IN
DEGREE PROJECT MEDICAL ENGINEERING, SECOND CYCLE, 30 CREDITS
STOCKHOLM SWEDEN 2020,
Acoustic Droplet Vaporization
An Assessment of How Ultrasound Wave
Parameters Influence the Vaporization Efficiency
Utvärdering av hur ultraljudsparametrar påverkar effektiviteten av akustisk vaporisering av
vätskedroppar SARA ÖQUIST
KTH ROYAL INSTITUTE OF TECHNOLOGY
SCHOOL OF ENGINEERING SCIENCES IN CHEMISTRY, BIOTECHNOLOGY AND HEALTH
i
Abstract
Acoustic droplet vaporization (ADV) is a process in which a phase shift of a liquified droplet into a gaseous microbubble, is triggered using an ultrasonic wave. In contrast to utilizing conventional contrast agents in ultrasound, the phase change contrast agents used in ADV can extravasate into tumor tissue, and they offer a greater circulatory lifespan, thereby increasing the potential applications in which they can be utilized.
In this project, the impact of different ultrasound parameters on the efficiency of ADV was investigated, using a programmable ultrasound system. Two different ultrasound sequences were designed, for imaging and vaporization of droplets. Furthermore, three different sets of experiments were performed. Firstly, the vaporization effect of different imaging voltages was investigated, whereby a setting of 15V was identified as an able voltage for the remaining experiments. Secondly, experiments concerning the effect of vaporizing frequency on the ADV efficiency were performed, including the use of single and dual frequencies. Lastly, different frequency settings were combined with varying the number of cycles, to assess how the choice of pulse length influences the vaporization.
The results from the project indicate that no substantial difference in ADV efficiency is achieved when using different frequency settings for perfluoropentane droplets encapsulated by cellulose nanofibers. However, the results provide clear indications of the benefit of using longer pulse durations on the vaporization efficiency. In conclusion, further studies are required before ADV can be translated into a clinical setting.
ii
iii
Acknowledgements
I would like to express my gratitude towards my supervisor, Dmitry Grishenkov, Assosciate Professor at the Department of Biomedical Engineering and Health systems at KTH, for his valuable time and input throughout the project. Moreover, I would like to thank the doctoral students in the CEMIT group, Ksenia Loskutova, for providing her help in preparing
droplets and characterizing these, and Hongjian Chen, for helping me when I struggled with the Verasonics system. Lastly, special thanks to Stina Hägglund, for accompanying me in the laboratory and acting as a sounding board when I have struggled in the project.
Sara Öquist May 2020
iv
v
Table of Contents
1. Introduction ... 1
1.1 Objectives ...
1
2. Methods ... 3
2.1 Materials ...
3
2.1.1 Droplets ...
3
2.1.2 Verasonics ...
4
2.1.3 Ultrasound Transducer ...
6
2.1.4 Phantom ...
6
2.2 Experimental Setup ...
6
2.3 Design of Ultrasound Sequences ...
7
2.3.1 Ultrasound Imaging ...
7
2.3.2 Vaporization Sequence ...
9
2.4 Experiments ...
9
2.5 Evaluating results ...
10
3. Results ... 12
3.1 Assessment of voltage for imaging ...
12
3.2 Efficiency of Vaporization ...
14
3.2.1 First Set of Experiments: Frequency ...
14
3.2.2 Second Set of Experiments: Frequency and Pulse Duration ...
15
4. Discussion ... 17
5. Conclusion ... 19
6. References ... 21
Appendix
Appendix A: State of the Art
Appendix B: Implementation of Ultrasound Sequences in Verasonics
vi
1
1. Introduction
For decades, the high echogenicity of microbubbles has been exploited to increase the contrast in ultrasound imaging for diagnostic purposes [1]. Furthermore, the behavior of microbubbles upon ultrasound exposure has been utilized in therapeutic applications. For example, microbubbles can be loaded with therapeutic drugs that are released when focused ultrasound exposure causes the microbubble to cavitate, enabling targeted drug delivery [2].
