060602_1.pdf

Perspective on fast-evolving photoacoustic tomography

Junjie Yao

and Lihong V. Wang

Duke University, Department of Biomedical Engineering, Durham, North Carolina,

United States

California Institute of Technology, Andrew and Peggy Cherng Department of Medical

Engineering, Department of Electrical Engineering, Pasadena, California, United States

Abstract:

Significance:

Acoustically detecting the rich optical absorption contrast in biological tissues,

photoacoustic tomography (PAT) seamlessly bridges the functional and molecular sensitivity

of optical excitation with the deep penetration and high scalability of ultrasound detection.

As a result of continuous technological innovations and commercial development, PAT has been

playing an increasingly important role in life sciences and patient care, including functional brain

imaging, smart drug delivery, early cancer diagnosis, and interventional therapy guidance.

Aim:

Built on our 2016 tutorial article that focused on the principles and implementations of

PAT, this perspective aims to provide an update on the exciting technical advances in PAT.

Approach:

This perspective focuses on the recent PAT innovations in volumetric deep-tissue

imaging, high-speed wide-field microscopic imaging, high-sensitivity optical ultrasound detec-

tion, and machine-learning enhanced image reconstruction and data processing. Representative

applications are introduced to demonstrate these enabling technical breakthroughs in biomedical

research.

Conclusions:

We conclude the perspective by discussing the future development of PAT

technologies.

The Authors. Published by SPIE under a Creative Commons Attribution 4.0 Unported License.

Distribution or reproduction of this work in whole or in part requires full attribution of the original pub-

lication, including its DOI.

[DOI:

10.1117/1.JBO.26.6.060602

]

Keywords:

photoacoustic tomography; optoacoustic imaging; volumetric imaging; high-speed

imaging; optical ultrasound detection; wearable device; machine learning.

Paper 210105-PERR received Apr. 5, 2021; accepted for publication Jun. 17, 2021; published

online Jun. 30, 2021.

1 Introduction

In the recent decade, photoacoustic tomography (PAT, also referred to as optoacoustic tomog-

raphy or thermoacoustic tomography) has emerged as one of the fastest-growing imaging

technologies and has become an enabling tool in many fundamental and translational studies,

particularly for early cancer diagnosis, functional brain imaging, drug delivery monitoring, and

interventional procedure guidance.

The imaging process in PAT typically starts with a short

laser pulse that illuminates biological tissue. As the excitation photons propagate through

the tissue, some are absorbed by endogenous or exogenous biomolecules, and their energy

is partially or completely converted into heat and thus a transient temperature rise, through non-

radiative relaxation of excited molecules [Fig.

1(a)

]. Generally, biomolecules with a lower or

zero fluorescent quantum yield and a larger Grüneisen parameter have more efficient thermal

conversion. When the excitation laser pulse width satisfies both thermal and stress confinement,

the resultant initial pressure rise is proportional to the transient temperature rise via the thermo-

elastic effect.

The pressure wave is then detected outside the tissue by an ultrasonic transducer

or transducer array to form a tomographic image that maps the optical energy deposition inside

the tissue. PAT has a 100% relative sensitivity to small optical absorption variations, which

*Address all correspondence to Junjie Yao,

junjie.yao@duke.edu

; Lihong V. Wang,

lvw@caltech.edu

PERSPECTIVE

Journal of Biomedical Optics

060602-1

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

means a given percentage change in the optical absorption coefficient yields the same percentage

change in the PA signal amplitude.

Because PAT does not rely on fluorescence emission, which

usually has a quantum yield much <

100%

, it can image nearly all molecules, fluorescent or not.

–

Although PAT has been implemented in numerous configurations and tailored for diverse

applications, its basic principles and major components remain similar. A typical PAT system

includes (i) a short-pulsed laser or multiple lasers at one or more optical wavelengths for efficient

PA wave generation, (ii) a wideband ultrasonic transducer or transducer array for PA signal

detection, (iii) a data acquisition system for signal amplification, filtering, and digitization, (iv) an

electronic system for trigger synchronization and data collection/streaming, and (v) a computa-

tional system for data processing, image reconstruction, and functional information quantifica-

tion. So far, PAT has been implemented with two major image formation methods [Fig.

1(b)

The first method, direct image formation (commonly referred to as photoacoustic microscopy, or

PAM), is based on mechanical scanning of a focused excitation light beam and a focused single-

element ultrasound transducer. A focused ultrasound transducer usually provides better detection

sensitivity than a flat transducer. PAM can be further classified into optical-resolution PAM (OR-

PAM) and acoustic-resolution PAM (AR-PAM), depending on the focal spot size of the optical

excitation and acoustic detection.

The second method, inverse reconstruction image formation

(commonly referred to as photoacoustic computed tomography, or PACT), is based on wide-field

light illumination and parallel acoustic detection by a multi-element ultrasound transducer array.

Each transducer element can be approximated as a point detector with a large acceptance angle.

Compared with PAM, PACT typically has a higher imaging speed and greater penetration,

but lower spatial resolutions. Other PAT implementations, such as photoacoustic endoscopy,

a miniaturized implementation of PAT for internal organ or intravascular imaging, can be imple-

mented in either a PAM or PACT configuration.

–

The imaging performance of major PAT

implementations is summarized in our previous tutorial article.

Readers are also referred to

a practical guide for implementing PAT systems.

Fig. 1

Principles, implementations, and representative applications of PAT. (a) Working principle

of PAT, from the laser excitation to the image reconstruction. (b) Three representative implemen-

tations of PAT: optical-resolution photoacoustic microscopy (OR-PAM), acoustic-resolution photo-

acoustic microscopy (AR-PAM), and photoacoustic computed tomography (PACT) with a linear

ultrasound transducer array. SOL, silicone oil layer; UT, ultrasound transducer; UTA, ultrasound

transducer array. (c) OR-PAM image of the microvasculature of a mouse ear bearing a xenotrans-

planted B16 melanoma tumor (white dashed box) at 584 nm. Depth is coded by colors: blue

(superficial) to red (deep). (d) White light photograph of the mouse ear. (e) OR-PAM image of

the melanoma at 600 nm. Blood vessels are invisible due to the relatively weak absorption of

hemoglobin at this wavelength. (e) OR-PAM image of oxygen saturation (sO

2

) of the principal

arterial-vein pair. (f) OR-PAM image of the blood flow velocity of the principal arterial-vein pair.

The directions of positive and negative flow are defined by the arrows. Reproduced with permis-

sion from Ref.

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-2

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

Seamlessly integrating the optical excitation with acoustic detection, PAT has several

advantages over other high-resolution optical imaging technologies: (i) PAT is maximally

sensitive to the rich optical absorption contra

st of biological tissue, and it is inherently well

suited for anatomic, functional, and molecular imaging [Fig.

1(c)

]; (ii) because biological

tissue is more transparent to sound than to li

ght in terms of the scattering mean free path,

PAT provides far greater penetration depth than optical microscopy; (iii) because of the high

scalability of optical excitation and ultras

ound detection, PAT can be implemented in many

different configurations, providing multi-s

cale observation of the sa

me biological process

with a consistent contrast mechanism; and (iv) PAT is functionally complementary to and

engineeringly compatible with other imagin

g modalities, especially ultrasound imaging.

PAT-capable multi-modal imaging can provide

a more comprehensive understanding of bio-

logical phenomena.

PAT has gained tremendous momentum in the last decade, driven by innovations in high-

power lasers, high-sensitivity ultrasound detection, high-speed scanning, large-scale computa-

tion, nanotechnology, protein engineering, and machine learning. In our tutorial published in

2016,

we systematically introduced the foundation of PAT technologies, including the imaging

principles from light to sound, the implementations at different length scales, and representative

applications in life sciences. For readers interested in developing and/or applying PAT for bio-

medical research, our tutorial and other comprehensive review articles can provide a practical

guide.

–

Built upon our tutorial, this perspective aims to provide an update on the develop-

ments in PAT technologies in the last several years. Limited by the paper length, we are not able

to cover all the exciting advances in PAT but will focus on several breakthroughs that have

allowed new imaging capabilities not available to traditional PAT systems, including (i) volu-

metric PATof deep tissues with nearly isotropic resolution, using a 2D ultrasound array; (ii) high-

speed PAT with microscopic resolution, wide field-of-view (FOV), and functional imaging

capability; (iii) high-sensitivity PAT with optical ultrasound detectors that have small sizes, wide

bandwidth, and high transparency; and (iv) novel image reconstruction and data processing

methods enabled by large-scale computation or machine learning, with improved image quality

and quantitative accuracy. We introduce these PAT innovations in the context of the longstanding

engineering challenges, summarize their much-improved imaging performance (usually by

orders of magnitude over traditional PAT), and present the representative applications in fun-

damental research and translational studies. We conclude with a brief discussion of remaining

challenges and future developments in PAT.

