Ultrasound in Med. & Biol., Vol. 43, No. 9, pp.

2086–2094, 2017
Ó 2017 World Federation for Ultrasound in Medicine & Biology
Printed in the USA. All rights reserved
0301-5629/$ - see front matter

http://dx.doi.org/10.1016/j.ultrasmedbio.2017.05.010

d Original Contribution

ACOUSTIC AND ELASTIC PROPERTIES OF GLYCEROL IN OIL-BASED GEL

PHANTOMS

LUCIANA C. CABRELLI, FELIPE W. GRILLO, DIEGO R. T. SAMPAIO, ANTONIO A. O. CARNEIRO,

and THEO Z. PAVAN
Departamento de F
ısica, Faculdade de Filosofia, Ci^encias e Letras de Ribeir~ao Preto, Universidade de S~ao Paulo, Ribeir~ao Preto,
SP, Brasil

(Received 17 January 2017; revised 3 May 2017; in final form 3 May 2017)

Abstract—Phantoms are important tools for image quality control and medical training. Many phantom materials
have been proposed for ultrasound; most of them use water as the solvent, but these materials have disadvantages
such as dehydration and low temporal stability if not properly stored. To overcome these difficulties, copolymer-in-
oil gel was proposed as an inert and stable material; however, speed of sound for these materials is still lower than
what is described for most biological tissues. Here, we propose the glycerol dispersion in oil-based gels to modify
the acoustic and elastic properties of copolymer-in-oil phantoms. We manufactured copolymer-in-oil gels using
styrene-ethylene/butylene-styrene (SEBS) in concentrations 8%–15%. We used 2 types of mineral oils with
different viscosities. Glycerol was added in a volume fraction 0%–30% of the total amount of liquid. The acoustic
(i.e., speed of sound, attenuation and backscattering) and the mechanical (i.e., density and Young’s modulus) prop-
erties of the samples were within the range of values observed for soft tissues. The acoustic parameters of the sam-
ples were dependent on oil viscosity and glycerol concentration. The speed of sound ranged 1423 m/s – 1502 m/s,
while the acoustic attenuation and the ultrasonic backscattering increased by adding glycerol. The density and the
Young’s moduli were less affected by the presence of glycerol. We conclude that glycerol can be used to control the
acoustic parameters of copolymer-in-oil gels. Additionally, it opens the possibility of incorporating other oil-
insoluble substances to control further properties of the phantom. (E-mail: theozp@usp.br) Ó 2017 World
Federation for Ultrasound in Medicine & Biology.
Key Words: Tissue-mimicking phantom, Copolymer-in-oil, Glycerol, Ultrasound, SEBS, Elasticity, Backscattering.

INTRODUCTION essential for training procedures, such as biopsy, suturing

and others.
Tissue-mimicking phantoms are important tools to char-
The human body is mostly composed of water;
acterize new imaging techniques, simulate clinical pro-
therefore, water-based materials are usually a good
cedures and provide imaging quality control (Kharine
choice to manufacture tissue-mimicking phantoms. In
et al. 2003; Pogue and Patterson 2006). In
this case phantoms can be manufactured using water as
ultrasonography, phantoms may also be used for
the solvent and the matrix can be a natural polymer,
training clinical procedures, such as biopsy and
such as gelatin and agar, or a synthetic polymer, such as
anesthesia (Culjat et al. 2010; Vieira et al. 2013).
polyvinyl alcohol (PVA) cryogel and polyacrylamide
Ultrasonic properties of these phantoms, such as speed
(Culjat et al. 2010). These materials have been reported
of sound and acoustic attenuation, are expected to be as
to have good tunable acoustic and elastic properties; how-
similar as possible to tissue. The average speed of
ever, can present disadvantages such as low temporal sta-
sound in soft tissue is 1540 m/s (Szabo 2004), varying
bility and mechanical fragility (Culjat et al. 2010;
from 1430 m/s in fat to 1585 m/s in muscle. Some imag-
Kharine et al. 2003; Madsen et al. 1978).
ing modalities, i.e., elasticity imaging, require phantoms
These hydrogels can be manufactured to mimic
presenting realistic elastic behavior, which is also
different viscoelastic properties of tissue. Therefore, a
wide range of papers describing elastography phantoms
have been reported. Gelatin and agar biopolymer have
Address correspondence to: Theo Z. Pavan, Av. Bandeirantes,
3900 - Monte Alegre, Ribeir~ao Preto, SP, CEP 14040-901, Brasil. been extensively studied for elastography phantoms
E-mail: theozp@usp.br (Hall et al. 1997; Madsen et al. 2003, 2005). They can