However, microbubbles are confined to intravascular applications, as their relatively large size prevents extravasation [1]. The utilization of microbubbles in diagnostic and therapeutic applications is further limited by diffusion of the encapsulated gas into the blood, resulting in a short circulation lifetime of the microbubbles [2].
To overcome the limitations of microbubbles, alternative contrast agents for ultrasonic applications have been investigated [3], whereby phase-change contrast agents (PCCAs) has been proposed as a promising alternative [4]. PCCAs consist of a liquid core that can
undergo a phase transition into a gas phase upon ultrasound exposure [3]. The phenomenon of the acoustically triggered phase shift of a droplet with a liquid core into a gaseous microbubble, is known as acoustic droplet vaporization (ADV) [5]. By utilizing ADV, it is possible to exploit the smaller size and greater stability of liquified droplets, e.g., allowing extravasation into tumor tissue, while efficiently achieve contrast enhancement offered by the gaseous microbubbles [1].
ADV shows great potential in both diagnostic and therapeutic applications [5]. For example, ADV can be utilized to achieve vessel occlusion for cancer treatment by triggering a phase transition of droplets localized in a feeder vessel to a tumor [1]. The formed microbubbles occlude small vessels, diminishing the blood flow to the tumor, a technique known as gas embolotherapy [1]. However, despite the potential of ADV in a clinical setting, there are several problems that have impeded the clinical translation of the phenomenon [2]. One such problem concerns the poor understanding of the ADV mechanism and inadequate knowledge of how to achieve efficient vaporization without harming surrounding tissues [1]. It has been reported that the pressure amplitudes required for successful vaporization, greatly exceeds the mechanical indices approved by the FDA for diagnostic ultrasound [6], whereby further knowledge of how to improve the ADV efficiency while lowering the pressure amplitudes, is required [7].
1.1 Objectives
In this thesis project, the main objective is to investigate how the use of different frequencies and pulse durations, influence the efficiency of ADV of perfluorocarbon droplets, using a programmable ultrasound system. To achieve this goal, three sub-objectives have been formulated. Firstly, to develop an ultrasonic imaging sequence that allows imaging of the droplets without vaporizing them. Secondly, to develop a vaporization sequence that achieves ADV. Lastly, to utilize the vaporization sequence, with different settings of
2 frequency and pulse durations, combined with the imaging sequence, to test how the
vaporization of droplets is influenced by these parameters.
3
2. Methods
This chapter will start with a description of the main materials and equipment used when conducting this project, including a description of the experimental set-up. Furthermore, the workflow of designing the ultrasound sequences, both for imaging and vaporization, will be described. Lastly, the experimental procedure of acquiring data will be explained, including information regarding how the data was analyzed and quantified.
2.1 Materials
2.1.1 Droplets
The PCCAs utilized in this project consist of a perfluoropentane core encapsulated by cellulose nanofibers (CNF) for stabilization purposes. The droplets were produced using the method of mechanical agitation, in which PCCAs are formed by mixing the materials of the core and shell materials [8]. This method is common to use in fabrication of PCCAs, as it is relatively fast and cheap [8]. For one of the samples, 0.7g of perfluoropentane was added to a container with 25g CNF suspension, containing 0.375% of CNF in Milli-Q water. The container was shaken by hand to stabilize perfluoropentane, which due to a low boiling point of 29.2℃ [5], is unstable at room temperature. An ultrasonic liquid processor (Vibracell W750, Sonics, USA) was utilized to further mix the sample. The mixing using the liquid processor, lasted for one minute with an amplitude of 80%. During this time, the sample container was placed in a bucket of ice to prevent heating from the extensive shaking of the equipment. The same procedure was used to produce several containers of perfluoropentane droplets.
Figure 1: Microscopic image of diluted droplet sample with pattern provided by the Neubauer chamber.
