You are on page 1of 15

This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 1

Exploring the Acoustic and Dynamic


Characteristics of Phase-Change Droplets
Ching-Hsiang Fan, Wei-Fu Kao, Shih-Tsung Kang, Yi-Ju Ho, Chih-Kuang Yeh, Senior Member,
IEEE

methods including embolotherapy, drug delivery, enhancement


Abstract—Acoustic droplet vaporization (ADV) provides the of the permeability of the blood–brain barrier (BBB), and the
on-demand production of bubbles for use in ultrasound (US)- destruction of tumor vessels [4-12] However, these desired
based diagnostic and therapeutic applications. The droplet-to- goals are only achievable by the successful occurrence of ADV.
bubble transition process has been shown to involve localized
internal gas nucleation, followed by a volume expansion of three-
Also, overdosing of ADV potentially results in side effects such
to fivefold and inertial bubble oscillation, all of which take place as irreversible cellular damage, cell death, or undesired gas
within a few microseconds. Monitoring these ADV processes is embolism [13, 14]. This situation means that developing a
important in gauging the mechanical effects of phase-change technique for the real-time, high-sensitivity, and accurate
droplets in a biological environment, but this is difficult to achieve detection of ADV event occurrences could benefit its clinical
using regular optical observations. In this study we utilized translation.
acoustic characterization (i.e., simultaneous passive cavitation
detection [PCD] as well as active cavitation detection [ACD]) to
ADV events have mainly been monitored using US B-mode
investigate the acoustic signatures emitted from phase-change imaging and optical observations. The bubbles generated from
droplets ADV, and determined their correlations with the physical the ADV process are capable of increasing the amplitude of
behaviors observed using high-speed optical imaging. The backscattered US by up to 20–30 dB, thereby serving as
experimental results showed that activation with three-cycle 5- effective US B-mode imaging contrast agents [10, 15].
MHz US pulse resulted in the droplets (diameter: 3.0–6.0 μm) Although this method can visualize ADV events in deep tissue,
overexpanding and undergoing damped oscillation before settling
to bubbles with a final diameter. Meanwhile, a broadband shock
its low sensitivity may result in it only been capable of detecting
wave was observed at the beginning of the PCD signal. The intense large numbers of events. Optical observations allow the
fluctuations of the ACD signal revealed that the shock wave arose occurrence of ADV events to be monitored with a high
from the inertial cavitation of nucleated small gas pockets in the temporal resolution, but it is only suitable for in vitro
droplets. It was particularly interesting that another shock-wave applications or superficial regions [16, 17].
signal with a much lower acoustic frequency (<2 MHz) was Interpreting acoustic signals emitted during the ADV process
observed at about 5 μs after the first half signal. This signal
coincided with the reduction of the ACD signal amplitude that
by passive cavitation detection (PCD) probably be a feasible
indicated the rebound of the transforming bubble. Since internal tool for overcoming the limitations of the present detection
gas nucleation is a crucial process of ADV, the first half signal may methods. PCD relies on the distinct Fourier spectra of acoustic
indicate the occurrence of an ADV event, and the second half signals, in which the strengths of stable cavitation hallmarks
signal may further reveal the degrees of expansion and oscillation (harmonic signals, subharmonic signal and ultraharmonic
of the bubble. These acoustic signatures provide opportunities for signal) and inertial cavitation markers (broadband signals) can
monitoring ADV dynamics based on the detection of acoustic
signals.
be measured in real time with high sensitivity and a good tissue
penetration depth. When utilizing microbubble-based contrast
Index Terms—acoustic droplets, ultrasound, acoustic droplet agents in ultrasonography, the activity of the microbubbles (e.g.,
vaporization, acoustic signatures, high-speed microscopy harmonic signals for steady-state oscillations or broadband
spectral content for destruction processes), the location of
microbubbles (nonlinear signals), the occurrence of brain
I. INTRODUCTION damage, and even the amount of delivered chemodrugs can be
characterized by analyzing the acoustic signals produced by
T he phospholipid shell with a liquid perfluorocarbon core of
droplet can be converted into gas bubbles through
ultrasound (US) sonication—called acoustic droplet
microbubbles during the onset of US sonication [18-21].
Previous studies have demonstrated that the violent physical
vaporization (ADV) [1-3]—has provided a multifunction behaviors occurring in ADV result in pressure waves and
platform for noninvasively addressing various diseases via produce acoustic signals [22, 23]. Reznik et al. suggested that

This paragraph of the first footnote will contain the date on which you Ching-Hsiang Fan is with Department of Biomedical Engineering, National
submitted your paper for review. This work was supported in part by support Cheng Kung University, Tainan, Taiwan, and Medical Device Innovation
by the Ministry of Science and Technology (MOST) of Taiwan, under grants Center, National Cheng Kung University, Tainan, Taiwan. Wei-Fu Kao, Shih-
nos. 108-2221-E-007-041-MY3, 108-2221-E-007-040-MY3, and 106-2218-E- Tsung Kang, Yi-Ju Ho, and Chih-Kuang Yeh * are with Department of
007-022-MY3. Biomedical Engineering and Environmental Sciences, National Tsing Hua
University, Hsinchu, Taiwan (e-mail: ckyeh@mx.nthu.edu.tw).

