Non-restrained measurement of

Young's modulus for soft tissue using a

photoacoustic technique
Cite as: Appl. Phys. Lett. 105, 103707 (2014); https://doi.org/10.1063/1.4893725
Submitted: 11 March 2014 • Accepted: 11 August 2014 • Published Online: 12 September 2014

Naoki Wadamori

ARTICLES YOU MAY BE INTERESTED IN

Syllables intelligibility in relation to the autocorrelation and cepstrum model: The case of
Chinese in Taiwan
Proceedings of Meetings on Acoustics 15, 015002 (2012); https://doi.org/10.1121/1.4733849

Quantitative photoacoustic elasticity and viscosity imaging for cirrhosis detection

Applied Physics Letters 112, 211902 (2018); https://doi.org/10.1063/1.5021675

Ultrasound (US) transducer of higher operating frequency detects photoacoustic (PA) signals
due to the contrast in elastic property
AIP Advances 6, 025210 (2016); https://doi.org/10.1063/1.4942106

Appl. Phys. Lett. 105, 103707 (2014); https://doi.org/10.1063/1.4893725 105, 103707

© 2014 Author(s).
 APPLIED PHYSICS LETTERS 105, 103707 (2014)

Non-restrained measurement of Young’s modulus for soft tissue using

a photoacoustic technique
Naoki Wadamoria)
Department of Electrical, Electronics and Information Engineering, Nagaoka University of Technology,
Nagaoka 940-2188, Japan
(Received 11 March 2014; accepted 11 August 2014; published online 12 September 2014)
A miniaturized sensor was developed to determine the Young’s modulus of tumors based on
photoacoustic spectroscopy. The sensor had a chamber height of 5.3 mm and diameter of 5.8 mm;
thus, this device is smaller than conventional endoscopes, the outer diameters of which are
typically about 10 mm. A non-restrained methodology for determining the Young’s modulus of
tumors was proposed based on the resonance frequency of the photoacoustic signal. The proposed
approach was applied to silicone rubbers with six different Young’s moduli (30, 64, 123, 224, 396,
and 574 kPa) and showed good resolution (62 kPa) and high reproducibility. These results show
that the photoacoustic technique can be applied to mechanically characterize soft tissue by diagnostic
endoscopy. VC 2014 Author(s). All article content, except where otherwise noted, is licensed under a

Creative Commons Attribution 3.0 Unported License. [http://dx.doi.org/10.1063/1.4893725]

Conventional transoral endoscopes can distinguish The amplitude of the PA signal was inversely proportional to
between benign and malignant tumors using electronic chro- the modulation frequency and the peak amplitude was
moscopy and endoscopic ultrasound, and can also be used affected by the mechanical characteristics of the sample sur-
for the early detection of digestive tract neoplasia. Ultrathin face vibration. The resonance frequency was used to deter-
endoscopes, such as transnasal endoscopes, have been devel- mine the Young’s modulus. The proposed approach was
oped to alleviate discomfort during diagnosis and therapy. applied to silicone rubbers with six different Young’s moduli
One of the most significant advantages of transnasal esopha- and was able to determine the moduli with good accuracy.
goscopy is the reduced potential for adverse events, with This approach should be applicable to the mechanical char-
only one major adverse event (an esophageal perforation) acterization of tumors using a smart endoscope.
having been reported in the literature.1 In contrast, conven- The mechanism of PA signal generation is shown in
tional endoscopes have been associated with multiple perfo- Fig. 1.7 When the sample is illuminated by the modulated
rations or even deaths.2 However, many optional diagnostic light, it is heated by non-radiative transitions. This heat is
instruments have been designed to fit through larger side modulated, generating thermal and elastic waves in the sam-
channels, which are only available with conventional endo- ple that propagate from the region of optical interaction and
scope systems; this has limited the applicability of transnasal produce a sound wave. This sound wave is called the PA
endoscopes. signal and is detected by a microphone or a piezoelectric
Information obtained by photoacoustic (PA) spectros- device. Therefore, the PA process involves complex energy
copy is affected not only by the optical absorption coefficient conversions associated with optical, thermal, and acoustic
but also by other physical parameters including the thermal processes.
expansion, specific heat, and Young’s modulus. Various The generation mechanism of the PA signal involves
studies have extensively reported on the elastic and thermo- surface vibrations of the sample that are directed toward the
elastic deformation of materials using diverse techniques.3–5 chamber, as shown in Fig. 2.
Although Macias et al. studied the effect of thermoelastic
bending of clamped circular magnetorheological membranes
on the PA signal theoretically and experimentally,5 their
study did not consider soft materials such as physiological
tissues. Moreover, the technique they used for a clamped
thin plate sample is difficult to use on soft tissue in vivo.
Recently, Yang et al. developed an endoscopic embodiment
for ex vivo PA elastography.6 However, their technique can-
not quantitatively evaluate elastic properties.
In the present study, PA spectroscopy was used to quan-
tify the Young’s modulus of a tissue phantom to demonstrate
its applicability for non-restrained and non-invasive mea-
surement. The PA signal produced by thermoelastic bending
of circular silicone rubbers with free edges was analyzed.

a)
[email protected] FIG. 1. PA signal generation mechanism.