2086
 Glycerol in phantoms d L. C. CABRELLI et al.
be used for homogenous or heterogeneous phantoms
(Madsen et al. 2003, 2005) and the degree of stiffness
(Hall et al. 1997), elastic non-linearity (Pavan et al.
2010) and viscoelasticity (Nguyen et al. 2014; Zhu
et al. 2015) can be tailored by using different recipes
and fabrication methods. PVA gels have also been
explored for elastography phantoms, where the
mechanical properties can also be adjusted by changing
the concentration and manufacturing procedure (Chu
and Rutt 1997; Eskandari et al. 2008; Pazos et al.
2009). Polyacrylamide (Kumar et al. 2010), gellan gum
(Chen and Shih 2013) and polyvinyl chloride (Li et al.
2016) are other materials that have been explored as
phantoms for elastography.
Oil-based phantoms usually present good temporal
stability and have been proposed for ultrasound (Kondo
et al. 2005; Oudry et al. 2009; Vieira et al. 2013).
Copolymer-in-oil gel is a type of oil-based phantom
that uses mineral oil as the solvent and has been proposed
as an alternative to manufacture phantoms mimicking
soft tissue. Several studies illustrated that these materials
are extremely stable in time and can be stored at room
temperature, requiring no special care (Cabrelli et al.
2017; Oudry et al. 2009). These oil-based gels have
tunable elastic, acoustic and optical properties (Cabrelli
et al. 2017; Oudry et al. 2009); however, they usually
present lower speed of sound than the average value for
soft tissues (Oudry et al. 2009). Solvent properties (e.g.,
density, viscosity, etc.) can influence the gel characteris-
tics (Oudry et al. 2009); furthermore, additives, such as
low-density polyethylene (LDPE), can be used to modify
its acoustical, optical and mechanical properties (Cabrelli
et al. 2017). In Cabrelli et al. (2017), we proposed the use
of styrene-ethylene/butylene-styrene (SEBS) gels for op-
tical imaging and revealed the viability for photoacous-
tics. In that study (Cabrelli et al. 2017) we verified that
variations in SEBS and LDPE amounts had little impact
on the gels’ speed of sound.
Glycerol is a colorless, nontoxic and viscous liquid
that presents speed of sound higher than 1900 m/s and
has been used as an additive to water-based phantoms.
In blood-mimicking fluids, for example, glycerol is added
to water to increase the speed of sound to values more
similar to blood (Ramnarine et al. 1998) and to mimic
the rheologic properties of blood (Oates 1991). It can
be added to agar/gelatin (Madsen et al. 1982, 2005) or
PVA (Cournane et al. 2010) phantoms to increase the
speed of sound with little impact in the acoustic attenua-
tion. Therefore, the aim of this research is to evaluate the
viability of using glycerol as an additive to copolymer-in-
oil gels. First described is the method used to obtain sta-
ble gels using mineral oil and glycerol. Then evaluated
are the acoustic and mechanical properties of gels that
produced varying the amount of glycerol and copolymer
2087
(i.e., SEBS) and using mineral oils with different
viscosities.
MATERIALS AND METHODS
Material preparation
Gel samples using the copolymer SEBS (Kraton G
1650 M, Kraton Polymers, Paul
ınia, SP, Brazil) in min-
eral oil were manufactured. The samples were prepared
mixing the SEBS copolymer, mineral oil and glycerol
and heated to 130
C in a vacuum oven (YK Oven Vac
4, YK Tecnologia, Sapiranga, RS, Brazil). Mineral oil
is the ideal solvent to SEBS copolymer, forming stable
gels. These gels can be easily obtained by raising the
SEBS/mineral oil mixture temperature to 130
C until
complete homogenization. In this case, it is not necessary
to stir the mixture to obtain stable gels. For example, to
obtain 500 mL of SEBS gel, the mixture rests in the vac-
uum oven at 130
C for around 3 h. Since glycerol is not
soluble in oil, the mixture SEBS/mineral oil/glycerol
had to be vigorously stirred at 130
C until a fluid gel
was formed. In the present study, the mixture was manu-
ally stirred using a plastic spatula. For that reason, the
mixture was removed from the vacuum oven and stirred
until a homogenous solution was obtained. This proced-
ure was repeated, every 5 min, until a viscous solution
without any visible solid copolymer was obtained. The
increased viscosity indicates the polymerization of the
gel was complete; therefore, the glycerol droplets were
trapped in the gel matrix. The resultant gel rested in the
oven under vacuum, at 130
C, to release trapped air bub-
bles. Finally it was poured, at 130
C, in molds used for
acoustic and mechanical measurements and cooled at
room temperature to obtain a solid gel. In our experi-
ments we obtained homogenous gels using glycerol in
liquid volume fractions of up to 30%.
We believe that SEBS in mineral oil gel acts as a sus-
pending agent or dispersant for the glycerol (Morrison
et al. 2002), resulting in homogeneous and stable gels
containing uniformly dispersed glycerol. For more details
about SEBS gels structure, the reader is referred to
Sudarsan et al. 2005 and Yu et al. 2008.
To evaluate the influence of the mineral oil viscosity
on the mechanical and the acoustic properties, the sam-
ples were produced using two types of pharmaceutical
grade paraffinic mineral oils with different viscosities:
EMCAplus 070 (Oxiteno, S~ao Paulo, SP, Brazil) with
lower viscosity and Naturol (Farmax, Divin
opolis, MG,
Brazil) with higher viscosity. The dynamic viscosity of
the mineral oils and the glycerol was measured using a
rheometer (RheolabQC, Anton Paar, Graz, Styria,
Austria) at a shear rate of 500 s21 at 25
C. The dynamic
viscosities of the mineral oils were 19.7 mPa$s and
48.8 mPa$s and that of glycerol was 849.3 mPa$s.
2088 Ultrasound in Medicine and Biology Volume 43, Number 9, 2017