2 Technical Advances in PAT

The technological development in PAT has been fueled by advances in almost every key system

component, from hardware to software, such as light sources with higher power, higher repeti-

tion rate, wider wavelength tuning range, and lower cost; novel ultrasound detectors with higher

sensitivity, larger frequency bandwidth, and lower cost; and advanced image reconstruction algo-

rithms with reduced artifacts, higher computation speed, and better quantification accuracy. For

example, there has been a strong interest in using low-cost laser diodes and light-emitting diodes

for PAM and PACT.

The low-cost light sources typically have much lower pulse energy (less

than a few mJ), longer pulse width (tens to hundreds of ns), and lower spatial/temporal coher-

ence, compared with the Class IV pulsed lasers typically used in PAT, but they can substantially

reduce the system cost, improve the portability, and thus facilitate the technical translation. The

diode-enabled PAT systems have been used for a wide range of applications where the imaging

depth or temporal resolution, spatial resolution, and/or the spectroscopic measurement accuracy

can be relaxed, including needle biopsy guidance,

melanoma imaging,

skin implant mon-

itoring,

and human finger imaging.

Limited by space, we will focus on several important developments that have overcome the

longstanding limits in traditional PAT. Interested readers are referred to comprehensive review

articles that provide in-depth analyses and discussions on low-cost light sources,

–

novel

ultrasound sensors,

PA contrast agents,

PA endoscopy,

deep learning enhanced

PAT,

–

as well as clinical translation of PAT.

–

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-3

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

2.1

Volumetric PACT with High-Speed and Isotropic Resolutions

PAT is inherently capable of volumetric or three-dimensional (3D) imaging, benefiting from the

time-resolved detection of the acoustic waves that provide the depth information of the targets.

For PAM, in which a single laser pulse generates a 1D depth-resolved image, 2D raster scanning

is employed to obtain a 3D image; for PACT with a 1D transducer array, in which a single laser

pulse generates a 2D cross-sectional image, orthogonal scanning along the elevational direction

is needed to obtain a 3D image. We will discuss the new developments in PAM in a later section,

but here we focus on volumetric PACT. While traditional volumetric PACT has been widely used

for functional brain imaging, small-animal whole-body imaging, and breast cancer diagnosis in

humans, the major drawbacks include the long imaging time needed for mechanical scanning

and the anisotropic spatial resolution (much worse elevational resolution) determined by the

cylindrical focusing of the transducer elements.

To accelerate the speed and improve resolution symmetry of volumetric PACT, recent efforts

have concentrated on applying 2D ultrasound transducer arrays coupled with high-power laser

sources and 3D image reconstruction. For PACT with a 2D transducer array, a single laser shot

can theoretically generate a 3D image,

and the resolutions can be nearly isotropic at the

center of the FOVor the well-resolved FOV. In practice, however, 2D transducer arrays typically

lack enough active elements to satisfy the spatial Nyquist sampling over a large volume, limited

by the transducer fabrication complexity and the number of data acquisition channels. Thus,

multiplexed data acquisition (electronic scanning)

and rotational scanning

are typically

needed to improve the spatial sampling density. Moreover, because repeated wide-field illumi-

nation may cause tissue damage due to accumulated heating,

the optical fluence (

∕

) per

pulse and the average fluence rate (

∕

) on the tissue surface need to be carefully controlled.

Different types of 2D transducer arrays have been explored for volumetric PACT, mostly

based on piezoelectric materials, as summarized in Table

. To maximize the detection aperture,

several groups have explored the feasibility of spherical ultrasound transducer arrays, usually

with the transducer elements sparsely distributed over the array surface.

Compared with

the planar 2D array,

the spherical array can provide higher spatial sampling density around its

center volume and better visualization of 3D structures with different orientations. Matsumoto

et al.

developed a volumetric PACT system using a sparse hemispherical detector array that is

scanned in a spiral pattern [Fig.

2(a)

], which can provide detailed blood oxygenation images on

the breast skin surface but suffers from slow imaging speed and low penetration depth [Figs.

2(c)

and

2(d)

]. Similarly, Schoustra et al.

upgraded the Twente Photoacoustic Mammoscope using

Table 1

Comparison of representative 2D ultrasound arrays in volumetric PACT.

2D ultrasound array

Hemi-spherical

Cup

Quad-arc

12-arc

Fabry

–

Perot

Array radius (mm)

130

120

25 (max)

Number of elements

500

256

1024

384

100

∕

2

Scanning scheme

Rotational None for small FOV Rotational Rotational

Raster

Spiral for large FOV

Element shape

Circular

Square

Rectangular Square

Circular

Element size (mm)

1.5

0.6

0.7

3.5

0.068

Element pitch (mm)

3.13

0.74

4.9

0.068

Central frequency (MHz)

2.25

Receiving bandwidth (%)

100

Noise equivalent

pressure (Pa)

0.5

Not available

200

References

and

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-4

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

12 arc-shaped transducer arrays arranged over a hemi-spherical surface, which can provide 3D

vascular images of healthy breasts within four minutes. To accelerate the imaging speed, the

Razansky group has developed a volumetric PACT system using a 2D ultrasonic transducer array

with 256 elements densely arranged on a partial cup, which has achieved a 3D imaging rate of

∼

50 Hz

This system has been implemented in both desktop and handheld configurations

and has been applied to capture in real time the heart beating of a mouse and the neuronal activ-

ities of a swimming zebrafish and a GCaMP-expressing mouse brain.

However, without

performing additional scanning, such a transducer arrangement results in a small well-resolved

FOV (

∼

4-mm

diameter) and can only be used to study small animal organs, such as the heart and

brain. Moreover, the limited view aperture (i.e., <

steradian solid angle) can further compro-

mise the image quality when a higher imaging speed is required.

To simultaneously improve the spatial sampling and imaging speed over a large FOV, the

Wang group has reported a novel design with a quad-arc-shaped 2D transducer array, which has

1024 elements and one-to-one mapped signal amplification and data acquisition [Fig.

2(b)

rotating the quad-arc-shaped array by 90 deg, the volumetric PACT system can provide a large

well-resolved FOV (

diameter

>

100 mm

) and

∼

steradian solid angle, with nearly isotropic

resolution of 370 to

390

. It takes only 2 to 10 s to generate a volumetric image, depending on

the targeted FOV, which is much faster than the previously reported systems. The newly devel-

oped volumetric PACT system has been applied for imaging a human breast within a single

breath hold [Figs.

2(e)

and

2(f)

]. So far, this is the volumetric PACT system with the largest

well-resolved FOV and the highest speed. Nevertheless, the imaging speed can be further

improved by adopting pulsed lasers with a higher repetition rate (

>

10 Hz

) as well as faster rota-

tional scanning stages. Meanwhile, functional and molecular imaging capability remains to be

demonstrated with high-speed, wavelength-tunable light sources. However, to comply with the

laser safety standard,

a higher laser repetition rate would lead to a lower maximum permissible

exposure (

∕

) on the tissue surface, and thus a lower signal-to-noise ratio. In other words,

a higher imaging speed in volumetric PACT may come at the cost of the final image quality and

penetration depth.

The imaging characteristics of several representative volumetric PACT systems are summa-

rized in Table

Fig. 2

Volumetric PACT with 2D ultrasound array. (a) Schematic of the hemispherical array with

512 elements sparsely arranged over the sensor surface.

(b) Schematic of the quad-arc array

with 1024 elements densely arranged along four arcs (with a separation of 90 deg) that are

mounted on a hemispherical surface.

(c)-(d) Projection images of the human breast vascular

oxygenation obtained by the volumetric PACT in (a), with a total scanning time of 120 s.

(e)

–

(f) Projection images of the human breast vasculature obtained by the volumetric PACT in

(b), with a total scanning time of 10 s. Adapted with permissions from Refs.

and

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-5

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

In addition to 2D ultrasound transducer arrays based on piezoelectric materials, optical ultra-

sound detectors, such as the Fabry

–

Perot interferometer,

micro-ring resonator,

and Bragg

grating fiber,

–

have been actively explored for volumetric PACT. Compared with piezoelec-

tric transducers, optical ultrasound detectors often have smaller size (<

100

), larger detection

bandwidth and receiving angle, higher detection sensitivity per unit area, better transmittance of

the PA excitation light, and simpler PA signal readout, all of which can help improve the res-

olution and penetration depth of volumetric PACT. We will discuss the developments in optical

ultrasound detectors in a later section.