0003-6951/2014/105(10)/103707/4 105, 103707-1 C Author(s) 2014

V
103707-2 Naoki Wadamori Appl. Phys. Lett. 105, 103707 (2014)

FIG. 2. Equivalent acoustic-mechanical model of a PA cell including sur-

face vibration of the sample. FIG. 4. Diagram of the PA system.

The driving-point mechanical impedance of the model To evaluate the PA signal as a function of the modulation
in Fig. 2 can be calculated as follows: frequency, the frequency response of the PA cell was meas-
  ured. A schematic and photograph of the PA cell for a smart
1 endoscope are shown in Figs. 3 and 5, respectively. The PA
Z_ m ¼ rm þ j xm  ; (1)
xCo cell consisted of a silicon-based microelectromechanical sys-
tem (MEMS) microphone and an SMA-type coupler threaded
where rm [N s m1] is the mechanical resistance, m [kg] is onto the SMA connector of the emitting fiber (BWF-1540;
the mass, and Co [N m1] is the compliance. Co is connected B&WTEK). A set of headphones (XBA-10; Sony) was con-
to Cm and CA /S2 in series, where Cm [N m1] is the compli- nected to the open end of the SMA-type coupler threaded onto
ance of the vibrating sample, CA is the acoustic capacitance a dummy plug. The gap between the headphones and the PA
of the chamber, and S [m2] is the effective area of the vibrat- cell was sealed with putty. The frequency was swept at a con-
ing sample. Because the mechanical resistance is very low stant rate from 400 to 2000 Hz with a spectrum analyzer
compared with the mechanical reactance at frequencies (CF7200; Onosokki). The output from the headphones entered
lower than the resonant frequency (i.e., x < x0 ) the PA cell and was recorded by the spectrum analyzer.
The experimental setup shown in Fig. 4 was used to mea-
1
Z_ m ’ j : (2) sure the PA signal in the presence of thermoelastic bending of
xCm the samples. A laser diode with a power of approximately 34
The inverse of the compliance is the spring constant mW and a wavelength of 1550 nm (BWF-1540; B&WTEK)
K ¼ 1/Cm. The Young’s modulus E [Pa] is given by was modulated at a frequency f by a function generator
E ¼ KL/S, where L [m] is the length of the sample and S (WF1974; NF Corporation). Such a high wavelength was
[m2] is the area of the sample. The resonance frequency chosen because it is not affected by water in the tissue. The pen-
depends on the Young’s modulus; a higher Young’s mod- etration depth into the skin is estimated to be 1.4–1.5-mm deep
ulus results in a higher resonance frequency, as expressed at a wavelength of 1550 nm.8 The PA signal was generated by
by optical absorption, detected with the MEMS microphone
(SPW0430HR5HB-B; Knowles), and transduced to an electric
1 signal amplified with a 40-dB preamplifier (ADA4841-1;
x0 ¼ pffiffiffiffiffiffiffiffiffi / E: (3)
mCo Analog Devices). This signal was further amplified with a
36-dB amplifier (PC-SAD8F1-A; SDS) through a low-pass filter
(PC-SAD8F1-A; SDS) and processed by a computer after being
converted into 16 bits by an A/D converter (PC-G164-1 M;
SDS). The spot of the laser beam (100–lm diameter) was

FIG. 5. Photograph of the fiber-coupled PA cell and sample on top of the

FIG. 3. Schematic view of the PA cell for a smart endoscope. PA cell.
103707-3 Naoki Wadamori Appl. Phys. Lett. 105, 103707 (2014)

TABLE I. Properties of the silicone rubbers studied.

Parameter S1 S2 S3

Young’s modulus (kPa) 34 62 114

Mass (g) 15.53 15.69 16.43
Diameter (mm) 44
Thickness (mm) 9.57 9.53 9.58
Parameter S4 S5 S6
Young’s modulus (kPa) 194 324 528
Mass (g) 16.74 17.15 16.95
Diameter (mm) 44
Thickness (mm) 9.57 9.58 9.56