To evaluate the influence of glycerol on the mechan-
ical and acoustical properties of the copolymer-in-oil
phantoms, volume fractions of glycerol varying 0%–
30% were used. The upper limit to obtain stable and ho-
mogeneous gels using the method described here was
30% of glycerol. The use of surfactants could be a future
study in an attempt to use a higher concentration of glyc-
erol. Three groups of samples using different concentra-
tions of SEBS and glycerol were prepared. A total of 16
gel samples using 2 types of mineral oils, with different
viscosities, were produced. The composition of the man-
ufactured gel samples are shown in Table 1.
Density measurements
To evaluate the densities of the samples, the volume
and mass of cylindrical samples with 2 cm in diameter
and 1.6 cm in height were measured. Volume measure-
ments were performed by mineral oil displacement of
the samples with a measuring cylinder. The samples
masses were measured with a precision balance (model
bl 3200 h, Shimadzu, Kyoto, Kansai, Japan).
minimize the friction with the samples during the tests.
The data acquisition started at the moment the compression
plate touched the sample. This contact point was automat-
ically identified by the equipment. The Young’s modulus
was calculated using the stress-strain curves.
Acoustic characterization
To evaluate the acoustic properties, the gel was
poured into PVC cylindrical molds with 7.5 cm in diam-
eter and 2.5 cm thick, sealed with PVC wrap film.
The speed of sound was estimated using the substi-
tution technique (Madsen et al. 1999) in an acrylic tank,
with dimensions of 11 cm 3 13 cm 3 11 cm, filled
with degassed water, using 2 narrowband transducers
with central frequencies of 1 MHz. The distance between
the transducers was 99.98 6 0.05 mm. The temperature
of the water and the samples was 21
C. The speed of
sound was calculated using
cw
cs 5
; (1)
11cwdDt

where cw is the water speed of sound, Dt is the time shift

Elastic measurements after immersing the sample in the ultrasonic wave path
To evaluate the elasticity of each sample, the stress- and d is the sample thickness.
strain curves were measured at room temperature via me- The acoustic attenuation coefficient was evaluated
chanical uniaxial compression using a texture analyser using 5 immersion unfocused narrowband ultrasound
(TA.XTplus, Stable Micro Systems, Godalming, Surrey, transducers (Panametrics, Olympus NDT Inc., Waltham,
UK) equipped with a 5-kg load cell. The same cylinder sam- MA, USA) with frequencies ranging 1 MHz–10 MHz be-
ples used to determine the density were tested. The samples ing driven by tone bursts of 12 cycles. Table 2 shows the
were compressed (deformations up to 18%) at a constant nominal and peak frequencies, 26 dB bandwidth and di-
speed of 0.1 mm/s using an acrylic platen with 40 mm in mensions of each transducer. The transmitted ultrasonic
diameter. Mineral oil was used to lubricate the platens to waves were detected by a needle hydrophone (0.5 mm
needle sensor, sn 1861, Precision Acoustics, Dorchester,
Table 1. Composition of the samples where the Dorset, UK) positioned opposite to the ultrasonic trans-
concentrations of SEBS are given in weight percentage of ducers. The hydrophone system was provided with a cali-
the whole material, while for glycerol it is given as bration certificate covering the frequency range
volume fraction of the total amount of liquid 1–20 MHz. In the frequency range 1–10 MHz, the sensi-
Oil viscosity tivity fluctuates 345–500 mV/MPa. The samples were
Set Phantom (mPa$s) SEBS (%) Glycerol (%) positioned between the ultrasound transducer and the hy-
A A1 19.7 10 0
drophone, with both samples and hydrophone located in
A2 19.7 10 5 the far field zone of the transducers. In all measurements
A3 19.7 10 10 we ensured the faces of the samples were parallel to the
A4 19.7 10 15
A5 19.7 10 20
A6 19.7 10 30 Table 2. Properties of the ultrasonic transducers used in
B B1 48.8 10 0
B2 48.8 10 5
the acoustic measurements
B3 48.8 10 10
Nominal Peak
B4 48.8 10 15
frequency frequency Diameter 26 dB
B5 48.8 10 20
(MHz) (MHz) size (in) bandwidth
B6 48.8 10 30
C C1 48.8 8 15
1.0 0.98 0.5 62.08%
C2 48.8 10 15
3.5 3.22 0.25 59.66%
C3 48.8 12 15
5.0 5.00 0.375 59.71%
C4 48.8 15 15
7.5 7.13 0.50 61.38%
10.0 10.0 0.25 70.76%
SEBS 5 styrene-ethylene/butylene-styrene.
 Glycerol in phantoms d L. C. CABRELLI et al.
ultrasonic transducers. The acoustic attenuation coeffi-
cient was obtained by
 