2.2

High-Speed Photoacoustic Microscopy over a Large FOV

Biological functions occur on a wide range of temporal and spatial scales, which requires

imaging technologies to provide best-matched imaging speeds and FOVs. For example, a single

neuron action potential lasts for 1 to 2 ms along a

10-

-diamter axon, neurovascular coupling

happens within hundreds of milliseconds over a functional circuit with a

100-

radius, and

the resting-state functional connectivity between the brain

’

s sub-regions occurs within tens of

seconds over a millimeter-level radius. Configured to work in 1D, 2D, or 3D imaging modes,

different implementations of PAT offer a wide range of imaging speeds with associated

tradeoffs.

In this section, we will focus on new developments in PAM that can offer high

imaging speed, large FOV, and functional imaging capability.

For PAM, different scanning mechanisms can be employed according to the desired imaging

speeds.

Unlike confocal or two-photon microscopy, PAM does not require depth scanning for

3D imaging due to its time-resolved acoustic detection. When high-speed imaging is needed in

OR-PAM, the focused excitation laser beam can be raster-scanned within the acoustic focal spot

(

∼

in diameter), which largely confines the FOV to single vessels.

Alternatively, cylin-

drically focused or unfocused acoustic detection can enlarge the FOV

—

up to

∼

40 mm

in diam-

eter as demonstrated thus far

—

at the expense of detection sensitivity.

In order to achieve a

high detection sensitivity over a large FOV, it is critical to maintain the confocal alignment of the

optical excitation and acoustic detection. Recently, 1D or 2D water-immersible resonant MEMS

(microelectromechanical systems) scanning mirrors that confocally steer both the excitation laser

beam and the emitted acoustic beam

have achieved a 2D imaging rate of 500 Hz and a 3D

imaging rate of

∼

1Hz

, with a moderate FOV of

∼

3mm

and uncompromised detection

sensitivity.

–

By using a pulse-width-based, single-wavelength method or a Raman-shifter-

based, two-wavelength method, MEMS-scanning OR-PAM can monitor the change in blood

oxygenation of mouse brain

in vivo

However, it is challenging for the resonant MEMS

Table 2

Comparison of representative volumetric PACT systems.

Volumetric PACT systems

by institute

Kyoto

University

of Zurich

Caltech

University

of Twente

UCL

Array shape

Hemi-spherical

Cup

Quad-arc

12-arc

Fabry

–

Perot

Lateral resolution (

270

200

390

1060

100

Axial resolution (

270

200

370

960

100

Diameter of well-resolved

FOV without scanning (mm)

N/A

Diameter of well-resolved

FOV with scanning (mm)

140

100

Laser repetition rate (Hz)

100

200

Imaging time (second)

120

2-10

240

Imaging depth (cm)

2.2

References

and

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-6

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

scanning mirrors to provide a larger FOV without sacrificing the system

’

s detection sensitivity,

and the scanning range drops sharply when the scanning frequency deviates from the resonant

frequency.

To address the tradeoff between the scanning speed and scanning range of MEMS scanners, a

recent work by Lan et al.

has reported the use of a water-immersible polygon mirror scanner in

OR-PAM that has achieved a 2D imaging rate of 1.2 kHz over a 12-mm scanning range and a 3D

imaging rate of 1 Hz over a

12 mm

FOV [Fig.

3(a)

]. The polygon scanner with six facets is

driven by a rotational DC motor, with each rotation providing six repeated 2D scans. Unlike the

resonant MEMS mirror, the polygon scanner can maintain its large scanning range at different

scanning frequencies, which is critical for imaging large organs, such as the blood oxygenation

change of the whole mouse cortex [Fig.

3(b)

]. By combining the polygon scanner with a Raman-

shifter-based, two-wavelength laser, Chen et al.

have demonstrated high-speed functional im-

aging of the hemodynamic response of the entire mouse ear to epinephrine, a commonly used

vasoconstrictor. Nevertheless, one drawback of the polygon scanner is the lack of adjustment of

its scanning range. Once the optical path is constructed, the scanning range is determined and

difficult to change, which poses a waste of scanning time on small targets. Different scanning

mechanisms used in high-speed OR-PAM are compared in Table

For AR-PAM, the imaging speed is mainly limited by the mechanical scanning speed and the

pulse repetition rate of the high-pulse-energy laser, the latter of which is limited by laser safety

on the tissue. In AR-PAM, mechanical scanning by a step motor or a voice-coil scanner can be

used, with a scanning step size

∼

times that used in OR-PAM.

A 2D imaging rate of 40 Hz

has been achieved by AR-PAM over a scanning range of

∼

9mm

, sufficient to capture the

oxygenation dynamics in a mouse heart within a heartbeat.

Recently, the 2D water-immersible

MEMS scanning mirrors have also been adapted to improve the imaging speed of AR-PAM

∼

-fold, with an FOV of

2.5 mm

–

When integrated with additional mechanical

scanning to

“

stitch

”

the MEMS scanning area, AR-PAM can image a

30 mm

area

within 70 s.

Fig. 3

High-speed PAM with novel scanning and non-scanning approaches. (a) Schematic of the

high-speed PAM using a water-immersible polygon scanner, in which a single rotation of the poly-

gon scanner provides six repeated 2D scans.

UT, ultrasound transducer. (b) High-speed imag-

ing of the mouse brain under hypoxia challenge obtained by the system in (a), showing reduced

blood oxygenation. (c) Schematic of the high-speed PAM using wide-field light illumination and

a single-element ultrasound detector through an ergodic relay.

(d) High-speed tracking of arterial

pulse wave obtained by the system in (c), showing the heated blood (coded in color) flowing in

the vessels. Adapted with permissions from Refs.

and

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-7

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

While the above fast-scanning

–

based approaches have significantly improved the imaging

speed of PAM systems, they are fundamentally limited by the laser

’

s pulse repetition rate when

the spatial Nyquist sampling needs to be satisfied. This limitation is particularly true for high-

speed OR-PAM, which often requires a small scanning step size of <

. For example, for the

recently published polygon-scanner

–

based PAM system,

the laser

’

s maximum pulse repetition

rate is 800 kHz and the B-scan (i.e., the fast-scanning axis) rate can reach as high as 2000 Hz

over a 10-mm scanning range. However, to satisfy the Nyquist sampling theorem, the B-scan rate

is limited to only 200 Hz if the FOV is kept the same, much lower than the maximal achievable

speed. One way to increase the scanning speed over a large FOV is to increase the scanning step

size at the cost of effective spatial resolution. Sparse sampling has thus become a necessary

compromise when imaging speed needs to be increased.

To relax the requirement on the laser

’

s pulse repetition rate, one solution is non-scanning PA

imaging based on an ergodic relay, which can simultaneously encode all of the PA signals from a

large FOVaccording to their unique time-delay characteristics.

–

In a recent work, Li et al.

demonstrated a high-speed implementation referred to as photoacoustic topography through an

ergodic relay (PATER). In PATER, for each single excitation laser pulse, the encoded PA signals

can be detected in parallel via a single-element ultrasound transducer and then decoded math-

ematically to reconstruct a 2D projection image [Fig.

3(c)

With a point-by-point scanning

calibration step, PATER has demonstrated a topographic frame rate of 2 kHz over a field of

view of

7.5 mm

, and has been applied to image the blood pulse wave velocity and track

the circulation of melanoma cells in the mouse brain [Fig.

3(d)

]. Because no optical or acoustic

beam scanning is needed in PATER, the imaging speed is essentially limited by the acoustic

transit time within the ergodic relay. Nevertheless, the current calibration method lacks the depth

information and thus only topographic images can be provided.

Limited by slow imaging speed and bulky system size, desktop PAM is mostly applied on

small animals under anesthesia or human subjects with the targeted region fixed (e.g., arm, hand,

or finger) to minimize the motion artifacts. Enabled by the elevated imaging speed and the result-

ant high imaging throughput, it has become possible to implement miniaturized PAM to image

otherwise challenging targets prone to motion artifacts, such as brain functions of freely moving

animals, longitudinal monitoring of rare circulating tumor cells of melanoma patients, and skin

cancer screening of difficult regions such as the neck and back. In recent years, various PAM

systems have been developed for handheld,

–

wearable,

–

and even head-mounted appli-

cations,

thanks to the advances in high-speed scanning methods. All these technical inno-

vations have allowed the miniaturization of PAM systems without sacrificing the imaging

performance. For example, to capture normal brain functions, it is critically important to record

the neural activities in freely behaving animals with high resolution and high throughput. Chen

et al.

have reported a wearable PAM system that is small enough to be mounted on the head of a

freely moving rat. A miniaturized MEMS scanning mirror provided high-speed, high-resolution

Table 3

Comparison of scanning mechanisms in OR-PAM.

