Optoacoustic (OA) methods have become powerful tools in biomedical research capable of retrieving functional information from biological tissues in vivo. Acquisition of OA signals generally relies on direct physical contact of a transducer or an acoustic coupling medium with the tissue surface, which prevents applicability, e.g., in open surgeries or wounded tissues. Non-contact OA imaging has been achieved with air-coupled piezoelectric transducers, which provide a straightforward approach for remote sensing of ultrasound vibrations. However, sensitivity was hampered by a suboptimal alignment between the illumination and detection fields. Herein, we devised an air-coupled transducer featuring a central aperture for light delivery with coaxially aligned optical and acoustic foci, thus providing optimal sensitivity for OA signal detection. Imaging of phantoms and a mouse ear in vivo is showcased by raster-scanning the transducer with light being delivered through a multimode optical fiber.

Optoacoustic (OA) imaging and sensing provides otherwise unavailable functional information from biological tissues by capitalizing on the specific light absorption properties of endogenous chromophores in living organisms. For example, distinctive absorption spectra in oxygenated and deoxygenated states of hemoglobin facilitate otherwise unattainable oxygen saturation readings, which can serve as bio-markers for many diseases.1,2 OA further enables identification of melanin, lipids, and other endogenous chromophores as well as coagulated tissues.3,4 These unique properties have fostered the development of multiple types of OA systems increasingly being used in preclinical and clinical studies.5,6

Acquisition of OA (pressure) signals is generally done with piezoelectric ultrasound transducers. These represent a mature technology providing sufficient robustness and sensitivity for detecting signals generated at depths up to several centimeters within biological tissues with laser energy levels conforming to safety standards.7 Several methods based on optical sensing of ultrasound have alternatively been suggested for acquisition of OA signals.8 Optical detectors complement the drawbacks of piezoelectric sensors by facilitating light transmission and miniaturization. Moreover, optical sensing has been used for remote detection of OA responses.9–11 Avoiding contact with sensitive tissues is essential in some applications, such as open surgeries or monitoring of burns. The optical properties of these and other tissues are, however, highly heterogeneous, leading to unstable responses in optical readings. This problem can be overcome by collecting the OA signals coupled to air, which can be achieved, e.g., with gas-coupled laser acoustic detection12 or with a microphone in a resonator chamber.13 Alternatively, air-coupled piezoelectric transducers operating in the MHz range have been suggested for non-contact detection of laser-generated ultrasound. Shock waves generated by laser ablation of tissues could easily be detected with these types of transducers.14 The feasibility of non-contact detection of OA responses was also demonstrated in tissue-mimicking phantoms15 and in hidden paint layers.16 Focused transducers further enabled the formation of OA images with an enhanced sensitivity.17 Imaging of superficial tissues has also been achieved by focusing the excitation light beam.18 Overall, air-coupled transducers provide a robust way to detect OA signals, but their sensitivity is very low. Herein, we introduce a focused air-coupled transducer featuring a central aperture enabling coaxial alignment of the excitation light beam and the acoustic sensitivity field for an optimal sensitivity. We analyze the performance of this approach for OA imaging and discuss the potential applications based on the results obtained in phantoms and in a mouse ear in vivo.

The layout of the experimental system used for the detection of OA signals is schematically depicted in Fig. 1(a). Basically, the output beam of an optical parametric oscillator (OPO) laser (Innolas GmbH, Krailling, Germany) tuned to an optical wavelength of 532 nm and a pulse repetition frequency (PRF) of 100 Hz was coupled into a multimode optical fiber with a core diameter of 105 µm and a numerical aperture (NA) of 0.22 (FG105UCA, Thorlabs, Inc., New Jersey, USA) and focused on the surface of the sample of interest. The generated OA signals were amplified and digitized at 100 MSPS with a specifically designed data acquisition system (DAQ) with 14 bit vertical resolution (Dasel SL, Madrid, Spain). The Q-switch output of the laser was used to trigger the DAQ, and the acquired data were transferred to a personal computer (PC) via Ethernet connection at 100 Mbit/s. A more detailed representation of the air-coupled transducer is further shown in the inset of Fig. 1(a). Basically, the output of the multimode fiber is fixed inside the aperture of the transducer. A microlens (LA1039, Thorlabs, Inc., New Jersey, USA) was attached to the edge of the aperture to focus light at a distance of ∼20 mm, corresponding to the acoustic focus of the transducer.

