic and impedance catheters into a single compact case
to avoid this issue in future testing.
Reflux volume estimation results compared very well with
the results measured by another objective independent
technique during the in vitro testing. However, for measuring
the diameter, the longitudinal transducer needs to be
positioned near the center of the lumen of the esophagus. The
movement of the transducer affects the measurement. It is
believed that a longitudinal ultrasonic transducer with a large
angle of divergence could overcome this problem . If the
transducer is not near the center of the lumen, two echo peaks
will be detected, from the average of which the diameter
could be estimated. Moreover, to verify the accuracy of the
reflux volume estimation in vivo, comparative measurement
of reflux volume using other methods needs to be pursued.
An innovative longitudinal ultrasonic catheter design for
dynamic volumetric assessment of gastroesophageal reflux
has been presented and tested in both in vitro and in vivo
human testing. The obtained in vitro results demonstrate the
feasibility of the probe for monitoring the dynamics of reflux
volume in the simulation model. In vivo human testing results
indicate promising suitability of the ultrasonic system for
detecting gastroesophageal reflux events as well as estimating
reflux volume dynamics.
 United State National Institute of Health, “Digestive disease statistics”.
 M. Patti, P. Fisichella, "Esophagus & Diaphragm," In: Current
Diagnosis & Treatment: Surgery, Doherty GM (Ed). [Online].
 W. Barlow, R. Orlando. “The pathogenesis of heartburn in nonerosive
reflux disease: a unifying hypothesis.” Gastroenterology, vol.128, pp.
771-8, Aug. 2005.
 A. Lazarescu, D. Sifrim. “Ambulatory Monitoring of GERD: Current
Technology.” Gastroenterol Clin N Am, vol. 37, pp. 793-805, 2008.
 D. Sifrim. “Relevance of volume and proximal extent of reflux in
gastro-oesophageal reflux disease.” Gut, vol. 54, pp.175-178, Jul. 2005.
 Y. Yorozu N. Tipnis, J. Liu J. Puckett, and R. Mittal. “Common cavity
pressure during gastroesophageal reflux: reassessment using
simultaneous pressure, impedance, and ultrasound imaging.” Am J
Physiol Gastrointest Liver Physiol, vol. 290, pp. 1149-56, 2006.
 R. Mittal, J. Liu, and J. Puckett. “Sensory and Motor Function of the
Esophagus: Lessons from Ultrasound Imaging.” Gastroenterology,
vol.128, pp. 487-497, Aug. 2005.
 W. Yang, T. Fung, K. Chian, et al. “Finite element simulation of food
transport through the esophageal body.” World J Gastroenterol, vol. 13,
pp. 1352-1359, 2007.
 S. Manoli, W. Lochner, S. Oswald, et al. “Estimation of Ventricular
Volume with an Intracardiac Ultrasonic Catheter Catheter,” Pflgers
Arch, vol 349, pp. 369-376, 1974.