Scanning methods

B-scan

rate (Hz)

Scanning

range (mm)

Detection

Sensitivity

Transducer

focusing

Ref

Mechanical motor

+++

Spherical

Voice-coil scanner

+++

Spherical

Galvo scanner (unfocused transducer)

100

∼

Unfocused

Galvo scanner (2D focused transducer)

180

0.1

+++

Spherical

Galvo scanner (1D focused transducer)

Cylindrical

Water-immersible MEMS scanner

400

Spherical

Water-immersible polygon scanner

1200

Spherical

More plus signs indicate better detection sensitivity.

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-8

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

imaging of the brain

’

s hemodynamic activities during and post ischemia challenge. Remarkably,

the motion artifacts were negligible during the 90-min imaging time.

2.3

Optical Detection of the Ultrasound Pressure

Piezoelectric ultrasound transducers still largely dominate the PAT technologies due to their wide

availability, high detection sensitivity, low fabrication cost, and ease of use. However, optical

ultrasound sensors have their unique advantages for PAT and have gained more momentum in

recent years.

Unlike ultrasonography, PAT does not need ultrasound transmission, and the PA

signals are usually broadband, so optical sensors can be used in receiving-only mode, taking full

advantage of their small size, large receiving angle, wide detection bandwidth, strong respon-

sivity in the low frequency band, and good compatibility with PA light path. More importantly,

the detection sensitivity of optical sensors usually has less dependence on the sensor size, which

leads to better sensitivity than piezoelectric transducers of the same size, especially at higher

frequencies (

>

2.5 MHz

Practically speaking, the optimal size of piezoelectric transducers

used in PACT is equivalent to a half-wavelength on the FOV boundary. Further size reduction

provides no clear benefit in spatial sampling density or receiving angular directivity.

Therefore,

when comparing the performance of optical ultrasound sensors with piezoelectric transducers,

we suggest half-wavelength sized piezoelectric transducers shall provide a fair comparison

unless the application is inherently space-constrained. For example, the optical sensors

’

high

detection sensitivity with a small form factor is particularly attractive for endoscopic and wear-

able PAT implementations, in which the working space is extremely limited. A thorough com-

parison of optical sensors and piezoelectric transducers can be found in the review article by

Wissmeyer et al.

So far, there have been two types of optical sensors demonstrated in PAT technologies: inter-

ferometric sensors and refractometric sensors. Taking advantage of the optical and acoustic inter-

actions in the PA effect, these optical sensors often probe a single step in the PA signal generation

and propagation process. The interferometric sensors typically have better detection sensitivity

than the refractometric sensors.

The above-mentioned Fabry

–

Perot interferometer, micro-ring

resonator, and Bragg grating fiber are all interferometric sensors that target the last step in the PA

signal propagation and have been applied in volumetric PACT and/or PAM as point-like detec-

tors. Refractometric sensors often exploit the earlier steps in PA signal generation, such as the

photothermal or thermoelastic effect in the tissues or coupling medium, and detect the change in

probing light beam

’

s transmission, reflection, or deflection.

–

100

Such changes, however, are

usually small. While interested readers are referred to the comprehensive review article on the

optical sensors in PAT technologies,

we would like to discuss three important limitations of

optical sensors

—

speed, scalability, and stability

—

as well as highlight some new studies aiming

to address these limitations.

For the PACT systems based on a planar Fabry

–

Perot interferometer, the nearly isotropic

spatial resolutions, approximately defined by the optical probing beam size, can be well main-

tained with 2D dense spatial sampling over the entire FOV.

However, the imaging speed is

traditionally limited by the point-by-point raster scanning of the probing beam and the low pulse

repetition rate of the PA excitation laser at 50 Hz.

101

A more recent work by the UCL group has

demonstrated a 32-fold higher imaging speed by employing a total of eight parallel probing

beams scanning simultaneously over the sensor [Figs.

4(a)

and

4(b)

102

A customized, high-

speed laser (200 Hz) also helps to improve the imaging speed. Such a speed-up strategy, how-

ever, has drastically increased the system

’

s complexity and cost, and the high-speed laser has

relatively low pulse energy. Wide-field detection of the interference pattern on the sensor surface,

using time-gated light illumination and a high-speed CCD camera, can potentially speed up the

imaging as well.

103

104

Other optical ultrasound sensors, including polymer micro-ring

and Bragg-grating fiber,

have recently been demonstrated as point-like detectors in PAT, often with sensor sizes that are

orders of magnitude smaller than their piezoelectric counterparts. However, one major obstacle

met by these optical sensors is the extreme difficulty in scaling up the production while main-

taining consistent optical properties, such as the optical resonant wavelength,

factor, and trans-

mission efficiency. Unlike piezoelectric materials that allow the manufacturing of high-density

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-9

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

arrays, it is difficult for optical sensors to be multiplexed. Slight fabrication inaccuracy, such as

of the Fabry-Perot polymer

’

s thickness or the micro-ring

’

s diameter, would drastically change its

operating parameters. This is particularly problematic for volumetric PACT, which requires

parallel signal detection to improve the imaging speed. To address this issue, Westerveld et al.

have developed a new micro-ring-resonator using silicon photonic technology [Fig.

4(c)

]. As a

proof of concept, a total of ten resonators can be fabricated onto a single optical bus waveguide

[Fig.

4(d)

]. This CMOS-compatible fabrication process may provide a viable path for scaling the

optical sensor to a 2D array for high-speed volumetric PACT [Fig.

4(c)

Another significant drawback of optical ultrasound sensors, particularly the interferometric

sensors, is low stability in the biological environment. For example, the micro-ring resonator is

sensitive to contamination on the sensor surface (e.g., dust, body fluid, or blood stain), which

induces scattering and absorption loss, and the Fabry

–

Perot interferometer is sensitive to the

environmental temperature drift, which changes the thickness and refractive index of the poly-

mer spacer. Such instability in the biological environment often leads to fast degradation of the

sensor sensitivity and prevents the use in longitudinal

in vivo

studies. To address this issue,

Li et al.

105

have developed a micro-ring resonator by soft nanoimprinting lithography, which

has significantly improved stability for

in vivo

applications. The micro-ring resonator is encap-

sulated by a protection layer made of both optically and acoustically transparent polydimethyl-

siloxane (thickness

). By isolating the micro-ring and waveguide from the potential

contaminants (e.g., blood), the micro-ring resonator has demonstrated impressively stable per-

formance when implanted on a mouse cortical surface for 28 days. Similarly, Westerveld et al.

recently demonstrated a micro-ring resonator using a thin layer of acoustic membrane to isolate

the ring structure from the environment, which can potentially improve the sensor

’

s stability in

water. To overcome the thermal stability of the Fabry

–

Perot interferometer, Chen et al.

106

have

incorporated an additional heating light source at 650 nm into the interferometer, which can

Fig. 4

PAT with optical ultrasound sensors. (a) Schematic of a PACT system using a Fabry

–

Perot

interferometer with eight parallel probing beams.

102

(b) A 3D human palm image obtained the sys-

tem in (a) with a total imaging time of 10 s. (c) Schematic and photograph of a micro-ring-resonator

using silicon photonic technology.

(d) The optical transmission spectrum of a multiplexed micro-

ring-resonator array with 10 sensors. (e) Representative 3D image of three stacked polyamide

sutures obtained by the micro-ring resonator in (a). Adapted with permissions from Refs.

and

102

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-10

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

modulate the polymer spacer

’

s thickness and thus compensate for the temperature-induced res-

onant spectral shift. Such thermal compensation can be performed in real time by a closed-loop

feedback.

2.4

Deep Learning Enhanced Image Reconstruction and Processing

Like many other technologies, PAT

’

s developments have been incorporating the fast-evolving

deep learning enabled by the prevalence of graphical processing unit (GPU) capabilities.

107

–

116

Deep learning is well-suited for addressing some long-standing challenges of PAT, such as

improving ill-posed reconstruction, removing limited-view artifacts, denoising channel data,

improving diffraction-limited spatial resolution, and upsampling sparse input data. Many of

these efforts have proven to be promising when traditional solutions either fail or make only

incremental progress. There have been several excellent reviews on the history and status of

deep learning technologies in PAT,

117

to which we refer interested readers. A detailed

comparison of different deep learning approaches in PAT can be found in the review article

by Gröhl et al.