FIG. 1.

Optoacoustic imaging via coaxial alignment of the light beam and the acoustic focus of the air-coupled transducer. (a) Layout of the imaging system. The inset shows an internal section of the transducer. L—laser, CL—coupling lens, CS—coupling stage, MF—multimode fiber, T—trigger, DAQ—data acquisition system, PC—personal computer, ACT—air-coupled transducer, UW—ultrasound waves, LB—light beam, IS—imaging sample, ML—microlens, PS—piezocomposite disk, and ACL—air coupling layer. (b) Layout of the ray-tracing-based simulation. (c) Simulated irradiance at the focusing plane.

FIG. 1.

Optoacoustic imaging via coaxial alignment of the light beam and the acoustic focus of the air-coupled transducer. (a) Layout of the imaging system. The inset shows an internal section of the transducer. L—laser, CL—coupling lens, CS—coupling stage, MF—multimode fiber, T—trigger, DAQ—data acquisition system, PC—personal computer, ACT—air-coupled transducer, UW—ultrasound waves, LB—light beam, IS—imaging sample, ML—microlens, PS—piezocomposite disk, and ACL—air coupling layer. (b) Layout of the ray-tracing-based simulation. (c) Simulated irradiance at the focusing plane.

Close modal

The ultrasonic transducer features a multilayer configuration with a 1–3 piezocomposite (PZT27-Araldit D), a λ/4 matching layer (Araldit D), and a second λ/4 matching layer (Millipore Type VS 0.025 µm), corresponding to a center frequency of 1 MHz. It has a concave shape without any backing material. The manufacturing process consisted in the following steps. First, a 20 mm diameter and 1.8 mm thick PZT27 disk was cut and attached to a thin elastic sacrificial sheet. A square net with 0.7 mm pitch and 150 µm width was then cut, and Araldite D was poured into the grooves. The cut disk was subsequently introduced into a spherical mold with a radius of 20 mm and cured at 60 °C. After curing, the same mold was used to fix the two matching layers in two consecutive stages. Finally, a hole with a diameter of 3.7 mm was drilled in the center of the disk with a high-speed diamond drill. The transducer was ultimately mounted in a steel box with a Microdot connector. A steel hollow cylinder was placed between the rear face of the transducer box and the piezocomposite hole to form the aperture for light delivery.

Numerical simulations based on ray tracing were performed to estimate the light beam profile at the optical focus. This further served to select a commercially available microlens with a suitable focal distance as well to determine the distance of the fiber output to such lens. The simulations were performed using Zemax [Fig. 1(b)]. To this end, the entrance pupil diameter, the wavelength, and the field of view (FOV) were selected in the system settings. The optical path structure parameters were defined considering the divergence angle corresponding to the NA of the fiber. The light field emitted at the output end of the multimode fiber was approximated as a Gaussian profile. The optimal configuration and the simulated beam shape at the focus are displayed in Figs. 1(b) and 1(c), respectively. For this configuration, the full-width at half maximum (FWHM) of the light beam at the focus was 158 µm, corresponding to the theoretically achievable lateral resolution. Optical diffraction was not considered in the simulations, i.e., the size of the beam at the focus is given by the laws of geometrical optics.