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 2

when dodecafluoropentane droplets were vaporizing into sonication system and a high-speed microscopy imaging system
bubbles, the changes in the amplitudes of the scattered was used to perform the experiments of this study [3]. The high-
fundamental and harmonic acoustic signals over time could be speed microscopy imaging system was setup by interfacing a
utilize to distinguish growing bubbles from stable bubbles [24]. high-speed camera (model FASTCAMSA4, Photron Ltd.,
Rapoport et al. found that the subharmonic frequencies and Tokyo, Japan) to a microscope (model IX71, Olympus
broadband noise detected during the ADV of perfluoropentane Corporation) for obtain serial optical images of the droplets
(PFP) droplets resulted from the oscillation and unstable growth ADV at 500,000 fps. The imaging system was attached a self-
of ADV bubbles, respectively [25]. Sheeran et al. demonstrated made water tank. For US sonication system, a 5-MHz US
that US-activated decafluorobutane droplets and transducer (model SU-128, Sonic Concepts, WA, USA) (Table
octafluoropropane droplets would overexpand and suffer 1) was confocally positioned with a waterproof 40× objective
unforced radial oscillations while settling to a stable bubble (Carl Zeiss, Tokyo, Japan). The vaporization pulses (5-MHz, 3
diameter. These radial oscillations generated several cycles, single pulse, and 7-8 MPa) were transmitted by the US
narrowband peaks of the spectral range from 0.1 to 1.3 MHz transducer. Our previous study has reported that the acoustic
[26]. vaporization threshold of droplet (1.5-3.0 µm) with 5-MHz
The findings of these previous studies were heterogeneous ultrasound sonication was approximately 6 MPa [27]. In order
due to inconsistencies in the experimental conditions, including to ensure the successfully ADV occurrence, we therefore chose
acoustic parameters, droplet components, observation periods, the acoustic pressure of 8 MPa to trigger ADV in this study.
and numbers of observed droplets. This situation makes it The US transducer was driven by an arbitrary waveform
necessary to characterize the correlation between ADV generator (model AWG 2005, Tektronix, CA, USA) and a
behaviors and acoustic signals before the acoustic signals can radiofrequency (RF) power amplifier (model A150, E&I, NY,
be used as indicators of ADV events in vivo. USA) to transmit the vaporization pulses. Note that the water
In this study we simultaneously used PCD and active tank was filled with degassed as well as deionized water at 37°C
cavitation detection (ACD) approaches to investigate the for mimicking the in vivo temperature throughout the
acoustic signals emitted from phase-change droplets during experiments. In order to avoid the occurrence of inertial
ADV, and determined their correlations with the physical cavitation from medium surrounding the droplet during
behaviors observed using high-speed optical imaging. The experiment, all the solutions used in this study have been
origin and categories of ADV acoustic signals were first degassed overnight. The acoustic pressure was measured by a
investigated under the setup of single-droplet ADV. The polyvinylidene difluoride type hydrophone (model HGL-0085,
correlation between ADV acoustic signals and the droplet ONDA Corporation, CA, USA) in with distilled and degassed
initial size were then revealed. Finally, we used the water at 25°C.
characteristic signatures in groups of droplets to develop a A droplet with diameter of 3.0, 4.0, 5.0, or 6.0 μm was
detection system based on M-mode imaging. administrated into a self-made agar tube (diameter: 200 μm)
and settled at the focus of microscopy via a microinjector
II. MATERIALS AND METHODS (model IM-6, Narishige, Tokyo, Japan) to ensure that only one
droplet appeared at the optical image. The desired size of
A. Preparation of droplets
droplet was optically selected before the ADV experiments by
In order to prepare droplets, 5 mol% DSPE-PEG2000 microscopy’s software (Photron FASTCAM Viewer, Photron
(distearoylphosphatidylethanolamine–polyethylene glycol Ltd.).
2000, Avanti Polar Lipids, AL, USA) and 95 mol% DSPC A 2.25-MHz or 25-MHz US transducer (model V305 or
(distearoylphosphatidylcholine, Avanti Polar Lipids) were well model V324, Olympus NDT, MA, USA) was confocally
dissolved in chloroform [3]. The lipid mixture was then drained arranged with a mutual focus for simultaneous PCD and ACD,
and dissolved with 1 wt% glycerol-containing phosphate- as illustrated in Figure 2a and 2b. Note that the beam axes of
buffered saline. Subsequently, 100 μL of PFP (ABCR, these two transducers were perpendicular to the agar tube in
Karlsruhe, Germany) was added into the solution at 4 °C. The order to reduce the backscattered energy of US. The 2.25 MHz
droplets were formed by 40 kHz sonication (100 W, 2510, transducer was used to detect ADV (as usually subhamornics
Branson, CT, USA) at 20 ℃ for 5 min. and wide-band occurrence are characteristic features for ADV
The size distribution and concentration of prepared droplets detection), and the 25 MHz transducer was suited for
were estimated using a Coulter counter (Multisizer 3, Beckman monitoring the bubble fluctuation during ADV. The frequency
Coulter, CA, USA). The morphology of the droplets was response of 2.25 MHz transducer was shown in Fig. S1.
visualized by a microscope (model IX71, Olympus Corporation, For PCD, the transducer was used to passively collect the
Tokyo, Japan). Figure 1 demonstrated that the diameter and acoustic signatures emitted during ADV. The signal was then
concentration of droplets were 2.8±1.4 μm and (1.2±0.1) × 109 digitized by an oscilloscope (model LT354, LeCroy
droplets/mL, respectively. Corporation, NY, USA) The sampling rate of oscilloscope was
set at 1 Gsample/s. The digitized signal was sent to a computer
B. High-speed photography and cavitation detection by a GPIB interface and then analyzed offline via MATLAB
An integrated acousto-optical system comprising an US software (MathWorks, MA, USA).
For ACD, the 25-MHz US transducer was operated in pulse-