Here we will highlight several of the most exciting advances.

There is a clear difference between the deep learning formulation in PACT and PAM. In

PACT, many challenges arise from solving the inverse problem, mostly with partial and/or sparse

detection geometries.

118

Deep learning in PACT can be used as (i) a pre-processing or post-

processing step in the image reconstruction, (ii) replacement of the traditional image reconstruc-

tion altogether, or (iii) one integrated step in the iterative reconstruction. For example, Gutta et al.

used a fully connected deep neural network (FC-DNN) as a pre-processing step to correct the

sonograms acquired by each transducer channel and broaden the bandwidth of the received chan-

nel data.

119

Davoudi et al.

120

used a fully convolutional neural network (U-Net) to reconstruct the

PACT data obtained by a ring array with limited view or sparse sampling, which resulted in

improvements in both spatial frequency coverage and the final image quality [Figs.

5(a)

and

5(b)

]. For PACT with a linear transducer array, a stabilized generative adversarial network

(GAN) model with gradient clipping has been employed as a post-processing step, which

Fig. 5

Limited-view PACT improved by deep learning methods. (a) Whole-body PACT images of

a mouse and the corresponding spatial frequency spectra obtained by a ring-array system with a

360 deg or 60 deg detection angle range. The images were reconstructed using the traditional

back projection method.

120

(b) A U-Net

–

based deep learning method was used to reconstruct the

PACT image with a 60 deg detection angle. (c) A comparison of PACT images obtained by a linear

array before and after a GAN method was used to reduce the limited-view artifacts.

121

Adapted

with permissions from Refs.

120

and

121

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-11

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

can reduce the limited-view and limited-bandwidth reconstruction artifacts of

in vivo

data

[Fig.

5(c)

121

Another key area of research is integrating the deep learning into the PA forward

operator for iterative

–

based image reconstruction, as demonstrated by Hauptmann et al.

122

123

and Bioink et al.

124

However, these iterative methods can be time

–

consuming.

Unlike PACT, PAM does not require inverse reconstruction, so deep learning models can

directly map time-resolved input signals to output images, and improve imaging speed, sig-

nal-to-noise ratio, and spatial resolution. One of the major utilizations of deep learning in

PAM is to improve sparsely sampled images, thereby shortening image acquisition time without

substantially degrading image quality. For example, DiSpirito et al.

125

have developed a modi-

fied fully dense U-Net architecture (FD U-Net), and demonstrated the feasibility to recover

microvessels in the mouse brain by acquiring only 2% of the pixels required by the Nyquist

sampling. For situations that lack ground truth data for model training, Vu et al.

126

have proposed

an innovative method that iteratively refines undersampled PAM images using a deep learning

prior. This work is of particular interest because it does not require training on a large

PAM dataset with ground truth. Deep learning has also recently been used by Song et al.

127

improve PAM images with extremely low excitation laser energy. Most recently, Sharma and

Pramanikhave

128

developed an FD U-Net to enhance the lateral resolution of AR-PAM, espe-

cially in the out-of-focus regions.

Notably, deep learning methods have also been investigated for improving quantitative PAT,

which has been difficult for deep-seated targets due to spectral coloring. Deep learning

approaches have been developed to either better estimate the optical fluence at different wave-

lengths or completely replace the traditional spectral unmixing algorithms. For example,

a sequential-learning recurrent neural network has been used to predict eigen-fluence maps

in deep tissue,

129

which were subsequently used for linear unmixing of the oxy- and deoxy-

hemoglobin concentrations.

129

In another work, Gröhl et al.

130

applied a fully connected neural

network on multi-spectral PA images, which improved the quantification accuracy of blood

oxygenation estimations on phantoms and

in vivo

porcine brain. Further, Bench et al.

131

applied

a 3D encoder-decoder style neural network to predict volumetric blood oxygenation; however,

this methodology has not yet been adapted to

in vivo

data due to the complexity of tissue

’

optical properties.

Nevertheless, one obstacle to the broad adoption of deep learning in PAT is the heavy reliance

on simulation data and the lack of large, open-source repositories of

in vivo

data. The gap

between simulation data and

in vivo

data makes model extrapolation to

in vivo

applications

difficult. Potential solutions to address this obstacle are for the community to (i) create a large,

open-source repository of various

in vivo

training examples or (ii) improve the quality of sim-

ulation data to better mimic

in vivo

cases. Ultimately, the incorporation of deep learning into PAT

requires the training of robust models that can readily adapt to a variety of

in vivo

conditions

—

many of which, such as sparsely-sampled, limited-view, and limited-bandwidth detection, are in

non-ideal environments.

3 Conclusion and Outlook

Harnessing the relevant advances in physics, chemistry, mathematics, and computer science,

PAT has experienced its fastest development in the last decade and become the enabling tech-

nology in many biomedical studies. Previously, the technical innovations in PAT were often

limited by the performance of key system components, such as the laser

’

s pulse repetition rate

and the ultrasound transducer

’

s sensitivity. Although many engineering solutions were explored

to address these long-standing technical challenges, they often require trade-offs between im-

aging parameters, such as the imaging speed versus the field of view, the spatial resolution versus

the penetration depth, and the detection sensitivity versus the detector size. Thanks to advances

in key technologies, such as high-power laser sources, fast scanning mechanisms, and minia-

turized optical ultrasound sensors, the traditional tradeoffs in PAT technical development

have become less constraining. The current perspective builds upon our previous tutorial on

the fundamentals of PAT and highlights several key technical developments that have generated

the most impact.

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-12

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

Innovations in volumetric PACT, with high speed and isotropic resolution, have addressed

one of its most prominent technical hurdles precluding clinical potential. Full spatial sampling

over a large FOV has enabled image quality similar to that of MRI or x-ray CT, particularly for

breast cancer imaging. The functional and molecular imaging provided by PACT will likely

complement the existing clinical imaging technologies and improve the detection specificity

of malignant cancers. The technical advances in PAM have lifted the traditional tradeoffs

between imaging speed, FOV, and detection sensitivity. Powered by super-fast pulsed lasers and

novel scanning mechanisms, PAM has achieved high-speed, high-resolution imaging over the

similar FOVof conventional CCD-based optical microscopes and can monitor the neurovascular

coupling of the entire mouse cortex. The substantial improvement in imaging throughput has

enabled implementations of PAM in portable and wearable formats, allowing for longitudinal

monitoring of biological functions in freely moving animals or awake patients, with negligible

motion artifacts. Both PACT and PAM can greatly benefit from the new advances in optical

ultrasound sensors with small size, large detection bandwidth, and wide receiving angle.

Innovations in the fabrication process, materials, and stabilization methods are critically

important to address the limitations in the optical sensor

’

s speed, scalability, and stability. The

relatively high detection sensitivity of small optical sensors is particularly beneficial for endo-

scopic and wearable PA applications. Moreover, the fast-evolving deep learning technologies

have been quickly adopted in PAT to improve the signal-to-noise ratio, inverse image reconstruc-

tion, and image post-processing. For technical challenges in PAT difficult to address using hard-

ware solutions, deep learning approaches may provide effective data-driven solutions that

impose minimal impact on the system

’

s complexity and cost.

Looking forward, we expect that PAT will grow at an accelerating speed in both technology

development and biomedical applications. With more commercially available PAT systems tail-

ored for clinical practice, the user base will also experience a fast expansion in the next several

years, resulting in a large number of published clinical studies. Of particular importance is the

first FDA-approved PAT system by Seno Medical Instruments, Inc., which has paved the way for

more commercial PAT systems to receive regulatory clearance. Developing low-cost PAT sys-

tems is an important step that can help improve its accessibility by the biomedical community.

In particular, low-cost light sources such as laser diodes and light emitting diodes can signifi-

cantly reduce the system cost of both PAM and PACT, and accelerate the technical translation to

clinical practice. The success of commercial PAT products will in turn provide strong incentives

for key industrial partners to develop products that are specially optimized for PAT, such as

high-power, high-speed lasers with relaxed coherence; low-cost CMUT or PMUT arrays with

a large number of elements;

high-channel-number, high-speed data acquisition systems with

built-in amplification capability; and high-speed GPU systems with large on-chip memory.