4 Upon fabrication, the droplet characteristics for one of the samples were evaluated, whereby the size distribution and the concentration was determined using images of the sample captured with an optical microscope. The equipment used for this was a Nikon Ni-E microscope with a Plan Apochromat Lambda 20x objective. A dilution of droplets and Milli-Q water with a 1:100 ratio was prepared for acquiring the microscopic images and was placed in an improved Neubauer chamber. One of the obtained images is provided in Figure 1. Upon acquiring several microscopic images of the droplets, the images were analyzed in ImageJ. The size indications provided by the different patterns of the Neubauer chamber, visible in Figure 1, was used for calibration. The resulting size distribution is shown in Figure 2. The mean value of the distribution was 2.67± 0.6µm.
The concentration was determined for three of the prepared samples, with results of 6.60∙108 droplets/ml, 4.21∙108 droplets/ml and 4.60∙108 droplets/ml, respectively. Since the exact concentration is not of significant importance for this project, it is assumed that the concentration for the samples used in this project is approximately 5∙108 droplets/ml, as deciding the concentration for each of the samples would be highly time-consuming. The concentration was decided based on the average of the determined concentrations.
2.1.2 Verasonics
In this project, the Verasonics V-1 system (Verasonics Inc., Kirkland, WA, USA) was used to develop and run the ultrasound sequences. The Verasonics System exposes essentially all aspects of an ultrasound system to the user, thereby making it possible to design customized sequences for transmitting, receiving, and processing ultrasound [9]. There are two main components of the Verasonics V-1 system: the Verasonics Data Acquisition system (VDAS),
Figure 2: Size distribution of CNF-coated PFP droplets.
5 which constitutes the hardware of the system, and the host computer, which contains the software environment. The system is programmed using Matlab (version R2013b, MathWorks Inc., Natick, MA, USA) on the host computer.
When developing a Matlab script for the Verasonics system, the user is required to provide the system with information regarding how the data acquisition and processing should be performed. When executing the script, a container of this information is created, called the Sequence object [10].
As illustrated in Figure 3, one of the main components of the sequence object is the event list, specifying what actions should be taken by the system and in what order. However, for the system to know how to perform those actions, additional information is required.
Thereby, the event list is complemented by several event objects, which also are illustrated
Figure 3: The Verasonics Sequence Object [10].
6 in Figure 3. For instance, transmit objects provide the system with information regarding what waveform should be used, which transducer elements should be activated, and if the elements should be activated simultaneously for a plane wave, or with delays for a focused beam.
2.1.3 Ultrasound Transducer
To transmit and receive the specified ultrasound waveform via the VDAS, it is required to connect an ultrasound transducer to the system. In this project, the Philips ATL L11-5 linear array transducer was used. The transducer consists of 128 elements and has a frequency range of 5-11MHz. Moreover, the transducer has a field-of-view of 34mm.
In this project, the ultrasound transducer was utilized for both transmitting and receiving ultrasound of the developed imaging sequences. Moreover, the transducer was used for the transmission of vaporization sequences, thereby, simultaneous imaging and vaporization was unattainable.
2.1.4 Phantom
To enable visualization of droplets inside a vessel, a tissue mimicking phantom was used.
The phantom, a Peripheral Vascular Doppler Flow Phantom of Model 524 from ATS Laboratories, consists of four different circular vessels with different diameters, 2, 4, 6 and 8mm, respectively. The vessels are located on a depth of 15mm below the scan surface [11].
The speed of sound in the phantom is 1450m/s, which was specified in the Matlab script. In the performed experiments, the largest vessel with a diameter of 8mm was used.
2.2 Experimental Setup
The equipment and materials described in the previous section, compose the main components in the complete experimental set-up in this project, illustrated in Figure 4.
The additional equipment used (nr. 1, 2, 5 and 6 in Figure 4) is briefly described below.
- Beaker containing the droplet solution. Diluted differently for different experiments, see section 2.3 and 2.4.
- Magnetic stirrer to assure a homogenous mixture of droplets during the experiments. The liquid droplets tend to descend to the bottom of the beaker, whereby mixing is required.