20 A0
a 5 log ; (2)
d A
where A0 and A are, respectively, the ultrasonic pressure
amplitude in volts acquired in water and after positioning
the samples between the hydrophone and the ultrasonic
transducer. The dependence of the acoustic attenuation
coefficient a and the ultrasonic frequency f were analyzed
by a( f ) 5 a0f n.
To understand the influence of glycerol dispersion
on ultrasound wave scattering, the apparent integrated
backscattering (AIB) of set A samples were evaluated.
This parameter is defined as the frequency average of
the ratio between the power spectrums of the sample
and a reference; and has been commonly used to charac-
terize tissue (Hoffman et al. 2011; Jiang et al. 2014; Naito
et al. 1994). These samples were scanned using a linear
array (L14-5/38) connected to an ultrasound machine
(Sonix RP, Ultrasonix, Richmond, BC, Canada). The
beamformed radio frequency (RF) backscattered signals
from the phantoms were acquired using the Texo
software development kit (version 5.7.4, Ultrasonix). To
calculate AIB, 10 frames composed of 20 scan-lines
were acquired for each phantom. After averaging the 10
frames, the RF signals were truncated using a hamming
window in 7.5 ms segments. The center of each segment
was positioned at 0.6-cm depth. The reference RF back-
scattered signal was obtained from a stainless steel plate
positioned at the same depth in a water tank. AIB was
estimated over the 26 dB bandwidth (Df26
dB 5 fH2fL) of the transducer with a measured center fre-
quency fc 5 4.0 MHz, according to
ð fH  
1 Pph ðf Þ
AIB 5 10 log df ; (3)
Df26dB fL Pr ðf Þ
where Pph and Pr are the power spectrum of the backscat-
tered RF signals of the phantom and the reference,
respectively.
Statistical analysis
Statistical tests were performed using analysis of
variance (one-way ANOVA) and post hoc analysis to
confirm differences between samples. We used a signifi-
cance value of 0.05 for all statistical tests. The statistical
analyses were performed using R statistical computing
environment.
RESULTS AND DISCUSSIONS
The gels manufactured only with mineral oil,
without glycerol, were translucent to visible light, as
2089
shown in Figure 1. The presence of glycerol resulted in
higher opacity of the material (Fig. 1). Optical micro-
scopy images of the gels revealed a uniform distribution
of glycerol microspheres with diameters ranging
5–30 mm.
Density
Table 3 shows the densities measured for the gel
samples. The samples had slightly increments in the den-
sity by increasing the glycerol percentage. Also, samples
of set A manufactured with the lower viscosity oil were
slightly less dense than samples of set B. However, these
differences, between oil type (sets A and B) or between
samples of the same set, were not statistically significant
(p . 0.05). These values are similar to that of soft tissues
of 1 g/cm3 (Culjat et al. 2010).
Speed of sound
The speeds of sound for the mineral oils with higher
and lower viscosities are 1432 m/s and 1395 m/s, respec-
tively. Figure 2 shows the speed of sound of the 3 sets of
materials. In all measurements the instrumental uncer-
tainty was 1.5 m/s. The results suggest that speed of
sound increased with oil viscosity, because samples of
set B were made with the oil of higher viscosity. This
result is in agreement with that observed by Oudry
et al. (2009). According to Parthasarathy and Bakhshi
(1953), a non-linear relationship exists between viscosity
and speed of sound; however, this relationship is depen-
dent on the type of substance analysed and density of
the material.
The speed of sound of sets A and B were fitted by a
linear function, resulting in cs 5 (1420.5 1 1.60x) m/s
for A and cs 5 (1460 1 1.37x) m/s for B, where x is
glycerol concentration. The slopes of the 2 fits show
that glycerol concentration similarly increased the
Fig. 1. Samples of set B. Sample B1 with 0% of glycerol (left),
sample B2 with 5% of glycerol (center) and sample B6 with
30% of glycerol (right). The presence of glycerol modified
some properties of the materials, e.g., adding glycerol increased
the phantom’s opacity.
2090 Ultrasound in Medicine and Biology Volume 43, Number 9, 2017