Enabled by these updated system components, which are often the bottlenecks in PAT technol-

ogy, the next wave of technological breakthroughs will naturally follow, including (i) real-time

volumetric PACT systems for human imaging, such as breast cancer screening; (ii) high-speed,

high-resolution PAM of neuronal activities enabled by novel voltage- or calcium-sensitive PA

probes; (iii) highly-compact endoscopic and intravascular PAT enabled by optical ultrasound

sensors; (iv) single-organelle or single-molecule PA imaging enabled by super-resolution

mechanisms; (v) large-scale, high-speed 3D modeling of PA signal generation and propagation

in a complex system; and (vi) robust quantitative PAT of tissue functions and molecular

compositions enabled by deep learning approaches. Finally, we envision that, in the big data

era, the next generation of PAT technologies will likely have artificial intelligence incorporated

at every step of system development. The light illumination, ultrasound detection, scanning

mechanism, data acquisition, and image formation can be optimized by the accompanying

machine learning models, which will make it possible to achieve the next generation of smart

PA technologies.

Disclosures

The authors have no financial conflicts of interest to disclose related to the content of this article.

L.V.W. has a financial interest in Microphotoacoustics, Inc., Cal-PACT, LLC, and Union

Photoacoustic Technologies, Ltd., none of which supported this work.

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-13

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

Acknowledgments

We thank Dr. Caroline Connor for editing the manuscript and Drs. Li Lin and Xiaoyi Zhu for

preparing the figures. This work was sponsored by the United States National Institutes of Health

(NIH) under Grant Nos. R35 CA220436 (Outstanding Investigator Award), U01 NS099717

(BRAIN Initiative), U01 EB029823 (BRAIN Initiative), and R01 EB028277 (to L.V.W), as well

as NIH Grant Nos. R01 EB028143, R01 NS111039, RF1 NS115581 (BRAIN Initiative), R21

EB027304, and R21EB027981; Duke Institute of Brain Science Incubator Award; American

Heart Association Collaborative Sciences Award (18CSA34080277); and the Chan Zuckerberg

Initiative under Grant No. 2020-226178 (to J.Y).

References

1. D. Das et al.,

“

Another decade of photoacoustic imaging,

”

Phys. Med. Biol.

(5), 05TR01

(2021).

2. L. H. V. Wang and J. J. Yao,

“

A practical guide to photoacoustic tomography in the life

sciences,

”

Nat. Methods

(8), 627

–

638 (2016).

3. Y. Zhou, J. Yao, and L. V. Wang,

“

Tutorial on photoacoustic tomography,

”

J. Biomed. Opt.

(6), 061007 (2016).

4. L. V. Wang,

“

Tutorial on photoacoustic microscopy and computed tomography,

”

IEEE J.

Sel. Top. Quantum Electron.

(1), 171

–

179 (2008).

5. S. Zackrisson, S. M. W. Y. van de Ven, and S. S. Gambhir,

“

Light in and sound out: emerg-

ing translational strategies for photoacoustic imaging,

”

Cancer Res.

(4), 979

–

1004

(2014).

6. P. Burgholzer, H. Grun, and A. Sonnleitner,

“

Sounding out fluorescent proteins,

”

Nat.

Photonics

(7), 378

–

379 (2009).

7. D. Razansky et al.,

“

Multispectral opto-acoustic tomography of deep-seated fluorescent

proteins

in vivo

”

Nat. Photonics

(7), 412

–

417 (2009).

8. J. J. Yao and L. H. V. Wang,

“

Photoacoustic microscopy,

”

Laser Photonics Rev.

(5),

758

–

778 (2013).

9. J. M. Yang et al.,

“

Simultaneous functional photoacoustic and ultrasonic endoscopy of

internal organs

in vivo

”

Nat. Med.

(8), 1297

–

1302 (2012).

10. P. Hajireza, W. Shi, and R. Zemp,

“

Label-free

in vivo

GRIN-lens optical resolution photo-

acoustic micro-endoscopy,

”

Laser Phys. Lett.

(5), 055603 (2013).

11. J. James et al.,

“

Coherent fiber bundle based integrated photoacoustic, ultrasound and fluo-

rescence imaging (PAUSFI) for endoscopy and diagnostic bio-imaging applications,

”

Laser Phys.

(8), 085608 (2014).

12. P. Shao et al.,

“

Integrated micro-endoscopy system for simultaneous fluorescence and

optical-resolution photoacoustic imaging,

”

J. Biomed. Opt.

(7), 076024 (2012).

13. J. M. Yang et al.,

“

Photoacoustic endoscopy,

”

Opt. Lett.

(10), 1591

–

1593 (2009).

14. K. Jansen, G. van Soest, and A. F. van der Steen,

“

Intravascular photoacoustic imaging:

a new tool for vulnerable plaque identification,

”

Ultrasound Med. Biol.

(6), 1037

–

1048

(2014).

15. R. Ansari et al.,

“

Photoacoustic endoscopy probe using a coherent fibre-optic bundle,

”

Proc. SPIE

9539

, 953905 (2015).

16. S. Manohar and D. Razansky,

“

Photoacoustics: a historical review,

”

Adv. Opt. Photonics

(4), 586

–

617 (2016).

17. D. Razansky, J. Baeten, and V. Ntziachristos,

“

Sensitivity of molecular target detection by

multispectral optoacoustic tomography (MSOT),

”

Med. Phys.

(3), 939

–

945 (2009).

18. B. Cox et al.,

“

Quantitative spectroscopic photoacoustic imaging: a review,

”

J. Biomed.

Opt.

(6), 061202 (2012).

19. N. C. Deliolanis et al.,

“

Deep-tissue reporter-gene imaging with fluorescence and opto-

acoustic tomography: a performance overview,

”

Mol. Imaging Biol.

(5), 652

–

660

(2014).

20. A. Taruttis and V. Ntziachristos,

“

Advances in real-time multispectral optoacoustic imaging

and its applications,

”

Nat. Photonics

(4), 219

–

227 (2015).

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-14

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

21. G. Wissmeyer et al.,

“

Looking at sound: optoacoustics with all-optical ultrasound detec-

tion,

”

Light Sci. Appl.

(1), 1

–

16 (2018).

22. Y. Zhu et al.,

“

Towards clinical translation of LED-based photoacoustic imaging: a review,

”

Sensors

(9), 2484 (2020).

23. W. Xia et al.,

“

Handheld real-time LED-based photoacoustic and ultrasound imaging

system for accurate visualization of clinical metal needles and superficial vasculature to

guide minimally invasive procedures,

”

Sensors

(5), 1394 (2018).

24. Y. Zhu et al.,

“

Light emitting diodes based photoacoustic imaging and potential clinical

applications,

”

Sci. Rep.

(1), 9885 (2018).

25. Y. Zhu et al.,

“

LED-based photoacoustic imaging for monitoring angiogenesis in fibrin

scaffolds,

”

Tissue Eng. Part C Methods

(9), 523

–

531 (2019).

26. M. Erfanzadeh and Q. Zhu,

“

Photoacoustic imaging with low-cost sources: a review,

”

Photoacoustics

–

11 (2019).

27. A. Fatima et al.,

“

Review of cost reduction methods in photoacoustic computed tomog-

raphy,

”

Photoacoustics

, 100137 (2019).

28. P. K. Upputuri and M. Pramanik,

“

Fast photoacoustic imaging systems using pulsed laser

diodes: a review,

”

Biomed. Eng. Lett.

(2), 167

–

181 (2018).

29. J. Brunker et al.,

“

Photoacoustic imaging using genetically encoded reporters: a review,

”

J. Biomed. Opt.

(7), 070901 (2017).

30. H. Wang et al.,

“

MEMS ultrasound transducers for endoscopic photoacoustic imaging

applications,

”

Micromachines

(10), 928 (2020).

31. Q. Fu et al.,

“

Photoacoustic imaging: contrast agents and their biomedical applications,

”

Adv. Mater.

(6), 1805875 (2019).

32. M. Li et al.,

“

Sound out the deep colors: photoacoustic molecular imaging at new depths,

”

Mol Imaging

, 1536012120981518 (2020).

33. T. Zhao et al.,

“

Minimally invasive photoacoustic imaging: current status and future

perspectives,

”

Photoacoustics

, 100146 (2019).

34. S. S. S. Choi and A. Mandelis,

“

Review of the state of the art in cardiovascular endoscopy

imaging of atherosclerosis using photoacoustic techniques with pulsed and continuous-

wave optical excitations,

”

J. Biomed. Opt.

(8), 080902 (2019).

35. C. Yang et al.,

“

Review of deep learning for photoacoustic imaging,

”

Photoacoustics

, 100215 (2021).

36. J. Gröhl et al.,

“

Deep learning for biomedical photoacoustic imaging: a review,

”

Photoacoustics

, 100241 (2021).

37. A. DiSpirito, III et al.,

“

Sounding out the hidden data: a concise review of deep learning

in photoacoustic imaging,

”

Exp. Biol. Med.