The experimental performance of the OA probe was subsequently tested. First, the shape of the optical beam was characterized. The image captured with a beam profiler (SP928, Ophir Optronics, USA) at the focal position of the light beam is shown in Fig. 2(a). The beam size was estimated by fitting the measured profile to a two-dimensional Gaussian function [Fig. 2(b)] with a FWHM of 175 and 190 µm along the major and minor axes, respectively, approximately matching the numerical simulations. The sensitivity of the OA probe for the detection of OA signals depends on the acoustic depth-of-field (DOF) of the air-coupled transducer. This was estimated by pulse-echo experiments. For this, the transducer was driven with a single bipolar pulse of 200 V (peak-to-peak), and the acoustic echoes from a polystyrene surface were amplified (15 dB) and digitized with the DAQ. The resulting peak-to-peak signal amplitude as a function of the distance from the acoustic focus is shown in red in Fig. 2(c), corresponding to a DOF of ∼2.2 mm. The effective OA DOF was then estimated by collecting the ultrasound waves excited via light absorption at a layer of black paint on top of a polystyrene surface. The optical and acoustic foci were aligned by fixing the position of the fiber to achieve the highest signal intensity. The per-pulse laser energy coupled into the fiber was ∼10 µJ. The resulting peak-to-peak OA signal amplitude as a function of the distance from the focus is shown in blue in Fig. 2(c). The measured value for the OA DOF (∼3.4 mm) match that obtained with ultrasound pulse-echo measurements considering the unidirectional propagation of ultrasound waves generated optoacoustically.

FIG. 2.

Experimental characterization of the coaxially aligned transducer. (a) Measured irradiance at the light focusing plane. (b) Two-dimensional Gaussian curve fitting the measured irradiance; full width at half maxima (FWHM) along the major and minor axes is indicated. (c) Peak-to-peak amplitude of the pulse-echo (red) and the optoacoustic (OA, blue) signal as a function of the distance from the focus. (d) Photograph of the characterization phantom. (e) OA image of the characterization phantom. (f) Time-resolved OA signal for the point indicated in red in (e), band-pass filtered between 0.8 and 1.2 MHz. The FWHM of the envelope is indicated. (g) One dimensional profile of the OA image indicated in blue in (e). Distance between resolvable peaks is indicated.

FIG. 2.

Experimental characterization of the coaxially aligned transducer. (a) Measured irradiance at the light focusing plane. (b) Two-dimensional Gaussian curve fitting the measured irradiance; full width at half maxima (FWHM) along the major and minor axes is indicated. (c) Peak-to-peak amplitude of the pulse-echo (red) and the optoacoustic (OA, blue) signal as a function of the distance from the focus. (d) Photograph of the characterization phantom. (e) OA image of the characterization phantom. (f) Time-resolved OA signal for the point indicated in red in (e), band-pass filtered between 0.8 and 1.2 MHz. The FWHM of the envelope is indicated. (g) One dimensional profile of the OA image indicated in blue in (e). Distance between resolvable peaks is indicated.

Close modal

The imaging capabilities were then assessed with experiments in which the air-coupled transducer along with the attached optical fiber was used as an OA probe raster-scanned in the horizontal plane. The probe was continuously moved at a speed of 1 mm/s in the horizontal direction and in steps of 0.1 mm in the vertical direction. A phantom consisting of two 50 µm black sutures (silk fiber 7/0 USP) forming a cross was used as an imaging sample [Fig. 2(d)]. The sutures were placed at a distance of ∼20 mm from the surface of the optical table to avoid interference from acoustic reflections or OA signals generated at this surface. The output energy of the beam was ∼20 µJ, and the signals were amplified 60 dB before digitization. Image formation was performed as follows: First, the OA signals were band-pass filtered between 0.8 and 1.2 MHz, corresponding approximately to the effective detection bandwidth of the transducer. Then, a three dimensional median filter was applied to the matrix formed by superimposing all collected signals. The pixel value at each x–y position was then taken as the peak-to-peak amplitude of the corresponding signal. The resulting image is shown in Fig. 2(e). An example of the filtered OA signal corresponding to a scanning point, marked in red in Fig. 2(e), is displayed in Fig. 2(f). The width of the signal is ∼4.6 µs as indicated. This correspond to a very low axial resolution (∼7 mm) for absorbers within biological tissues (∼1540 m/s speed of sound), as expected due to the narrow detection bandwidth of the transducer. In practice, light can only be focused at depths within <1 mm inside biological tissues; hence, it is only possible to provide a two-dimensional image in this region with the air-coupled-based OA probe. Figure 2(g) shows a horizontal profile of the image, marked in orange in Fig. 2(e). It is shown that the two sutures can be distinguished for a separation of ∼160 µm, matching the expected lateral resolution considering the measured diameter of the light beam.