Table 3. Density of the gel samples phantom and the error bars are the combined uncer-
Density Density Density
tainties. To improve visualization, the error bars for set
Sample (g/cm3) Sample (g/cm3) Sample (g/cm3) C results are not shown (Fig. 3c). Table 4 shows the values
obtained for a0 and n of all samples. The acoustic atten-
A1 0.77 6 0.11 B1 0.84 6 0.12 C1 0.89 6 0.12
A2 0.77 6 0.11 B2 0.87 6 0.12 C2 0.87 6 0.12
uation was clearly increased with glycerol concentration
A3 0.78 6 0.11 B3 0.87 6 0.12 C3 0.88 6 0.12 in the samples of sets A and B. Samples A6 and B6 ex-
A4 0.81 6 0.12 B4 0.86 6 0.12 C4 0.89 6 0.12 hibited the highest attenuations. This result was expected
A5 0.82 6 0.12 B5 0.90 6 0.13
A6 0.88 6 0.13 B6 0.93 6 0.13
because glycerol viscosity is higher than those for min-
eral oils 1 and 2. Samples of sets A and B illustrated a
monotonically increase of a0 with the concentration of
speed of sound for the two set of samples. The speed of glycerol. The acoustic attenuation was consistently
sound achieved the highest value of 1502 m/s for the higher for samples of sets B compared with the samples
sample B6, increasing 41 m/s if compared with sample of set A manufactured using the same composition of
B1. Similar variation in speed of sound was observed in SEBS and glycerol. Post hoc analysis suggested differ-
samples of set A. In this case, the speed estimated for ences between samples with 0%–20% glycerol
sample A1 was 1423 m/s, while for sample A6 it was (p , 0.05). This result is also expected because of higher
1473 m/s. Increasing SEBS concentration, for a fixed viscosity of the mineral oil used to produce the samples of
amount of glycerol (15%), increased the speed of sound set B. The results obtained for set C show that varying the
from 1474 m/s (8% of SEBS) to 1481 m/s (15% of amount of SEBS had little impact in the attenuation coef-
SEBS), resulting in cs 5 (1467 1 1.03y) m/s, where y ficient with no significant statistical differences between
is SEBS concentration. Therefore, the slope of the speed the samples (p . 0.05). No trends in the non-linear
of sound versus concentration for the glycerol in SEBS dependence of acoustic attenuation with frequency were
gels is much lower for increased concentration of SEBS. observed for different concentrations of glycerol and
Although lower than the average values adopted for SEBS; values of n varied between 1.61 and 1.86. Statisti-
soft tissues (1540 m/s) (Szabo 2004), the range of values cal tests confirmed this parameter was not dependent
obtained here is in agreement with a variety of soft tissues (p . 0.05) on the amount of glycerol and SEBS or oil
such as fat and breast tissues (Szabo 2004). For example, type.
the mean value for breast tissue can vary with age from The literature reports a0 for soft tissues varying from
1490 m/s to 1450 m/s for women of 25–30 y and 0.15 dB/cm/MHz2n for blood to 0.75 dB/cm/MHz2n for
.65 y, respectively (Katz-Hanani et al. 2014). These breast tissue (Szabo 2004), which is in the range of values
values are within the range of the glycerol in SEBS gels obtained here. The values obtained for n are equivalent to
manufactured in this study. those reported for kidney and breast, but still higher than
other soft tissues, such as liver and muscle (Szabo 2004).

Acoustic attenuation Apparent integrated backscattering

Acoustic attenuation measurements were performed Figure 4 shows B-mode images of the phantoms A1,
in the range of 1–10 MHz for each set of samples and the A2 and A4. Figure 4a demonstrates absence of backscat-
results are shown in Figure 3. The values are the average tered ultrasonic waves from the phantom made with no
of 5 measurements taken at different positions of the addition of glycerol (sample A1). In this image we can

a b 1500
1500 Set A Set C
Set B
Speed of sound (m/s)

1490
Speed of sound (m/s)

1480
1480
1460
1470
1440
1460
1420
1450
0 5 10 15 20 25 30 7 8 9 10 11 12 13 14 15 16
Glycerol (%) SEBS (%)

Fig. 2. Speed of sound of (a) samples of sets A and B; (b) samples of set C.
 Glycerol in phantoms d L. C. CABRELLI et al. 2091

Fig. 3. Acoustic attenuation of samples of (a) set A, (b) set B and (c) set C.