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 3

echo mode to interrogate the morphological evolution of statistical analyses were performed by one-way ANOVA, with
bubbles using a low-amplitude US pulse (200 cycles, single the Bonferroni post-hoc test applied for multiple comparisons
pulse, and 1 MPa). A waveform generator (model AWG 2041, (SPSS version 13.0, SPSS, IBM, NY, USA). A p value of less
Tektronix) was utilized to produce a detection pulse that was than 0.05 was considered to be indicative of a significant
amplified with a radiofrequency (RF) power amplifier (model difference.
A150, E&I,) to trigger the transducer. The RF signal was firstly
obtained via a diode limiter/transformer diplexer circuit (model III. RESULTS
DIP-3, Matec Instruments NDT, MA, USA) and amplified by a
A. Two acoustic signatures appeared during droplet
pulser/receiver (model 5073PR, Olympus NDT). The signal vaporization by PCD observation
was then digitized by the oscilloscope and stored in the
computer. The signal was then processed offline by MATLAB Figure 2C shows typical PCD and ACD signals for single 5-
software. μm droplets subjected to ADV upon exposure to US at 8 MPa.
A 0.5-μs sinusoidal signal was observed at the beginning of the
PCD signal. The intense fluctuations of the ACD signal reveal
C. Data acquisition and analysis dramatic size expansion of the nucleated droplet, denoted as the
In PCD, the obtained signals were translated to the frequency gas nucleation signal. It was particularly interesting that another
domain by the fast Fourier transform. Because inertial sinusoidal signal was observed at about 5 μs after the first half
cavitation would produce a wideband spectral distribution signal. That signal coincided with the reduction of the ACD
among the fundamental and harmonic peaks, the inertial signal amplitude that indicated rebound of the transforming
cavitation dose could be estimated from the area under the bubble, as evident in high-speed microscopy images (Fig. 2C,
receiver operating characteristics curve within the -6 dB bottom), and it was denoted as the bubble rebound signal.
bandwidth of the 2.25-MHz transducer, which did not contain We then investigated the droplet-size-dependence properties
contributions from the fundamental and harmonic components. of the PCD signals for droplets with typical sizes ranging from
In ACD, the amplitude of the received signals was positively 3.0 to 6.0 μm (US parameters: 8 MPa, three cycles, single pulse).
correlated with the size of the ADV bubbles. Therefore, the time The time-domain signals demonstrated that increasing the size
interval between the highest signal amplitude (ADV-bubbles of the droplets did not change the amplitude of the gas
expansion) and lowest signal amplitude (ADV-bubbles nucleation signals (Fig. 3A). The spectral analyses revealed that
compression) could be estimated and then compared with the the gas nucleation signals spectra that strongly resembled the -
spectra dose of PCD. The relative changes in the signal 6 dB bandwidth of 2.25-MHz transducer for all sizes of droplets,
amplitude after ADV (referred to as the compression ratio of suggesting that this signal was a wideband signal (Fig. 3B). The
the rebound bubbles) was calculated as: quantified spectral doses were also consistent in these groups,
at 230.8±8.8, 234.0±8.5, 233.7±5.4, and 230.3±5.5 a.u. for
Compression ratio of rebound bubbles droplets with typical diameters of 3.0, 4.0, 5.0, and 6.0 μm,
𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑚𝑚𝑚𝑚𝑚𝑚 /𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑚𝑚𝑚𝑚𝑚𝑚
= respectively (Fig. 3C). No wideband signals were observed in
𝑇𝑇𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 − 𝑇𝑇𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸
PBS only group (50.3±3.8 a.u.) (Fig. 3b), suggesting that no
inertial cavitation event was involved in these observations.
where Expansion voltage max is the highest signal amplitude
(ADV-bubbles expansion), Compression voltage min is the
lowest signal amplitude (ADV-bubbles compression), B. The amplitudes of the ADV-bubble rebound signal is
T Compression is the delay time of bubble compression, and associated with droplet size
T Expansion is the delay time of bubble expansion. The amplitudes of the bubble rebound signals decreased as
the droplet size increased for 3.0-, 4.0-, and 5.0-μm droplets,
D. US M-mode imaging of ADV respectively. The amplitude of the bubble rebound signal for
6.0-μm droplets approached that of the PBS signal with a
For M-mode imaging, the droplet concentration was set at similar amount of noise. Furthermore, we noticed that
2×105 droplets/mL. Each M-mode image consisted of 100 RF enlargement of the droplets also delayed the onset time of the
signals. In order to reduce stationary signals, a high-pass bubble rebound signals, being 4.6, 5.0, 5.5, and 5.9 μs for 3.0-,
Butterworth filter (cutoff frequency: 100 Hz) was performed 4.0-, 5.0-, and 6.0-μm droplets, respectively. The distribution of
over a slow time index. A baseband demodulation was used to the frequency components of the bubble rebound signals
yield the envelope-detected images demonstrating the acoustic decreased as the droplets increased in size. The frequency
signatures of phase-change droplets during ADV and the response of the signals for 3.0-μm droplets spanned the spectral
signal-to-noise ratio (SNR) within regions of interest was also range from 0 to 3 MHz (Fig. 3d), whereas a lower frequency
calculated. ranges from 0 to 2 MHz appeared for 6.0-μm droplets. The
narrowed frequency distribution was also associated with a
E. Statistical analyses decreasing spectral dose of the signals, being 160.5±3.7,
All results are presented as mean and standard deviation (SD) 152.0±13.7, 144.0±23.1, and 91.8±26.1 a.u. for 3.0-, 4.0-, 5.0-,
values from at least three independent measurements. All and 6.0-μm droplets, respectively (Fig. 3e). Because this signal
appeared following the vaporization pulse had passed, it was