The feasibility to provide in vivo images of biological tissues was further tested by imaging the ear of a six-week-old female athymic nude-Fox1nu mice [Envigo RMS B.V., Netherlands, Fig. 3(a)], which was housed in a ventilated cage inside a temperature-controlled room under a 12-h dark/light cycle. For the experiment, the mouse was anesthetized with isoflurane (4% v/v for induction and 1.5% v/v for maintenance, Abbott, Cham, Switzerland) in a mixture of medical air and oxygen with flow rates of 0.8 and 0.2 l/min, respectively. The head of the mouse was fixed into a custom-made stereotactic holder coupled to a breathing mask (Narishige International, Japan). Vet ointment (Bepanthen, Bayer AG, Leverkusen, Germany) was applied on the eyes of the mice to prevent any dehydration during scanning and to protect them from laser light. The ear was gently fixed with a double-sided tape to a microscope slice placed at the focal plane of the OA probe. Blood oxygen saturation, heart rate, and body temperature were continuously monitored (PhysioSuite, Kent Scientific) and the temperature was maintained at ∼36 °C with a heating pad. The experiment was performed in full accordance with the Swiss Federal Act on Animal Protection and approved by the Cantonal Veterinary Office Zurich (license number ZH 161/18). In a first measurement, the OA probe was scanned in the same manner as that employed for imaging the phantom, i.e., it was moved at a speed of 1 mm/s in the vertical direction and in steps of 0.1 mm in the horizontal direction. The output energy of the beam was ∼20 µJ. OA signals were amplified by 80 dB before digitization. Image formation was performed in the same manner as in the phantom experiment. The resulting image is shown in Fig. 3(b) (left). Major vessels could be observed with a contrast-to-noise (CNR) ratio of 9.32 (19.36 dB) without averaging for the energy density employed (∼76 mJ/cm2). The images formed by averaging consecutive signals (5 and 25) along the fast-scanning axis are also shown in Fig. 3(b). The CNR is clearly enhanced by averaging at the expense of the spatial resolution. To avoid this, the scanning can alternatively be performed step by step. Another region of the ear was scanned with steps of 0.1 mm in the horizontal and vertical directions. A 200 µm core optical fiber with a NA of 0.22 (FG200UEA, Thorlabs, Inc., New Jersey, USA) was used for this scan, where ∼25 µJ energy per pulse was coupled. Major blood vessels could also be observed without averaging [Fig. 3(c), left] with a CNR of 7.29 (17.25 dB). The energy density was slightly lower in this case (∼61 mJ/cm2), but the energy and diameter of the beam as well as the vessel size further have an effect on the generated OA signal intensity. Averaging in this step-by-step scan enabled visualizing smaller microvascular structures [Fig. 3(c)], which, however, was achieved to the detriment of the acquisition time.

FIG. 3.

Imaging performance in biological tissues. (a) Photograph of the mouse ear scanned in the experiments. (b) OA image for the region indicated in blue in (a) acquired by continuous scan in the vertical direction. Images with 1 (no averaging), 5, and 25 averages are shown. (c) OA image for the region indicated in green in (a) acquired with a step-by-step scan. Images with 1 (no averaging), 5, and 25 averages are shown.

FIG. 3.

Imaging performance in biological tissues. (a) Photograph of the mouse ear scanned in the experiments. (b) OA image for the region indicated in blue in (a) acquired by continuous scan in the vertical direction. Images with 1 (no averaging), 5, and 25 averages are shown. (c) OA image for the region indicated in green in (a) acquired with a step-by-step scan. Images with 1 (no averaging), 5, and 25 averages are shown.