see only the reflection of the phantom’s surface. The im- all very similar (data not shown), resulting in poor
ages in Figures 4(b, c) show a speckled pattern because of graphic presentation, which suggests no relation be-
scattered waves caused by the presence of dispersed glyc- tween Young’s modulus and oil type or glycerol concen-
erol droplets inside the phantoms. Dispersing oil in tration. The samples displayed non-linear elastic
gelatin/agar gels is a common strategy used to adjust behavior; therefore, the Young’s modulus was estimated
the acoustic and mechanical properties of water-based fitting a linear function 0%–2.5% of strain. Table 5
phantoms. The oil droplets formed in these gels also in- shows the Young’s moduli estimated for all samples.
crease the ultrasonic scattering (Madsen et al. 1982; The Young’s moduli of samples are within the range
Pavan et al. 2012). found for a great variety of soft tissues (Fung 2013;
Figure 5 shows AIB estimated for the samples of set Krouskop et al. 1998). Further adjustment to
A. The graph shows an increment of about 20 dB in AIB manufacture softer or stiffer samples can be achieved
for the phantoms containing glycerol. There is no by using SEBS concentrations below or above the
apparent dependence of AIB with glycerol concentration. limits adopted here. Varying the SEBS concentration is
Because the backscattered signal from sample A1 can be also related to small variations in the speed of sound
considered insignificant (confirmed by visual inspection and attenuation of the material (Cabrelli et al. 2017;
of the RF signals), we can assume values close to Oudry et al. 2009). All samples exhibited a weak
265 dB as the noise level of our data. Variations in degree of elastic non-linearity, which can be a limitation
AIB as a result of changes in SEBS concentration were to adequately reproduce the pronounced non-linear
insignificant because the glycerol is the source of behavior of some tissues.
backscattering. The Young’s modulus of set C samples increased
with SEBS concentration following a power law Y 5
(0.8 x1.65) kPa. In Oudry et al. (2009) the stiffness of
Elasticity
the gels followed a power law with SEBS concentra-
Figure 6 shows the stress-strain curves for the sam-
tion with an exponent of 2.46, suggesting that adding
ples of set C. Stress-strain curves for sets A and B were
glycerol to SEBS gels decreases the value of the
exponent. Although more results are needed to
Table 4. Values of a0 and n estimated for all samples.
confirm this speculation, this result is in interesting
Sample a0 (dB/cm/MHz2n) n agreement with the results found in Pavan et al.
(2010). However, in Pavan et al. (2010) the authors
A1 0.10 6 0.01 1.78 6 0.04
A2 0.15 6 0.02 1.73 6 0.05 analysed the stiffness of agar/gelatin gels containing
A3 0.19 6 0.02 1.78 6 0.04 dispersed oil droplets varying the amount of agar
A4 0.21 6 0.03 1.81 6 0.07
and found a decrease in the exponent by increasing
A5 0.29 6 0.04 1.86 6 0.06
A6 0.52 6 0.07 1.83 6 0.07 the amount of oil in the dispersion. The Young’s
B1 0.18 6 0.01 1.88 6 0.03 modulus of sets A and B samples did not depend on
B2 0.26 6 0.02 1.76 6 0.03
B3 0.39 6 0.02 1.76 6 0.02
the mineral oil viscosity (p . 0.05) and increased
B4 0.45 6 0.02 1.71 6 0.02 around 10 kPa, in both cases (27 kPa–37 kPa), by
B5 0.50 6 0.04 1.76 6 0.03 increasing the glycerol concentration 0%–30%. On
B6 0.59 6 0.03 1.86 6 0.03
C1 0.46 6 0.05 1.62 6 0.07
the other hand, adding oil to gelatin/agar gels
C2 0.42 6 0.04 1.66 6 0.06 decrease the stiffness of the samples (Madsen et al.
C3 0.42 6 0.06 1.73 6 0.08 2003; Pavan et al. 2010). Post hoc analysis showed
C4 0.42 6 0.07 1.73 6 0.07
differences between samples without glycerol (A1
2092 Ultrasound in Medicine and Biology Volume 43, Number 9, 2017

Fig. 4. Ultrasound B-mode images of phantoms (a) A1, (b) A2 and (c) A4. (The images are 2-cm depth.) The image of
sample A1 shows negligible backscattering, while the images of phantoms A2 and A4 demonstrate enhancement in back-
scattering by adding glycerol to the SEBS gels.

and B1) and samples with glycerol (p , 0.05).
Significant difference was found between A6 (30%
of glycerol) and the other samples in set A, while
no differences were found between samples A2–A5.
The phantoms evaluated in this study have shown
good stability. For example, all samples were retested af-
ter 5 mo and the Young’s moduli did not vary more than
6 kPa. Also, SEBS phantoms with dispersed glycerol
droplets prepared about 1 y ago and stored with no special
care still look and feel as new.
Here we show the viability of producing stable oil-
based gels with dispersed glycerol in concentrations up
to 30%. The influence of glycerol in the acoustic and me-
chanical properties of the gel were described. Future
studies will focus on evaluating the acoustic properties
for frequencies above 10 MHz and the viscosity of the
phantoms. Besides using glycerol to tune the acoustic
behavior of the gels, the presence of this substance can
be explored to incorporating other oil-insoluble liquids
and solids to the phantoms. This strategy can be adopted
to further adjust the physical and chemical properties of
Set A
-45
-50
the gels to be used in a variety of imaging modalities.
Currently, our group is investigating how this strategy
can be useful to the production of SEBS gel phantoms
for optical and magnetic resonance imaging.
CONCLUSION
Tissue-mimicking materials composed of SEBS
copolymer and mineral oil, using glycerol as an additive
for tuning the acoustical properties, were developed. In
all samples the glycerol was well dispersed, resulting in
homogeneous and stable gel phantoms. Increasing the
concentration of glycerol increased the speed of sound
and acoustic attenuation with a minor effect on the den-
sity and Young’s modulus of the samples. In this study,
the speed of sound up to 1503 m/s was obtained. Ultra-
sound images and apparent integrated backscattering
measurements showed that the presence of glycerol
enhanced ultrasonic scattering. The speed of sound and
acoustic attenuation of the samples also depended on
the mineral oil viscosity. In addition, to modify ultrasonic
and mechanical properties of the gels, mixing glycerol in
the samples opens the possibility of incorporating other
15
C1
C2
C3
Stress (kPa)

10 C4
AIB (dB)

-55

-60
5
-65

-70
0
0 5 10 15 20 30
Glycerol (%) 0 2 4 6 8 10 12 14 16 18
Strain (%)
Fig. 5. Apparent integrated backscattering obtained for samples
of set A. Fig. 6. Stress-strain curves obtained for the samples of set C.
 Glycerol in phantoms d L. C. CABRELLI et al. 2093