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 4

independent of the US pulse, but potentially was contributed to 5). This type of activity was observed for droplets of all sizes
by radial oscillations of droplet-formed bubbles. Besides, the and increased as the size of droplet decreased. For the 3.0-µm
extremely low intensity of spectral dose (45.5±4.0 a.u.) in PBS droplets, the contraction phase seemed to induce an additional
group suggested that oscillations of droplet-formed bubbles momentum in the formed bubble that resulted in it re-expanding
were not caused by the occurrence of inertial cavitation from from 4 to 6 μs and then returning to the final bubble size beyond
medium surrounding the droplet. 6 μs. These results were consistent with the data of previous
studies which demonstrated that the volumetric dampening of
C. ADV-bubble rebound signal was resulted from the
droplet during ADV as a function of droplet size [26]. Besides,
rebound of ADV bubble by ACD observation
these results also indicated that bubble rebound resulted from
Further investigation of the morphological evolution during the overexpansion of droplet ADV.
droplet ADV with the ACD setup helped to understand the
origin of the bubble rebound signals. Figure 4A demonstrates
the superimposition of ACD and PCD signals in the time E. Acoustic signatures of droplet vaporization-based US M-
domain for droplets with typical sizes of 3.0, 4.0, 5.0, and mode imaging
6.0 μm (US parameters: 8 MPa, three cycles, single pulse). Next, we acquired the acoustic signals of groups of
Upon US activation, the amplitude of ACD signals started droplets (concentration: 2×105 droplets/mL) to verify if the
increasing over time (from 0 to 3 μs) in each case, suggesting acoustic signals would be distorted by the simultaneous
that droplets were produced during the expansion phase. It was occurrence of multiple ADV events. We found that the gas
apparent that the ACD signals steadily decayed from 3 to 5 μs, nucleation signal was higher for groups of droplets than for a
indicating that the ADV bubbles were being compressed. The single droplet, probably due to the occurrence of signal
amplitude of the ACD signals subsequently increased again, superimposition in group-droplet ADV (Fig. 6A–D). In the time
suggesting re-expansion of the ADV bubbles, which was called domain, more than one rebound bubble signal could be
ADV-bubble rebound. It is notable that the rebound of ADV observed for groups of droplets, which is probably due to the
oscillations of ADV bubbles of multiple sizes.
bubbles appeared right after the vaporization pulse had passed
Finally, the acoustic signals of group-droplet ADV
and was independent of the pulse, suggesting that the rebound
(concentration: 2×105 droplets/mL) were collected to develop
phenomenon was a natural oscillation instead of representing
M-mode imaging. Figure 6E shows that the SNR was higher for
forced or driven oscillations. gas nucleation signals than for rebound bubble signals (16.6 dB
Combining with the PCD data, the time point at which the vs. 9.3 dB). Also, the appearance time of rebound bubble
second signals appeared strongly overlapped the transition of signals was highly inconsistent compared with the gas
the ADV-bubbles compression and re-expansion. The strong nucleation signals, suggesting that the latter is more suitable as
correlation between the PCD delay times (3.2±0.3, 3.6±0.2, an indicator for the occurrence of ADV.
4.0±0.3, and 4.3±0.2 μs for 3.0-, 4.0-, 5.0-, and 6.0-μm droplets,
respectively) and the ACD delay times (2.3±0.7, 2.7±0.4,
2.7±0.4, and 3.8±0.2 μs, respectively) also supported this IV. DISCUSSION
observation (R2 = 0.872) (Fig. 4B). Moreover, Fig. 4C This study successfully combined the detection of acoustic
demonstrates that the amplitude of the second signal detected
signals and high-speed optical imaging to determine the
by PCD (0.10±0.01, 0.08±0.02, 0.06±0.02, and 0.02±0.00 V for
physical behaviors of phase-change droplets during ADV. Our
3.0-, 4.0-, 5.0-, and 6.0-μm droplets, respectively) and the
results demonstrate a strategy for monitoring the transient
ADV-bubbles compression ratio (2.3±0.9, 1.6±0.1, 1.4±0.1,
physical dynamics of phase-change droplets based on the
and 0.9±0.2 μm–1, respectively) decreased with the droplet size
detection of acoustic signals. This strategy potentially provides
(R2 = 0.921). From these data it can be concluded that the ADV
benefits over existing techniques for use in medical applications.
bubbles underwent a cycle of oscillation and simultaneously
Compared to US B-mode imaging for detecting ADV events,
emitted acoustic signals. Therefore, we considered that the
which require a certain amount of bubbles to produce adequate
observed second acoustic signal for PCD resulted from the
contrast enhancement, our proposed approach can detect single-
rebound of ADV bubbles.
droplet ADV. Besides, the low frequency components of
acoustic signatures after droplet vaporization probably
D. Volumetric overexpansion and constriction of ADV- potentially can be used to deep tissue imaging and translated to
bubble observed from high-speed microscopy imaging larger animals and humans.
We also utilized high-speed microscopy imaging to visualize Although Fig. 6 demonstrated that the acoustic signatures of
the above-mentioned ADV bubble behaviors for 3.0-, 4.0-, 5.0- droplet vaporization could be as an indicator of droplet
μm, and 6.0-μm droplets under US sonication. The image vaporization, the current droplet-based sequences may tend to
sequence and analysis of the diameter–time curve revealed that have lower resolution than a CEUS or B-mode sequence. For
the droplets showed volumetric overexpansion from 0 to 2 μs instance, we found that a two-part droplet response playing out
to form a bubble, with the formed bubble constricting from 2 to over time that would either look like poor axial resolution or
4 μs to settle to a stable diameter that was smaller than the clutter artifacts below the vaporization region (Fig. 6e).
maximum size reached during the initial expansion phase (Fig. However, in the CEUS or B-mode sequence the axial resolution
will be dictated by the point spread function alone, whereas the