1355

–

1367 (2021).

38. W. Choi et al.,

“

Clinical photoacoustic imaging platforms,

”

Biomed. Eng. Lett.

(2),

139

–

155 (2018).

39. A. B. E. Attia et al.,

“

A review of clinical photoacoustic imaging: current and future

trends,

”

Photoacoustics

, 100144 (2019).

40. J. Kim et al.,

“

Towards clinical photoacoustic and ultrasound imaging: probe improvement

and real-time graphical user interface,

”

Exp. Biol. Med.

245

(4), 321

–

329 (2020).

41. S. Manohar and S. S. Gambhir,

“

Clinical photoacoustic imaging,

”

Photoacoustics

100196 (2020).

42. X. L. Dean-Ben and D. Razansky,

“

Adding fifth dimension to optoacoustic imaging:

volumetric time-resolved spectrally enriched tomography,

”

Light Sci. Appl.

, e137

(2014).

43. Y. Wang et al.,

“

In vivo

three-dimensional photoacoustic imaging based on a clinical

matrix array ultrasound probe,

”

J. Biomed. Opt.

(6), 061208 (2012).

44. J. Laufer et al.,

“

Three-dimensional noninvasive imaging of the vasculature in the mouse

brain using a high resolution photoacoustic scanner,

”

Appl. Opt.

(10), D299

–

D306

(2009).

45. R. A. Kruger et al.,

“

Photoacoustic angiography of the breast,

”

Med. Phys.

(11),

6096

–

6100 (2010).

46. L. I. O. America,

“

American National Standard for the safe use of lasers,

”

American

National Standard Institute, New York (2007).

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-15

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

47. Y. Matsumoto et al.,

“

Visualising peripheral arterioles and venules through high-resolution

and large-area photoacoustic imaging,

”

Sci. Rep.

, 14930 (2018).

48. A. Ron et al.,

“

Flash scanning volumetric optoacoustic tomography for high resolution

whole-body tracking of nanoagent kinetics and biodistribution,

”

Laser Photonics Rev.

(3), 2000484 (2021).

49. L. Lin et al.,

“

High-speed three-dimensional photoacoustic computed tomography for

preclinical research and clinical translation,

”

Nat. Commun.

(1), 882 (2021).

50. S. M. Schoustra et al.,

“

Twente Photoacoustic Mammoscope 2: system overview and three-

dimensional vascular network images in healthy breasts,

”

J. Biomed. Opt.

(12), 121909

(2019).

51. E. Zhang, J. Laufer, and P. Beard,

“

Backward-mode multiwavelength photoacoustic scan-

ner using a planar Fabry

–

Perot polymer film ultrasound sensor for high-resolution three-

dimensional imaging of biological tissues,

”

Appl. Opt.

(4), 561

–

577 (2008).

52. A. P. Jathoul et al.,

“

Deep

in vivo

photoacoustic imaging of mammalian tissues using

a tyrosinase-based genetic reporter,

”

Nat. Photonics

(4), 239

–

246 (2015).

53. M. Heijblom et al.,

“

The state of the art in breast imaging using the Twente Photoacoustic

Mammoscope: results from 31 measurements on malignancies,

”

Eur. Radiol.

(11),

3874

–

3887 (2016).

54. D. Piras et al.,

“

Photoacoustic imaging of the breast using the twente photoacoustic mam-

moscope: present status and future perspectives,

”

IEEE J. Sel. Top. Quantum Electron.

(4), 730

–

739 (2010).

55. S. Gottschalk et al.,

“

Rapid volumetric optoacoustic imaging of neural dynamics across

the mouse brain,

”

Nat. Biomed. Eng.

(5), 392

–

401 (2019).

56. I. Ivankovic et al.,

“

Real-time volumetric assessment of the human carotid artery: handheld

multispectral optoacoustic tomography,

”

Radiology

291

(1), 45

–

50 (2019).

57. X. L. Dean-Ben, S. J. Ford, and D. Razansky,

“

High-frame rate four dimensional opto-

acoustic tomography enables visualization of cardiovascular dynamics and mouse heart

perfusion,

”

Sci. Rep.

, 10133 (2015).

58. R. Ansari et al.,

“

Miniature all-optical flexible forward-viewing photoacoustic endoscopy

probe for surgical guidance,

”

Opt. Lett.

(22), 6238

–

6241 (2020).

59. N. Huynh et al.,

“

Sub-sampled Fabry

–

Perot photoacoustic scanner for fast 3D imaging,

”

Proc. SPIE

10064

, 100641Y (2017).

60. W. J. Westerveld et al.,

“

Sensitive, small, broadband and scalable optomechanical ultra-

sound sensor in silicon photonics,

”

Nat. Photonics

, 341

–

345 (2021).

61. R. Shnaiderman et al.,

“

A submicrometre silicon-on-insulator resonator for ultrasound

detection,

”

Nature

585

(7825), 372

–

378 (2020).

62. Y. Liang et al.,

“

Fiber-laser-based ultrasound sensor for photoacoustic imaging,

”

Sci. Rep.

, 40849 (2017).

63. Y. Liang et al.,

“

Photoacoustic computed tomography by using a multi-angle scanning

fiber-laser ultrasound sensor,

”

Opt. Express

(6), 8744

–

8752 (2020).

64. L. H. V. Wang and S. Hu,

“

Photoacoustic tomography:

in vivo

imaging from organelles to

organs,

”

Science

335

(6075), 1458

–

1462 (2012).

65. B. Rao et al.,

“

Real-time four-dimensional optical-resolution photoacoustic microscopy with

Au nanoparticle-assisted subdiffraction-limit resolution,

”

Opt. Lett.

(7), 1137

–

1139 (2011).

66. J. Y. Liang et al.,

“

Random-access optical-resolution photoacoustic microscopy using

a digital micromirror device,

”

Opt. Lett.

(15), 2683

–

2686 (2013).

67. W. Song et al.,

“

Integrating photoacoustic ophthalmoscopy with scanning laser ophthal-

moscopy, optical coherence tomography, and fluorescein angiography for a multimodal

retinal imaging platform,

”

J. Biomed. Opt.

(6), 061206 (2012).

68. W. Qin et al.,

“

Large-field-of-view optical resolution photoacoustic microscopy,

”

Opt.

Express

(4), 4271

–

4278 (2018).

69. C. Lee, J. Y. Kim, and C. Kim,

“

Recent progress on photoacoustic imaging enhanced

with microelectromechanical systems (MEMS) technologies,

”

Micromachines

(11),

584 (2018).

70. J. Y. Kim et al.,

“

Fast optical-resolution photoacoustic microscopy using a 2-axis water-

proofing MEMS scanner,

”

Sci. Rep.

, 7932 (2015).

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-16

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

71. J. Yao et al.,

“

High-speed label-free functional photoacoustic microscopy of mouse brain

in action,

”

Nat. Methods

(5), 407

–

410 (2015).

72. L. Lin et al.,

“

Handheld optical-resolution photoacoustic microscopy,

”

J. Biomed. Opt.

(4), 041002 (2016).

73. C. Zhang et al.,

“

Multiscale high-speed photoacoustic microscopy based on free-space

light transmission and a MEMS scanning mirror,

”

Opt. Lett.

(15), 4312

–

4315 (2020).

74. B. Lan et al.,

“

High-speed widefield photoacoustic microscopy of small-animal hemo-

dynamics,

”

Biomed. Opt. Express

(10), 4689

–

4701 (2018).

75. Y. Li et al.,

“

Snapshot photoacoustic topography through an ergodic relay for high-

throughput imaging of optical absorption,

”

Nat. Photonics

(3), 164

–

170 (2020).

76. J. Chen et al.,

“

Wide-field polygon-scanning photoacoustic microscopy of oxygen

saturation at 1-MHz A-line rate,

”

Photoacoustics

, 100195 (2020).

77. K. Maslov et al.,

“

Optical-resolution photoacoustic microscopy for

in vivo

imaging of

single capillaries,

”

Opt. Lett.

(9), 929

–

931 (2008).

78. L. D. Wang et al.,

“

Fast voice-coil scanning optical-resolution photoacoustic microscopy,

”

Opt. Lett.

(2), 139

–

141 (2011).

79. Z. X. Xie et al.,

“

Laser-scanning optical-resolution photoacoustic microscopy,

”

Opt. Lett.

(12), 1771

–

1773 (2009).

80. J. Yao et al.,

“

Water-immersible MEMS scanning mirror enhanced optical-resolution

photoacoustic microscopy,

”

OSA Tech. Dig., Biomed. Opt.