Table 5. Young’s modulus* Jiang Y, Liu C, Li R, Wang W-P, Ding H, Qi Q, Ta D, Dong J, Wang W.
Analysis of apparent integrated backscatter coefficient and backscat-
Sample Young’s modulus (kPa) tered spectral centroid shift in calcaneus in vivo for the ultrasonic
evaluation of osteoporosis. Ultrasound Med Biol 2014;40:
A1 26.0 6 0.6 1307–1317.
A2 32.3 6 0.8 Katz-Hanani I, Rothstein T, Gaitini D, Gallimidi Z, Azhari H. Age-
A3 34.3 6 0.2 related ultrasonic properties of breast tissue in vivo. Ultrasound
A4 34.1 6 0.7 Med Biol 2014;40:2265–2271.
A5 34.8 6 0.6 Kharine A, Manohar S, Seeton R, Kolkman RGM, Bolt RA,
A6 37.5 6 0.4 Steenbergen W, de Mul FF. Poly(vinyl alcohol) gels for use as tissue
B1 27.7 6 0.8 phantoms in photoacoustic mammography. Phys Med Biol 2003;48:
B2 34.2 6 0.5 357–370.
B3 32.3 6 0.5 Kondo T, Kitatuji M, Kanda H. New tissue mimicking materials for ul-
B4 35.0 6 0.4 trasound phantoms. IEEE Ultrasonics Symposium 2005;3:
B5 36.8 6 0.6 1664–1667.
B6 37.2 6 0.5 Krouskop TA, Wheeler TM, Kallel F, Garra BS, Hall T. Elastic moduli
C1 25.7 6 0.5 of breast and prostate tissues under compression. Ultrason Imaging
C2 35.0 6 0.4 1998;20:260–274.
C3 50.3 6 0.5 Kumar K, Andrews ME, Jayashankar V, Mishra AK, Suresh S. Measure-
C4 71.4 6 0.7 ment of viscoelastic properties of polyacrylamide-based tissue-
mimicking phantoms for ultrasound elastography applications.
* Estimated from a linear fit of the stress-strain curves. In this case IEEE Trans Instrum Meas 2010;59:1224–1232.
strains up to 2.5% were considered. Li W, Belmont B, Greve JM, Manders AB, Downey BC, Zhang X, Xu Z,
Guo D, Shih A. Polyvinyl chloride as a multimodal tissue-
mimicking material with tuned mechanical and medical imaging
oil-insoluble substances, which can be used to adjust a va- properties. Med Phys 2016;43:5577–5592.
Madsen EL, Dong F, Frank GR, Garra BS, Wear KA, Wilson T,
riety of chemical and physical properties for different im- Zagzebski JA, Miller HL, Shung KK, Wang SH, Feleppa EJ,
aging modalities. Liu T, O’Brien WD, Topp KA, Sanghvi NT, Zaitsev AV, Hall TJ,
Fowlkes JB, Kripfgans OD, Miller JG. Interlaboratory comparison
of ultrasonic backscatter, attenuation, and speed measurements.
Acknowledgments—This study was supported by CAPES, CNPq grant J Ultrasound Med 1999;18:615–631.
476671/2013-2, FAPESP grant 2013/18854-6 and USP NAP-FisMed Madsen EL, Frank GR, Krouskop TA, Varghese T, Kallel F,
grant. We thank professor Delia Rita Tapia Bl
acido in the Chemistry Ophir J. Tissue-mimicking oil-in-gelatin dispersions for use
Department, Faculdade de Filosofia, Ci^encias e Letras de Ribeir~ao in heterogeneous elastography phantoms. Ultrason Imaging
Preto, Universidade de S~ao Paulo, Ribeir~ao Preto, SP, Brasil, for allow- 2003;25:17–38.
ing us to use the texture analyser, and professor Luciano Bachmann in Madsen EL, Hobson MA, Shi H, Varghese T, Frank GR. Tissue-
the Physics Department, Faculdade de Filosofia, Ci^encias e Letras de mimicking agar/gelatin materials for use in heterogeneous elastog-
Ribeir~ao Preto, Universidade de S~ao Paulo, for taking the microscopy raphy phantoms. Phys Med Biol 2005;50:5597–5618.
images of the phantom. Madsen EL, Zagzebski JA, Banjavie RA, Jutila RE. Tissue mimicking
materials for ultrasound phantoms. Med Phys 1978;5:391–394.
Madsen EL, Zagzebski JA, Frank GR. Oil-in-gelatin dispersions for use
REFERENCES as ultrasonically tissue-mimicking materials. Ultrasound Med Biol
1982;8:277–287.
Cabrelli LC, Pelissari PIBGB, Deana AM, Carneiro AAO, Pavan TZ. Morrison DS, Lu L, Ray RH. Hydrocarbon gels as suspending and
Stable phantom materials for ultrasound and optical imaging. Phys dispersing agents and products. US patent 6 433 068. August
Med Biol 2017;62:432–447. 13, 2002.
Chen RK, Shih AJ. Multi-modality gellan gum-based tissue-mimicking Naito J, Masuyama T, Tanouchi J, Mano T, Kondo H, Yamamoto K,