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 5

droplet sequence will likely stretch on for longer as the time However, a noninvasive imaging tool with good spatial and
response evolves. temporal resolutions capable of detecting the vaporization and
One previous published article observed that with shorter recondensation of phase-changing nanodroplets is still lacking.
ADV pulses, droplets would over-expand and undergo Given the properties of rebound bubble signals detected in this
unforced radial oscillation while settling to a final bubble study, we consider that rebound bubble signals can be used as
diameter [26]. The phenomenon of size dependent oscillation an acoustic signature of the recondensating droplets for the real-
generated a unique acoustic signal and the magnitude of the time detection of the physical behaviors of phase-change
acoustic signal increases as the boiling point of droplet core droplets in a biological environment.
decreasing. However, a confirmation between the amplitude of In this study, the smallest droplets for which phase changes
oscillation and the frequency of oscillation for a single droplet were observed had diameters of around 3.0 μm, since it is
may not be apparent even using highly dilute samples of difficult to precisely estimate the sizes of smaller particles using
droplets (lowest concentration of 0.05% in their study). optical observations. Excessive US sonication (i.e., high
To characterize the acoustic signals emitted from droplets acoustic pressures and long exposure times) will cause the
during ADV, we conducted this study under the setup of single- collapse or oscillation of ADV bubbles [17]. In the present
droplet ADV. Then, we revealed the correlations between ADV study, we determined the suitable US sonication pressure at the
acoustic signals and the different droplet sizes. Finally, we acoustic threshold for ADV and a pulse length of three cycles.
further used the characteristic signatures in groups of droplets Applying these optimized sonication parameters resulted in no
to develop a detection system based on ultrasound M-mode disruption of ADV bubbles being observed. Moreover, the
imaging. Besides, one previous study observed that the residual damping of ADV bubbles also was not affected by the US
oscillation after droplet vaporization but that did not appear in sonication.
our study. One of the reasons is potentially responsible for this Our data showed that increasing the diameter of the droplets
difference probably because the boiling point of from 3.0 to 6.0 μm did not increase the amplitude of gas
perfluorocarbon used in our study (PFP, 28 ℃) is largely higher nucleation signals, but it did decrease the amplitude of the
than that used in the referenced study (decafluorobutane, -2 ℃; rebound signals. The gas nucleation signal was generated from
octafluoropropane: -36.7 ℃) [26]. Such high boiling point of the rapid transition of PFP. Since these droplets were
PFP potentially produced less momentum on expansion and did insonicated at the same acoustic pressure (8 MPa), the
not produce residual oscillation after droplet vaporization. amplitudes of these signals should be similar. On the other hand,
Previous studies have developed the approach that could be the rebound signals resulted from the damped oscillation of
viewed as the acoustic signatures during droplet vaporization to ADV bubbles. The degree of the oscillations was negatively
generate high-sensitivity, high-contrast images of droplet dependent on the droplet diameter. Together these observations
vaporization events. As the droplet was vaporized, it indicate that the gas nucleation signals and rebound signals can
overexpanded and oscillated down to its final resting diameter, be used to detect the occurrence of droplet ADV and the
producing low frequency echo between 0.25 and 2.5 MHz formation of ADV bubbles, respectively.
regardless of the excitation frequency (5.0 or 8.0 MHz) [28, 29]. While we have demonstrated that the physical behavior of
In 2019, Rojas et al. demonstrated that this approach still could acoustic droplets during ADV can be characterized by
provide a high contrast-to-tissue ratio image in vivo [30]. measuring acoustic signals, three major limitations still remain
However, a long tail signal appeared in their received time in this study. First, the low frame rate of optical imaging.
domain signal suggested that the wideband signal during Further increases in the frame rate of optical imaging are needed
droplet vaporization and the un-forced ADV-bubble oscillation to assess the origin of acoustic signals in detail for developing
signal probably mixed together [28]. In this study, we ADV-specific detection techniques. Second, the current study
successfully separated these two signals via increasing the size was the absence of in vivo tests. The ADV emitted acoustic
of droplet (0.1 µm vs. 3.0-6.0 µm) because larger droplets signals might be distorted by heterogenous tissues within the
would generate larger microbubbles that oscillated at low body. Third, the detection setup is hard to present in vivo
frequencies when compared with small droplets. because it is required to separate two frequency detectors co-
ADV is receiving increasing attention due to its ability to aligned within body. Forth, the acoustic pressure used in this
produce microbubbles on demand in a noninvasive and study might induce inertial cavitation in a non-degassed
localized manner. A novel application is the utilization of laser- medium within the body. A previous study demonstrated that
activated nanometer-sized phase-change droplets undergoing the inertial cavitation threshold of 4.7-MHz plused US in
repeatedly vaporization and recondensation to their native human blood was higher than 6.3 MPa [33]. The next goal of
liquid nanodroplet state. This process can transiently produce this project will focus on: (1) combining with low-boiling-point
sonography-detectable microbubbles for time periods ranging phase-change contrast agents to reduce the acoustic energy of
from several milliseconds to hundreds of milliseconds, droplet vaporization [18]; (2) developing novel acoustic
potentially providing a new methodology for high-resolution droplets that can precise tunability in vaporization behavior; (3)
US molecular imaging in deep tissue [31]. Moreover, this integrating the transmitter with the detector into a single
repeated behavior also produces mechanical forces that can transducer; (4) combing above two strategies for in vivo droplet
disrupt the endothelial lining of the BBB, inducing BBB vaporization detection. Designing an algorithm and PCD setup
opening and allowing agents to be delivered to brain tissue [32]. for obtaining the ADV emitted acoustic signals inside body also