, BSu3A.54 (2012).

81. L. D. Wang et al.,

“

Video-rate functional photoacoustic microscopy at depths,

”

J. Biomed.

Opt.

(10), 106007 (2012).

82. M. Moothanchery et al.,

“

Acoustic resolution photoacoustic microscopy based on micro-

electromechanical systems scanner,

”

J. Biophotonics

(2), e201960127 (2020).

83. J. W. Baik et al.,

“

Ultrawide-field acoustic-resolution photoacoustic microscopy,

”

Proc.

SPIE

10878

, 1087815 (2019).

84. M. Moothanchery et al.,

“

High-speed simultaneous multiscale photoacoustic microscopy,

”

J. Biomed. Opt.

(8), 086001 (2019).

85. A. DiSpirito, III et al.,

“

Reconstructing undersampled photoacoustic microscopy images

using deep learning,

”

IEEE Trans. Med. Imaging

(2), 562

–

570 (2020).

86. B. T. Cox and P. C. Beard,

“

Photoacoustic tomography with a single detector in a rever-

berant cavity,

”

J. Acoust. Soc. Am.

125

(3), 1426

–

1436 (2009).

87. Y. Li et al.,

“

Photoacoustic topography through an ergodic relay for functional imaging

and biometric application

in vivo

”

J. Biomed. Opt.

(7), 070501 (2020).

88. M. Brown et al.,

“

Reverberant cavity photoacoustic imaging,

”

Optica

(6), 821

–

822

(2019).

89. K. Park et al.,

“

Handheld photoacoustic microscopy probe,

”

Sci. Rep.

(1), 1

–

15 (2017).

90. S. Liu et al.,

“

Handheld photoacoustic imager for theranostics in 3D,

”

IEEE Trans. Med.

Imaging

(9), 2037

–

2046 (2019).

91. Y. Zhou et al.,

“

Noninvasive determination of melanoma depth using a handheld photo-

acoustic probe,

”

J. Invest. Dermatol.

137

(6), 1370 (2017).

92. A. Dangi et al.,

“

Towards a low-cost and portable photoacoustic microscope for point-of-

care and wearable applications,

”

IEEE Sens. J.

(13), 6881

–

6888 (2020).

93. Q. Chen, H. Xie, and L. Xi,

“

Wearable optical resolution photoacoustic microscopy,

”

J. Biophotonics

(8), e201900066 (2019).

94. S. Liu et al.,

“

Toward wearable healthcare: a miniaturized 3d imager with coherent

frequency-domain photoacoustics,

”

IEEE Trans. Biomed. Circuits Syst.

(6), 1417

–

1424

(2019).

95. Q. Chen, H. K. Xie, and L. Xi,

“

Wearable optical resolution photoacoustic microscopy,

”

J. Biophotonics

(8), e201900066 (2019).

96. J. Tang et al.,

“

Wearable 3-D photoacoustic tomography for functional brain imaging in

behaving rats,

”

Sci. Rep.

(1), 1

–

10 (2016).

97. A. M. Winkler, K. Maslov, and L. V. Wang,

“

Noise-equivalent sensitivity of photoacous-

tics,

”

J. Biomed. Opt.

(9), 097003 (2013).

98. X. Zhu et al.,

“

Ultrasonic detection based on polarization-dependent optical reflection,

”

Opt. Lett.

(3), 439

–

441 (2017).

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-17

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021

99. S. M. Maswadi et al.,

“

All-optical optoacoustic microscopy based on probe beam deflec-

tion technique,

”

Photoacoustics

(3), 91

–

101 (2016).

100. P. Hajireza et al.,

“

Non-interferometric photoacoustic remote sensing microscopy,

”

Light

Sci. Appl.

(6), e16278

–

e16278 (2017).

101. E. Z. Zhang et al.,

“

In vivo

high-resolution 3D photoacoustic imaging of superficial

vascular anatomy,

”

Phys. Med. Biol.

(4), 1035

–

1046 (2009).

102. N. Huynh et al.,

“

Photoacoustic imaging using an 8-beam Fabry

–

Perot scanner,

”

Proc.

SPIE

9708

, 97082L (2016).

103. R. Nuster et al.,

“

Full field detection in photoacoustic tomography,

”

Opt. Express

(6),

6288

–

6299 (2010).

104. M. Lamont and P. C. Beard,

“

2D imaging of ultrasound fields using CCD array to map

output of Fabry

–

Perot polymer film sensor,

”

Electron. Lett.

(3), 187

–

189 (2006).

105. H. Li et al.,

“

Disposable ultrasound-sensing chronic cranial window by soft nanoimprint-

ing lithography,

”

Nat. Commun.

(1), 4277 (2019).

106. B. Chen, Y. Chen, and C. Ma,

“

Photothermally tunable Fabry

–

Pérot fiber interferometer

for photoacoustic mesoscopy,

”

Biomed. Opt. Express

(5), 2607

–

2618 (2020).

107. M. M. Najafabadi et al.,

“

Deep learning applications and challenges in big data analytics,

”

J. Big Data

(1), 1 (2015).

108. M. I. Razzak, S. Naz, and A. Zaib,

“

Deep learning for medical image processing: overview,

challenges and future,

”

ArXiv170406825 (2017).

109. J.-G. Lee et al.,

“

Deep learning in medical imaging: general overview,

”

Korean J. Radiol.

(4), 570 (2017).

110. Y. Rivenson et al.,

“

Deep learning microscopy,

”

Optica

(11), 1437

–

1443 (2017).

111. B. J. Erickson et al.,

“

Machine learning for medical imaging,

”

Radiographics

(2),

505

–

515 (2017).

112. G. Litjens et al.,

“

A survey on deep learning in medical image analysis,

”

Med. Image Anal.

,60

–

88 (2017).

113. B. Sahiner et al.,

“

Deep learning in medical imaging and radiation therapy,

”

Med. Phys.

(1), e1

–

e36 (2019).

114. J. Martorell-Marugán et al.,

“

Deep learning in Omics data analysis and precision medi-

cine,

”

Computational Biology

, H. Husi, Ed., pp. 37

–

53, Codon Publications, Singapore

(2019).

115. C. Shen et al.,

“

An introduction to deep learning in medical physics: advantages, potential,

and challenges,

”

Phys. Med. Biol.

(5), 05TR01 (2020).

116. P. Pradhan et al.,

“

Deep learning a boon for biophotonics?,

”

J. Biophotonics

(6),

e201960186 (2020).

117. A. Hauptmann and B. T. Cox,

“

Deep learning in photoacoustic tomography: current

approaches and future directions,

”

J. Biomed. Opt.

(11), 112903 (2020).

118. M. Agranovsky and P. Kuchment,

“

Uniqueness of reconstruction and an inversion pro-

cedure for thermoacoustic and photoacoustic tomography,

”

Inverse Prob.

, 2089

–

2102

(2007).

119. S. Gutta et al.,

“

Deep neural network-based bandwidth enhancement of photoacoustic

data,

”

J. Biomed. Opt.

(11), 116001 (2017).

120. N. Davoudi, X. L. Deán-Ben, and D. Razansky,

“

Deep learning optoacoustic tomography

with sparse data,

”

Nat. Mach. Intell.

(10), 453

–

460 (2019).

121. T. Vu et al.,

“

A generative adversarial network for artifact removal in photoacoustic com-

puted tomography with a linear-array transducer,

”

Exp. Biol. Med.

245

(7), 597

–

605 (2020).

122. A. Hauptmann et al.,

“

Approximate k-space models and deep learning for fast photoacous-

tic reconstruction,

”

Lect. Notes Comput. Sci.

11074

, 103

–

111 (2018).

123. A. Hauptmann et al.,

“

Model-based learning for accelerated, limited-view 3-d photoacous-

tic tomography,

”

IEEE Trans. Med. Imaging

(6), 1382

–

1393 (2018).

124. Y. E. Boink, S. Manohar, and C. Brune,

“

A partially-learned algorithm for joint photo-

acoustic reconstruction and segmentation,

”

IEEE Trans. Med. Imaging

(1), 129

–

139

(2020).

125. A. DiSpirito, III et al.,

“

Reconstructing undersampled photoacoustic microscopy images

using deep learning,

”

IEEE Trans. Med. Imaging

(2), 562

–

570 (2021).

Yao and Wang: Perspective on fast-evolving photoacoustic tomography

Journal of Biomedical Optics

060602-18

June 2021

•

Vol. 26(6)

Downloaded From: https://www.spiedigitallibrary.org/journals/Journal-of-Biomedical-Optics on 06 Jul 2021