- Peristaltic pump and tubing to create a flow of droplet solution through the tissue mimicking phantom.
- Equipment for fixation of the ultrasound transducer to keep the transducer alignment constant through a series of experiments.
7
2.3 Design of Ultrasound Sequences
This section provides an explanation of the workflow of designing the ultrasound sequences for imaging and vaporization used in the final experiments.
2.3.1 Ultrasound Imaging
Pulse Inversion
For the imaging of droplets and potential formed microbubbles upon ADV, a harmonic imaging technique called pulse inversion was implemented. Pulse inversion exploits the nonlinear behavior of microbubbles, which causes generation of ultrasound echoes with harmonic distortion [12]. The nonlinear behavior of microbubbles is described in more detail in Section 1.4 in Appendix A. When utilizing pulse inversion, two consecutive ultrasound pulses are transmitted, the second pulse being phase inverted 180° in relation to the first. The echoes from both pulses are averaged, whereby the echoes from non-distorting structures will cancel each other, illustrated in Figure 5. Meanwhile, the nonlinear behavior of e.g.
microbubbles will result in a signal containing the harmonic frequencies of the echo, Figure 5 [12].
Figure 4: Experimental set-up composed of 1) Magnetic stirrer, 2) Beaker with droplet solution, 3) Ultrasound transducer, 4) Tissue mimicking phantom, 5) Fixation equipment for transducer, 6) Peristaltic pump
1 .
3
4
6 5
2
.
8 Imaging Sequence in Verasonics
The imaging sequence developed in this project contains functionality for imaging droplets using a focused beam and pulse inversion. Moreover, the sequence includes functionality for saving images that can be used for analysis. The imaging sequence utilizes a transmission frequency of 5.63MHz, corresponding to a second harmonic of 11.26MHz. Thereby, both frequencies are on one end of the bandwidth of the transducer, respectively. The pulse length of the ultrasound wave was set to one cycle. More detailed information regarding how the imaging sequence was implemented in Verasonics is provided in Section 1 in Appendix B.
Assessment of Voltage for Imaging
A series of experiments was performed to identify an adequate voltage for imaging the droplets. A suiting voltage for the imaging sequence should not cause any vaporization of the droplets, i.e. no visual changes in the vessel should occur and the mean pixel intensity should be constant over time.
The first step of the experiment consisted of preparing a sample of diluted droplets. The droplets were diluted with a dilution ratio of 1:50, resulting in a concentration of about 107 droplets/ml. Thereafter, the experimental set-up was arranged as described in Section 2.2.
Experiments were performed for voltage settings of 10V, 15V, 20V, 30V, 40V and 50V, consecutively. For each voltage, the Verasonics program was executed and the peristaltic pump was stopped, upon which 10 images were captured at regular time intervals over a
Figure 5: Principle of Pulse Inversion. The left-hand side correspond to non-distorted ultrasound echoes, whereas the right-hand side corresponds to echoes from distorting structures, e.g.
microbubbles with a nonlinear behavior. Reprinted from [12], © [2015] with permission from John Wiley and Sons.
9 total period of 30 seconds. After each experimental run, the pump was started again to allow for replenishment of droplets. The images were then evaluated according to Section 2.4.
2.3.2 Vaporization Sequence
The vaporization sequence includes functionality for transmitting high-power ultrasound transmission to vaporize the droplets. Furthermore, the vaporization sequence allows using two different frequencies for ultrasound transmission. Due to limitations in Verasonics, the use of two frequencies requires two separate transmissions. Therefore, the vaporization sequence consists of two consecutive transmissions from odd-numbered transducer elements and even-numbered elements, respectively. Additional functionality of the vaporization sequence includes saving ultrasound images, both triggered independently and after
vaporization. More detailed information concerning the implementation of the vaporization sequence in Verasonics is provided in Section 2 in Appendix B.