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 6

would be benefit to improve the in vivo application. One of the 34, no. 3, pp. 435-45, Mar 2008, doi:
10.1016/j.ultrasmedbio.2007.08.004.
challenges presented by ADV approaches is the ability to [6] M. Zhang et al., "Initial investigation of acoustic droplet
successfully generate vaporization in the deep tissue (>10 cm vaporization for occlusion in canine kidney," (in eng), Ultrasound
range) within human body when significant patient-to-patient Med Biol, vol. 36, no. 10, pp. 1691-703, Oct 2010, doi:
10.1016/j.ultrasmedbio.2010.06.020.
variability exists. Future attempts to reduce vaporization energy [7] M. Zhang et al., "Acoustic Droplet Vaporization for the
or develop novel droplet may address this issue. Recently, Enhancement of Ultrasound Thermal Therapy," (in eng), Proc IEEE
Shakya et al. proposed a tune-able droplet vaporization system Ultrason Symp, vol. 2010, pp. 221-224, Oct 11 2010, doi:
via encapsulating a fluorocarbon (FC) or hydrocarbon (HC) 10.1109/ultsym.2010.0054.
[8] C. C. Chen, P. S. Sheeran, S. Y. Wu, O. O. Olumolade, P. A. Dayton,
endoskeleton into C 5 F 12 droplets [34]. FC skeletons could and E. E. Konofagou, "Targeted drug delivery with focused
inhibit vaporization, whereas HC skeletons trigger vaporization ultrasound-induced blood-brain barrier opening using acoustically-
near the rotator melting transition, potentially precisely activated nanodroplets," (in eng), J Control Release, vol. 172, no. 3,
pp. 795-804, Dec 28 2013, doi: 10.1016/j.jconrel.2013.09.025.
controlling vaporization behavior of droplets. On the other hand, [9] Y. J. Ho, Y. C. Chang, and C. K. Yeh, "Improving Nanoparticle
several studies have demonstrated biomedical applications of Penetration in Tumors by Vascular Disruption with Acoustic
drug-loaded phase-change droplets, but they have lacked a real- Droplet Vaporization," (in eng), Theranostics, vol. 6, no. 3, pp. 392-
403, 2016, doi: 10.7150/thno.13727.
time method for assessing the released cargos. In the field of [10] W. T. Chen, S. T. Kang, J. L. Lin, C. H. Wang, R. C. Chen, and C.
microbubble-assisted drug delivery, the acoustic signals K. Yeh, "Targeted tumor theranostics using folate-conjugated and
produced by microbubbles could be utilized to estimate the camptothecin-loaded acoustic nanodroplets in a mouse xenograft
model," (in eng), Biomaterials, vol. 53, pp. 699-708, 2015, doi:
amount of chemodrug that is delivered. Therefore, future 10.1016/j.biomaterials.2015.02.122.
applications of our findings should be useful in classifying the [11] R. D. Airan et al., "Noninvasive Targeted Transcranial
extent of drug-loaded phase-change droplets. Neuromodulation via Focused Ultrasound Gated Drug Release from
Nanoemulsions," (in eng), Nano Lett, vol. 17, no. 2, pp. 652-659,
Feb 8 2017, doi: 10.1021/acs.nanolett.6b03517.
[12] A. Moncion et al., "Controlled release of basic fibroblast growth
V. CONCLUSION factor for angiogenesis using acoustically-responsive scaffolds," (in
eng), Biomaterials, vol. 140, pp. 26-36, Sep 2017, doi:
In this study we observed that a broadband shock wave 10.1016/j.biomaterials.2017.06.012.
appeared at the beginning of the PCD signal. The intense [13] D. Thakkar, R. Gupta, K. Monson, and N. Rapoport, "Effect of
ultrasound on the permeability of vascular wall to nano-emulsion
fluctuations of the ACD signal revealed that the shock wave droplets," (in eng), Ultrasound Med Biol, vol. 39, no. 10, pp. 1804-
arose from the inertial cavitation of the nucleated small gas 11, Oct 2013, doi: 10.1016/j.ultrasmedbio.2013.04.008.
pockets in the droplet. Another shock wave signal with a much [14] D. Qin et al., "In situ observation of single cell response to acoustic
lower acoustic frequency (<2 MHz) was observed at about 5 μs droplet vaporization: Membrane deformation, permeabilization, and
blebbing," (in eng), Ultrason Sonochem, vol. 47, pp. 141-150, Oct
after the first half signal. That signal coincided with the 2018, doi: 10.1016/j.ultsonch.2018.02.004.
reduction of the ACD signal amplitude that indicated the [15] A. H. Lo, O. D. Kripfgans, P. L. Carson, E. D. Rothman, and J. B.
rebound of the transforming bubble. Since internal gas Fowlkes, "Acoustic droplet vaporization threshold: effects of pulse
nucleation is a crucial process of ADV, the first half signal may duration and contrast agent," (in eng), IEEE Trans Ultrason
Ferroelectr Freq Control, vol. 54, no. 5, pp. 933-46, May 2007, doi:
indicate the occurrence of an ADV event, and the second half 10.1109/tuffc.2007.339.
signal may further reveal the degrees of expansion and [16] C. H. Wang, S. T. Kang, Y. H. Lee, Y. L. Luo, Y. F. Huang, and C.
oscillation of the bubble. These acoustic signatures provide K. Yeh, "Aptamer-conjugated and drug-loaded acoustic droplets for
opportunities for the real-time monitoring of ADV dynamics in ultrasound theranosis," (in eng), Biomaterials, vol. 33, no. 6, pp.
1939-47, Feb 2012, doi: 10.1016/j.biomaterials.2011.11.036.
vivo based on the detection of acoustic signals. [17] Y. J. Ho and C. K. Yeh, "Theranostic Performance of Acoustic
Nanodroplet Vaporization-Generated Bubbles in Tumor
Intertissue," (in eng), Theranostics, vol. 7, no. 6, pp. 1477-1488,
2017, doi: 10.7150/thno.19099.
REFERENCES [18] W. T. Shi, F. Forsberg, J. S. Raichlen, L. Needleman, and B. B.
[1] O. D. Kripfgans, J. B. Fowlkes, D. L. Miller, O. P. Eldevik, and P. Goldberg, "Pressure dependence of subharmonic signals from
L. Carson, "Acoustic droplet vaporization for therapeutic and contrast microbubbles," (in eng), Ultrasound Med Biol, vol. 25, no.
diagnostic applications," (in eng), Ultrasound Med Biol, vol. 26, no. 2, pp. 275-83, Feb 1999, doi: 10.1016/s0301-5629(98)00163-x.
7, pp. 1177-89, Sep 2000, doi: 10.1016/s0301-5629(00)00262-3. [19] Y. S. Tung, J. J. Choi, B. Baseri, and E. E. Konofagou, "Identifying
[2] N. Rapoport, Z. Gao, and A. Kennedy, "Multifunctional the inertial cavitation threshold and skull effects in a vessel phantom
nanoparticles for combining ultrasonic tumor imaging and targeted using focused ultrasound and microbubbles," (in eng), Ultrasound
chemotherapy," (in eng), J Natl Cancer Inst, vol. 99, no. 14, pp. Med Biol, vol. 36, no. 5, pp. 840-52, May 2010, doi:
1095-106, Jul 18 2007, doi: 10.1093/jnci/djm043. 10.1016/j.ultrasmedbio.2010.02.009.
[3] S. T. Kang, Y. L. Huang, and C. K. Yeh, "Characterization of [20] L. Deng, M. A. O'Reilly, R. M. Jones, R. An, and K. Hynynen, "A
acoustic droplet vaporization for control of bubble generation under multi-frequency sparse hemispherical ultrasound phased array for
flow conditions," (in eng), Ultrasound Med Biol, vol. 40, no. 3, pp. microbubble-mediated transcranial therapy and simultaneous
551-61, Mar 2014, doi: 10.1016/j.ultrasmedbio.2013.10.020. cavitation mapping," (in eng), Phys Med Biol, vol. 61, no. 24, pp.
[4] C. M. Carneal, O. D. Kripfgans, J. Krücker, P. L. Carson, and J. B. 8476-8501, Dec 21 2016, doi: 10.1088/0031-9155/61/24/8476.
Fowlkes, "A tissue-mimicking ultrasound test object using droplet [21] T. Sun et al., "Closed-loop control of targeted ultrasound drug
vaporization to create point targets," (in eng), IEEE Trans Ultrason delivery across the blood-brain/tumor barriers in a rat glioma
Ferroelectr Freq Control, vol. 58, no. 9, pp. 2013-25, Sep 2011, doi: model," (in eng), Proc Natl Acad Sci U S A, vol. 114, no. 48, pp.
10.1109/tuffc.2011.2045. E10281-e10290, Nov 28 2017, doi: 10.1073/pnas.1713328114.
[5] K. J. Haworth, J. B. Fowlkes, P. L. Carson, and O. D. Kripfgans, [22] R. Asami, T. Ikeda, T. Azuma, S. Umemura, and K.-i. Kawabata,
"Towards aberration correction of transcranial ultrasound using "Acoustic signal characterization of phase change nanodroplets in
acoustic droplet vaporization," (in eng), Ultrasound Med Biol, vol. tissue-mimicking phantom gels," Japanese Journal of Applied
Physics, vol. 49, no. 7S, p. 07HF16, 2010.