phantom with targeted mechanical, electrical, and thermal proper- Nagano R, Hori M, Inoue M, Kamada T. Analysis of transmural
ties. Phys Med Biol 2013;58:5511–5525. trend of myocardial integrated ultrasound backscatter for differenti-
Chu KC, Rutt BK. Polyvinyl alcohol cryogel: An ideal phantom mate- ation of hypertrophic cardiomyopathy and ventricular hypertrophy
rial for MR studies of arterial flow and elasticity. Magn Reson Med due to hypertension. J Am Coll Cardiol 1994;24:517–524.
1997;37:314–319. Nguyen MM, Zhou S, Robert J, Shamdasani V, Xie H. Development of
Cournane S, Cannon L, Browne JE, Fagan AJ. Assessment of the accu- oil-in-gelatin phantoms for viscoelasticity measurement in ultra-
racy of an ultrasound elastography liver scanning system using a sound shear wave elastography. Ultrasound Med Biol 2014;40:
PVA-cryogel phantom with optimal acoustic and mechanical prop- 168–176.
erties. Phys Med Biol 2010;55:5965–5983. Oates CP. Towards an ideal blood analogue for Doppler ultrasound
Culjat MO, Goldenberg D, Tewari P, Singh RS. A review of tissue sub- phantoms. Phys Med Biol 1991;36:1433–1442.
stitutes for ultrasound imaging. Ultrasound Med Biol 2010;36: Oudry J, Bastard C, Miette V, Willinger R, Sandrin L. Copolymer-in-oil
861–873. phantom materials for elastography. Ultrasound Med Biol 2009;35:
Eskandari H, Salcudean SE, Rohling R. Viscoelastic parameter estima- 1185–1197.
tion based on spectral analysis. IEEE Trans Ultrason Ferroelectr Parthasarathy S, Bakhshi NN. Relation between velocity of sound and
Freq Control 2008;55:1611–1625. viscosity in liquids. Proc Phys Soc Sect B 1953;66:368.
Fung YC. Biomechanics: Mechanical properties of living tissues. New Pavan TZ, Madsen EL, Frank GR, Carneiro AAO, Hall TJ. Nonlinear
York: Springer Science1Business Media; 2013. elastic behavior of phantom materials for elastography. Phys Med
Hall TJ, Bilgen M, Insana MF, Krouskop TA. Phantom materials for Biol 2010;55:2679–2692.
elastography. IEEE Trans Ultrason Ferroelectr Freq Control 1997; Pavan TZ, Madsen EL, Frank GR, Jiang J, Carneiro AAO, Hall TJ.
44:1355–1365. A nonlinear elasticity phantom containing spherical inclusions.
Hoffman JJ, Johnson BL, Holland MR, Fedewa RJ, Nair A, Miller JG. Phys Med Biol 2012;57:4787–4804.
Layer-dependent variation in the anisotropy of apparent integrated Pazos V, Mongrain R, Tardif JC. Polyvinyl alcohol cryogel: Optimizing
backscatter from human coronary arteries. Ultrasound Med Biol the parameters of cryogenic treatment using hyperelastic models.
2011;37:632–641. J Mech Behav Biomed Mater 2009;2:542–549.
2094 Ultrasound in Medicine and Biology
Pogue BW, Patterson MS. Review of tissue simulating phantoms for op-
tical spectroscopy, imaging and dosimetry. J Biomed Opt 2006;11:
041102–041116.
Ramnarine KV, Nassiri DK, Hoskins PR, Lubbers J. Validation of a new
blood-mimicking fluid for use in Doppler flow test objects. Ultra-
sound Med Biol 1998;24:451–459.
Sudarsan AP, Wang J, Ugaz VM. Thermoplastic elastomer gels: An
advanced substrate for microfluidic chemical analysis systems.
Anal Chem 2005;77:5167–5173.
Szabo TL. Diagnostic ultrasound imaging—Inside out. Academic Press
Series in Biomedical Engineering. Boston, MA: Elsevier; 2004.
Volume 43, Number 9, 2017
Vieira SL, Pavan TZ, Junior JE, Carneiro AAO. Paraffin-gel tissue-
mimicking material for ultrasound-guided needle biopsy phantom.
Ultrasound Med Biol 2013;39:2477–2484.
Yu X, Zheng Z, Lu J, Wang X. Morphology and mechanical properties
of styrene–ethylene/butylene–styrene triblock copolymer/high-
density polyethylene composites. J Appl Polym Sci 2008;107:
726–731.
Zhu Y, Dong C, Yin Y, Chen X, Guo Y, Zheng Y, Shen Y, Wang T,
Zhang X, Chen S. The role of viscosity estimation for oil-in-
gelatin phantom in shear wave based ultrasound elastography. Ultra-
sound Med Biol 2015;41:601–609.