2.4 Experiments
For investigating how the vaporization efficiency is influenced by the choice of frequency and pulse duration, two sets of experiments were performed. The first set of experiments aimed at investigating how different frequency settings affects the vaporization. The objective of the second set of experiments was to investigate how the combination of pulse duration and frequency settings affects the vaporization. For each set of experiments, several experimental runs were performed, each using different waveform specifications for the vaporization sequence. The waveform specifications used in each set of experiments is illustrated in Table 1. All experiments used 15V as a setting for the imaging sequence, based on the results presented in section 3.1.
Each experiment used the experimental set-up described in Section 2.2, and a droplet solution diluted with a dilution ratio of 1:10, resulting in a concentration of about 5∙107 droplets/ml. One experimental run consisted of several steps. First, the waveform was updated in Matlab according to Table 1, and the Verasonics program was executed. Second, the peristaltic pump was stopped, and twenty images were acquired and saved. Thereafter, the vaporization was triggered, including subsequent saving of ten images. The vaporization was initiated twenty times, after which the experimental run was completed by a final acquisition of twenty images. In a total, each run of the experiment resulted in a total of 240 images.
10
Table 1: Waveform settings used for 1st and 2nd set of experiments
1
stset of experiments
Index Type (single or dual frequency)
Frequency Pulse duration [number of cycles]
1 Single 5MHz 32
2 Single 10MHz 32
3 Dual 5MHz+10MHz 32
4 Dual 5MHz+8MHz 32
5 Dual 8MHz+10MHz 32
2
ndset of experiments
Index Type (single or dual frequency)
Frequency Pulse duration [number of cycles]
1 Single 5MHz 32
2 Single 5MHz 320
3 Single 5MHz 640
4 Dual 5MHz+8MHz 32
5 Dual 5MHz+8MHz 320
6 Dual 5MHz+8MHz 640
7 Single 8MHz 32
8 Single 8MHz 320
9 Single 8MHz 640
2.5 Evaluating results
As previously described, all performed experiments generated a set of images saved in specified directories on the host computer. The images were evaluated using a separate
Figure 6: ROI in vessel with droplet solution.
11 Matlab-script with functionality for calculating the mean pixel intensity in a selected region of interest (ROI) and visualize the result in a Matlab Figure. The position of the selected ROI is illustrated in Figure 6.
For the series of experiments described in Section 2.4, the data was also analyzed
quantitatively. For each collection of images, percental changes of the mean pixel intensity was calculated. First, an average of the first twenty images was calculated, v1. Second, four additional averages, v2-v5 were calculated, the first three corresponding to the images acquired after the fifth, tenth and fifteenth vaporization, respectively. The last average value corresponded to the twenty images acquired after the last vaporization. Finally, the percental change, xi, of the mean pixel intensity was calculated according to equation 1.
𝑥𝑖 = 𝑣𝑖
𝑣1− 1 𝑖 = 2,3,4,5 (1)
12
3. Results
In this chapter, the results from the different experiments performed in this project are presented. The results are presented in chronological order, starting with the results from the experiments regarding assessment of what voltage is suitable for imaging, as these results were a requirement before continuing any further experiments with vaporization.
3.1 Assessment of voltage for imaging
The results regarding assessment of a suitable voltage for the imaging sequence is presented in Figure 7.
To easier distinguish differences in intensity between different voltage settings, the data was normalized by dividing all values for one voltage setting with the maximum value for that setting. The normalized data is presented in Figure 8.
Figure 7: Results from experiments to find adequate voltage for the imaging sequence.
13 Figure 8 shows that the mean pixel intensity for voltages of 10V and 15V is relatively
constant, whereas the remaining voltage settings show clear intensity variations. The result was supported by reviewing the captured images. For example, the images presented in Figure 9, show the difference in droplet presence for two consecutive images at 40V, illustrating how the droplets are affected when exposed to this voltage for a period of three seconds.
Furthermore, the increasing behavior of the intensity, spotted in the graphs for 20V and 30V in Figure 8, origins from the appearance of microbubbles, as microbubbles are more
Figure 8: Normalized mean pixel intensities from experiments to find adequate voltage for the imaging sequence.