Close modal

The results presented in this work are consistent with those reported in Ref. 18, where the energy density at 532 nm was set to 12–18 mJ/cm2, signals were averaged 50 times, and larger structures in the rabbit ear were imaged. The energy density levels required for detecting OA signals from vascular structures in the mouse ear with single-shot excitation (no averaging) were higher in the present study, exceeding safety exposure levels in humans (20 mJ/cm2 for 532 nm) by a factor of 3–4. The same CNR could alternatively be achieved using energy levels below the safety limit and averaging ∼20 times. Coaxial alignment of optical and acoustic foci has been shown to provide important advantages with respect to off-axis alignment when using water coupling.19 A similar performance is then expected with air-coupled transducers. First, a higher sensitivity can be achieved, particularly when using focused ultrasound sensors. It is also possible to combine optical-resolution and acoustic-resolution imaging modes. More importantly, coaxial alignment facilitates positioning the transducer relative to the sample and covering a large DOF, which enables in vivo imaging of different types of tissues featuring arbitrary curvatures. Another limit that must be observed if this approach is eventually used in the clinical setting is the mean power optical density (200 mW/cm2 for 532 nm). This prevents repetitive measurements in the same point if the PRF of the laser is relatively high (>10 Hz) unless the energy per pulse is reduced. A higher PRF can be used if the light beam is rapidly scanned. The PRF of the laser is a key factor determining the acquisition time; hence, a large value is generally preferred. Another key factor is the number of pixels in the image, which is effectively determined by the light beam diameter. We have shown that a relatively high beam diameter (∼180 µm) is required to excite detectable signals. For a pixel size in this range, an OA image for a FOV of a few millimeters can be acquired in a reasonable time (a few minutes) if a relatively large PRF is employed. Higher resolution imaging with a smaller beam size, ultimately limited by optical diffraction, is possible. However, a larger number of averaged signals and a finer scan are required, leading to a large acquisition time that becomes impractical at some point. Alternatively, the resolution can be enhanced with super-resolution imaging approaches, e.g., based on the analysis of signal fluctuations,20 localization,21 or model-based algorithms.22 The optical spectral dimension of OA can also be exploited to retrieve valuable information. Tissue composition can be analyzed from the OA signals at multiple wavelengths. Considering that no water coupling is required with air-coupled transducers, a large spectral range can be covered even expanding toward a mid-infrared range where rich chemical information can be retrieved.23,24 OA sensing of tissue composition can be performed at a single point without forming an image and, thus, can be completed in a reasonable time even if long averaging is required.

In conclusion, we have shown that an air-coupled transducer having a central aperture for light delivery with coaxially aligned optical and acoustic foci provides sufficient sensitivity for detecting signals from microvascular structures in biological tissues. Image formation with relatively low energy densities could be achieved in a reasonable time by raster-scanning the transducer along with an attached optical fiber. This anticipates the applicability of remote OA imaging and sensing with air-coupled transducers in situations where tissue contact needs to be avoided, such as open surgeries or monitoring of burned tissues.

This work was supported by the Helmut Horten Stiftung (X.L.D.B.; Project Deep Skin), the National Natural Science Foundation of China (J.L.; Grant Nos. 81771880 and 82171989), the Tianjin Municipal Government of China (J.L.; Grant No. 19JCQNJC12800), the Spanish Universities Ministery (F.M.d.E.; Salvador de Madariaga grant), and the European Research Council (D.R.; Grant No. ERC-2015-CoG-682379).

The authors have no conflicts to disclose.