focused directly on the surface of the sample, which was

placed on top of the PA cell as shown in Fig. 5.
The PA cell was used to examine the Young’s modulus
of six silicone rubbers (34, 62, 114, 194, 328, and 528 kPa).
Table I shows the diameters, masses, and thicknesses of the
silicone rubbers.
The frequency response of the PA cell was measured and
the results are presented in Fig. 6. Figure 6 shows that the var- FIG. 7. Amplitude of the measured PA signal as a function of modulation
iation range is about 61 dB. In the range of 400–2000 Hz, the frequency for six different silicone rubber samples.
PA cell exhibited a flat frequency response.
The amplitudes of the PA signal are presented in Fig. 7 sample: a larger Young’s modulus corresponds to a higher
as a function of the modulation frequency for the six silicone resonance frequency.
rubber samples. The frequency response was obtained point- A frequency responsivity of 0.56 kPa Hz–1 was calcu-
by-point by measuring the PA signal at each modulation fre- lated using the resonance frequencies and Young’s moduli in
quency and computing the amplitude of the PA signal by Fig. 8. From these results, the resolution can be estimated to
Fourier transform. The measurements were repeated five be about 62 kPa. The frequency response can be obtained
times for each sample and the average amplitude values are with high reproducibility regardless of the broad peaks
shown in Fig. 7. It can be seen that the PA signal amplitude shown in Fig. 8.
is inversely dependent on the modulation frequency and the The frequency range of biological signals is typically
peak position is affected by the vibration of the sample 0–200 Hz (e.g., pneumogram: 0–40 Hz, electromyogram:
surface. 10–200 Hz). Because this frequency range is lower than the
The maximum value of the detected peak was selected resonant frequencies, the tumor signal can be easily extracted
as the resonance frequency for each sample. The averages from the background.
and standard deviations of the resonance frequencies are Typical values of the Young’s modulus are about 10 kPa
shown in Fig. 8 and Table II for each Young’s modulus. The for parenchyma, 20 kPa for muscle, and 106 kPa for carci-
resonance frequency depends on the Young’s modulus of the noma of the breast.9 Moreover, the Young’s modulus was

FIG. 6. Frequency response of the PA cell. FIG. 8. Young’s modulus as a function of modulation frequency.
103707-4 Naoki Wadamori Appl. Phys. Lett. 105, 103707 (2014)

TABLE II. Experimental values of the resonance frequency for the six rub- smaller than the tip of a conventional endoscope. A simple
ber silicone samples. and reliable methodology was proposed for determining the
Young’s modulus of tumors based on the resonance fre-
Sample No. Young’s modulus (kPa) Resonance frequency (Hz)
quency of the PA signal. The proposed approach was applied
S1 34 530 6 1.14 to silicone rubbers with six different Young’s moduli and
S2 62 574 6 2.92 showed good accuracy and high reproducibility over a wide
S3 114 741 6 0.55 measuring range. The highest accuracy for the measured res-
S4 194 837 6 0.89 onance frequencies and therefore Young’s modulus was
S5 324 1071 6 1.14
obtained for the first observed resonance peak. This new
S6 528 1409 6 0.55
application of the PA technique should encourage its use in
the mechanical characterization of soft tissue in diagnostic
reported to be 69–75 kPa in cholangiocellular carcinoma, endoscopy. In future work, theoretical models of the PA cell
69 kPa in metastatic tumors in the esophagus, and 64–72 kPa that include surface vibration will be developed.
in metastatic tumors in the stomach; the differences in
Young’s modulus were statistically significant (P ¼ .049 by This work was supported by JSPS KAKENHI Grant
the Kruskal–Wallis test).10 Considering these results, our ap- Nos. 24680058 and 24650292 and by STARC Grant No.
paratus is suitable for measuring the wide range of Young’s 1340.
moduli in various tissue types.
1
The proposed methodology based on using the reso- A. Zaman, M. Hahn, R. Hapke, K. Knigge, M. B. Fennerty, and R. M.
nance frequency to determine the Young’s modulus can be Katon, Gastrointest. Endosc. 49, 279 (1999).
2
V. K. Sharma, C. C. Nguyen, M. D. Crowell, D. A. Lieberman, P. D.
used to distinguish tumors from tissue with high reproduci- Garmo, and D. E. Fleischer, Gastrointest. Endosc. 66, 27 (2007).
bility and resolution. The developed device, shown in Fig. 3 3
S. P. Timoshenko and S. Woinowsky-Krieger, The Theory of Plates and
and 5, is smaller than a conventional endoscope, the tip di- 4
Shells (McGraw-Hill, New York, 1959).
ameter of which is typically about 10 mm. Y. Sun and M. Saka, J. Sound Vib. 329, 328 (2010).
5
J. D. Macias, J. Ordonez-Miranda, and J. J. Alvarado-Gil, J. Appl. Phys.
Further work is underway to improve the spatial resolu- 112, 124910 (2012).
tion for determining the size of tumors and to develop a 6
J. Yang, K. Maslov, H. Yang, Q. Zhou, K. K. Shung, and L. V. Wang,
mathematical model to compare the measured and theoreti- Opt. Lett. 34, 1591 (2009).
7
cal data. Taking the heterogeneity of the tissue into consider- A. Rosencwaig, Photoacoustics and Photoacoustic Spectroscopy (Wiley,
New York, 1990).
ation in this method is very complex and requires further 8
M. P. Goldman, R. E. Fitzpatrick, E. V. Ross, S. L. Kilmer, and R. A.
study. Weiss, Lasers and Energy Devices for the Skin (CRC Press, Boca Raton,
In this study, a miniaturized sensor, based on PA spec- 2013).
9
P. Wells and H. Liang, J. R. Soc. Interface 8, 1521 (2011).
troscopy for embedment into endoscopes, was developed to 10
R. Masuzaki, R. Tateishi, H. Yoshida, T. Sato, T. Ohki, T. Goto, H.
measure differences in the Young’s modulus for a tumor Yoshida, S. Sato, Y. Sugioka, H. Ikeda, S. Shiina, T. Kawabe, and M.
mimicking phantom. The developed device was sufficiently Omata, Hepatol. Int. 1, 394 (2007).