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 7

[23] M. L. Fabiilli, K. J. Haworth, N. H. Fakhri, O. D. Kripfgans, P. L. [30] J. D. Rojas and P. A. Dayton, "Vaporization Detection Imaging: A
Carson, and J. B. Fowlkes, "The role of inertial cavitation in acoustic Technique for Imaging Low-Boiling-Point Phase-Change Contrast
droplet vaporization," (in eng), IEEE Trans Ultrason Ferroelectr Agents with a High Depth of Penetration and Contrast-to-Tissue
Freq Control, vol. 56, no. 5, pp. 1006-17, May 2009, doi: Ratio," (in eng), Ultrasound Med Biol, vol. 45, no. 1, pp. 192-207,
10.1109/tuffc.2009.1132. Jan 2019, doi: 10.1016/j.ultrasmedbio.2018.08.017.
[24] N. Reznik, R. Williams, and P. N. Burns, "Investigation of [31] G. P. Luke, A. S. Hannah, and S. Y. Emelianov, "Super-Resolution
vaporized submicron perfluorocarbon droplets as an ultrasound Ultrasound Imaging in Vivo with Transient Laser-Activated
contrast agent," (in eng), Ultrasound Med Biol, vol. 37, no. 8, pp. Nanodroplets," (in eng), Nano Lett, vol. 16, no. 4, pp. 2556-9, Apr
1271-9, Aug 2011, doi: 10.1016/j.ultrasmedbio.2011.05.001. 13 2016, doi: 10.1021/acs.nanolett.6b00108.
[25] N. Rapoport, D. A. Christensen, A. M. Kennedy, and K. H. Nam, [32] K. A. Hallam, E. M. Donnelly, A. B. Karpiouk, R. K. Hartman, and
"Cavitation properties of block copolymer stabilized phase-shift S. Y. Emelianov, "Laser-activated perfluorocarbon nanodroplets: a
nanoemulsions used as drug carriers," (in eng), Ultrasound Med Biol, new tool for blood brain barrier opening," (in eng), Biomed Opt
vol. 36, no. 3, pp. 419-29, Mar 2010, doi: Express, vol. 9, no. 9, pp. 4527-4538, Sep 1 2018, doi:
10.1016/j.ultrasmedbio.2009.11.009. 10.1364/boe.9.004527.
[26] P. S. Sheeran, T. O. Matsunaga, and P. A. Dayton, "Phase change [33] C. X. Deng, Q. Xu, R. E. Apfel, and C. K. Holland, "In vitro
events of volatile liquid perfluorocarbon contrast agents produce measurements of inertial cavitation thresholds in human blood," (in
unique acoustic signatures," (in eng), Phys Med Biol, vol. 59, no. 2, eng), Ultrasound Med Biol, vol. 22, no. 7, pp. 939-48, 1996, doi:
pp. 379-401, Jan 20 2014, doi: 10.1088/0031-9155/59/2/379. 10.1016/0301-5629(96)00104-4.
[27] C. H. Fan, Y. T. Lin, Y. J. Ho, and C. K. Yeh, "Spatial-Temporal [34] G. Shakya, S. E. Hoff, S. Wang, H. Heinz, X. Ding, and M. A.
Cellular Bioeffects from Acoustic Droplet Vaporization," (in eng), Borden, "Vaporizable endoskeletal droplets via tunable interfacial
Theranostics, vol. 8, no. 20, pp. 5731-5743, 2018, doi: melting transitions," (in eng), Sci Adv, vol. 6, no. 14, p. eaaz7188,
10.7150/thno.28782. Apr 2020, doi: 10.1126/sciadv.aaz7188.
[28] C. B. Arena, A. Novell, P. S. Sheeran, C. Puett, L. C. Moyer, and P.
A. Dayton, "Dual-frequency acoustic droplet vaporization detection
for medical imaging," (in eng), IEEE Trans Ultrason Ferroelectr
Freq Control, vol. 62, no. 9, pp. 1623-33, Sep 2015, doi:
10.1109/tuffc.2014.006883.
[29] A. Novell, C. B. Arena, O. Oralkan, and P. A. Dayton, "Wideband
acoustic activation and detection of droplet vaporization events
using a capacitive micromachined ultrasonic transducer," (in eng), J
Acoust Soc Am, vol. 139, no. 6, p. 3193, Jun 2016, doi:
10.1121/1.4953580.