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
M.
Li
,
Y.
Tang
, and
J.
Yao
,
Photoacoustics
10
,
65
(
2018
).
2.
I.
Olefir
,
S.
Tzoumas
,
C.
Restivo
,
P.
Mohajerani
,
L.
Xing
, and
V.
Ntziachristos
,
IEEE Trans. Med. Imaging
39
(
11
),
3643
(
2020
).
3.
J.
Weber
,
P. C.
Beard
, and
S. E.
Bohndiek
,
Nat. Methods
13
(
8
),
639
(
2016
).
4.
X. L.
Deán‐Benand
and
D.
Razansky
,
Exp. Dermatol.
30
(
11
),
1598
1609
(
2021
).
5.
W.
Choi
,
E.-Y.
Park
,
S.
Jeon
, and
C.
Kim
,
Biomed. Eng. Lett.
8
(
2
),
139
(
2018
).
6.
X. L.
Deán-Ben
,
S.
Gottschalk
,
B.
Mc Larney
,
S.
Shoham
, and
D.
Razansky
,
Chem. Soc. Rev.
46
(
8
),
2158
(
2017
).
7.
R.
Manwar
,
K.
Kratkiewicz
, and
K.
Avanaki
,
Micromachines
11
(
7
),
692
(
2020
).
8.
G.
Wissmeyer
,
M. A.
Pleitez
,
A.
Rosenthal
, and
V.
Ntziachristos
,
Light: Sci. Appl.
7
(
1
),
53
(
2018
).
9.
P.
Hajireza
,
W.
Shi
,
K.
Bell
,
R. J.
Paproski
, and
R. J.
Zemp
,
Light: Sci. Appl.
6
(
6
),
e16278
(
2017
).
10.
J. L.
Johnson
,
J.
Shragge
, and
K.
van Wijk
,
J. Biomed. Opt.
22
(
4
),
041014
(
2017
).
11.
C.
Tian
,
F.
Ting
,
C.
Wang
,
S.
Liu
,
Q.
Cheng
,
D. E.
Oliver
,
X.
Wang
, and
G.
Xu
,
IEEE Sens. J.
16
(
23
),
8381
(
2016
).
12.
J. L.
Johnson
,
K.
van Wijk
,
J. N.
Caron
, and
M.
Timmerman
,
J. Opt.
18
(
2
),
024005
(
2016
).
13.
K.
Sathiyamoorthy
,
E. M.
Strohm
, and
M. C.
Kolios
,
J. Biomed. Opt.
22
(
4
),
046001
(
2017
).
14.
F. J. O.
Landa
,
X.
Luís Deán-Ben
,
F.
Montero de Espinosa
, and
D.
Razansky
,
Opt. Lett.
41
(
12
),
2704
(
2016
).
15.
R. G. M.
Kolkman
,
E.
Blomme
,
T.
Cool
,
M.
Bilcke
,
T. G.
van Leeuwen
,
W.
Steenbergen
,
K. A.
Grimbergen
, and
G. J.
den Heeten
,
J. Biomed. Opt.
15
(
5
),
055011
(
2010
).
16.
G. J.
Tserevelakis
,
P.
Siozos
,
A.
Papanikolaou
,
K.
Melessanaki
, and
G.
Zacharakis
,
Ultrasonics
98
,
94
(
2019
).
17.
X.
Luís Deán-Ben
,
G. A.
Pang
,
F.
Montero de Espinosa
, and
D.
Razansky
,
Appl. Phys. Lett.
107
(
5
),
051105
(
2015
).
18.
H.
Ma
,
K.
Xiong
,
J.
Wu
,
X.
Ji
, and
S.
Yang
,
Appl. Phys. Lett.
114
(
13
),
133701
(
2019
).
19.
S.
Jeon
,
J.
Kim
,
D.
Lee
,
J. W.
Baik
, and
C.
Kim
,
Photoacoustics
15
,
100141
(
2019
).
20.
T.
Chaigne
,
B.
Arnal
,
S.
Vilov
,
E.
Bossy
, and
O.
Katz
,
Optica
4
(
11
),
1397
(
2017
).
21.
X.
Luís Deán-Ben
,
J.
Robin
,
R.
Ni
, and
D.
Razansky
, arXiv:2007.00372 (
2020
).
22.
W.
Li
,
U. A.
Hofmann
,
J.
Rebling
,
A.
Ozbek
,
Y.
Gong
,
D.
Razansky
, and
X.
Luis Dean-Ben
,
Laser Photonics Rev.
(
in press
), arXiv:2009.10173.
23.
J.
Shi
,
T. T. W.
Wong
,
Y.
He
,
L.
Li
,
R.
Zhang
,
C. S.
Yung
,
J.
Hwang
,
K.
Maslov
, and
L. V.
Wang
,
Nat. Photonics
13
(
9
),
609
(
2019
).
24.
M. A.
Pleitez
,
A. A.
Khan
,
A.
Soldà
,
A.
Chmyrov
,
J.
Reber
,
F.
Gasparin
,
M. R.
Seeger
,
B.
Schätz
,
S.
Herzig
,
M.
Scheideler
, and
V.
Ntziachristos
,
Nat. Biotechnol.
38
(
3
),
293
(
2020
).