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 8

List of Table and Figure Captions

TABLE Ⅰ. Transducer characteristics.

Fig. 1. Size distribution and concentration of the phase-change droplets used in this study. Inset: bright-field microscopy image of
the phase-change droplets.

Fig. 2. (a) Experimental setup of passive cavitation detection (PCD). (b) Experimental setup of active cavitation detection (ACD).
(c) Typical examples of a PCD signal, an ACD signal, and high-speed bright field images for a 5-μm droplet exposed to a three-
cycle ultrasound (US) signal at 8 MPa.

Fig. 3. (a) Individual voltage traces for the signals received from 3.0-, 4.0-, 5.0-, and 6.0-μm droplets produced by acoustic droplet
vaporization (ADV) at a US pressure of 8.0 MPa. (b) Spectral analysis and the (c) corresponding spectral dose for the first half of
the PCD signal (N=6). (d) Spectral analysis and the (e) corresponding spectral dose for the second half of the PCD signal (N=6).
*, p<0.05. PBS, phosphate-buffered saline. Data are mean and SD values.

Fig. 4. (a) Superimposition of voltage traces of droplet ADV received for PCD and ACD with different parameters. (b) Comparison
of delay times measured for PCD and ACD (N=6). (c) Comparison of voltage amplitudes and rebound bubble compression ratios
measured for PCD and ACD (N=6). *, p<0.05. Data are mean and SD values.

Fig. 5. (a) High-speed microscopy images (500,000 fps) and (b) changes in the diameters of droplets acquired during ADV for
typical droplet sizes of 3.0 and 6.0 μm (US parameters: 8 MPa, three cycles, single pulse) (N=6).

Fig. 6. (a) Individual voltage traces, (b) time–frequency analysis, (c) spectra of gas nucleation signals, and (d) spectra of rebound
bubble signals received from PCD for single droplets and groups of droplets. (e) M-mode images produced via droplet ADV-
emitted acoustic signals of groups of droplets. Red dot rectangular: region of gas nucleation signals. Blue dot rectangular: region
of rebound bubble signals.

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 9

TABLE Ⅰ
TRANSDUCER CHARACTERISTICS.
Transducer model Sonic Concepts SU- Olympus NDT V305 Olympus NDT V324
128
Center frequency (MHz) 5 2.25 25
Focal length (mm) 2.3 5.6 1.3
Focal width (mm) 0.3 1.9 0.2
Focal depth (mm) 21.6 35.2 13.2
Outer diameter (mm) 25.4 19.1 6.4
–6 dB bandwidth (%) 53 75.6 45.1

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 10

Fig. 1. Size distribution and concentration of the phase-change droplets used in this study. Inset: bright-field microscopy image
of the phase-change droplets.

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 11

Fig. 2. (a) Experimental setup of passive cavitation detection (PCD). (b) Experimental setup of active cavitation detection (ACD).
(c) Typical examples of a PCD signal, an ACD signal, and high-speed bright field images for a 5-μm droplet exposed to a three-
cycle ultrasound (US) signal at 8 MPa.

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 12

Fig. 3. (a) Individual voltage traces for the signals received from 3.0-, 4.0-, 5.0-, and 6.0-μm droplets produced by acoustic
droplet vaporization (ADV) at a US pressure of 8.0 MPa. (b) Spectral analysis and the (c) corresponding spectral dose for the
first half of the PCD signal (N=6). (d) Spectral analysis and the (E) corresponding spectral dose for the second half of the PCD
signal (N=6). *, p<0.05. PBS, phosphate-buffered saline. Data are mean and SD values.

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 13

Fig. 4. (a) Superimposition of voltage traces of droplet ADV received for PCD and ACD with different parameters. (b) Comparison
of delay times measured for PCD and ACD (N=6). (c) Comparison of voltage amplitudes and rebound bubble compression ratios
measured for PCD and ACD (N=6). *, p<0.05. Data are mean and SD values.

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 14

Fig. 5. (a) High-speed microscopy images (500,000 fps) and (b) changes in the diameters of droplets acquired during ADV for
typical droplet sizes of 3.0 and 6.0 μm (US parameters: 8 MPa, three cycles, single pulse) (N=6).

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TUFFC.2020.3032441, IEEE
Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 15

Fig. 6. (a) Individual voltage traces, (b) time–frequency analysis, (c) spectra of gas nucleation signals, and (d) spectra of rebound
bubble signals received from PCD for single droplets and groups of droplets. (e) M-mode images produced via droplet ADV-
emitted acoustic signals of groups of droplets. Red dot rectangular: region of gas nucleation signals. Blue dot rectangular: region
of rebound bubble signals.

0885-3010 (c) 2020 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
Authorized licensed use limited to: Auckland University of Technology. Downloaded on November 08,2020 at 02:30:19 UTC from IEEE Xplore. Restrictions apply.

You might also like