Figure 9: The disappearance of droplets caused by continuous exposure to the imaging sequence for three seconds at 40V. A, represents image index 1. B, represents image index 2.
14 echogenic in comparison to droplets. The formation of a microbubble using 30V is
visualized in Figure 10.
3.2 Efficiency of Vaporization
3.2.1 First Set of Experiments: Frequency
The normalized results from the first set of experiments, regarding how different frequency settings affect the vaporization efficiency, is shown in Figure 11.
Figure 10: Appearance of a microbubble upon exposure to imaging sequence at 30V. A, represents image index 5. B, represents image index 6.
Figure 11: Normalized results obtained from the first set of experiments regarding frequency settings.
15 As appearing from Figure 11, all settings of frequency results in a distinct change of mean pixel intensity. Moreover, a significant change in intensity can be spotted at indices corresponding to the first vaporization event. However, due to the large number of data points and overlapping graphs, it is difficult to draw a conclusion of how the frequency affects the vaporization efficiency based on Figure 11 solely.
As described in section 2.5, the data from the experiments were also evaluated quantitively based on percental changes in calculated average values, according to equation 1. The results from this quantification is presented in Table 2. Table 2 shows that the total percentual change in mean pixel intensity varies between 28.7% and 33.3%, where a dual frequency of 5MHz and 10MHz corresponds to the lowest value, whereas a single frequency of 5MHz corresponds to the highest value. Among the dual frequencies, values of 8MHz and 10MHz resulted in the largest percentual differences.
Table 2: Quantified results from first set of experiments.
Frequency v1 v2 x2 v3 x3 v4 x4 v5 x5
5MHz 60.16 47.23 -21.7% 43.22 -28.1% 42.53 -29.3% 40.10 -33.3%
10MHz 60.18 48.83 -18.9% 45.27 -24.8% 45.62 -24.2% 41.76 -30.6%
5MHz+10MHz 58.99 49.67 -15.8% 45.19 -23.4% 45.55 -22.8% 42.03 -28.7%
5MHz+8MHz 57.29 45.04 -21.4% 42.19 -26.3% 40.35 -29.6% 39.75 -30.6%
8MHz+10MHz 55.59 47.91 -13.8% 45.23 -18.6% 41.43 -25.5% 37.54 -32.5%
3.2.2 Second Set of Experiments: Frequency and Pulse Duration
The results from the second set of experiments regarding frequency and pulse duration, was plotted in three different plots for better visualization. The normalized result is presented in Figure 12.
Figure 12: Results obtained from the second set of experiments regarding frequency settings and pulse durations
16 For all settings, a sharp decline in intensity after the first vaporization can be noted.
Moreover, there are some variations in the initial intensity. For better evaluation of the efficiency using the different settings, the data was also evaluated quantitively, according to the description provided in Section 2.5. The result of the data quantification is presented in Table 3.
Table 3: Quantified results from second set of experiments.
Frequency Number
of cycles v1 v2 x2 v3 x3 v4 x4 v5 x5
5MHz 32 54.98 47.49 -13.6% 43.73 -20.4% 40.88 -25.6% 40.10 -27.1%
5MHz 320 56.39 46.65 -17.3% 45.61 -19.1% 40.75 -27.7% 39.88 -29.3%
5MHz 640 57.29 44.99 -21.5% 42.57 -25.7% 40.73 -28.9% 37.78 -34.1%
5MHz+8MHz 32 53.48 44.95 -15.9% 40.76 -23.8% 37.81 -29.3% 36.92 -30.9%
5MHz+8MHz 320 53.98 44.91 -16.8% 39.25 -27.3% 38.20 -29.2% 37.78 -30.0%
5MHz+8MHz 640 56.19 44.68 -20.5% 41.58 -26.0% 38.55 -31.38 37.02 -34.1%
8MHz 32 57.63 48.52 -15.8% 42.17 -26.8% 41.31 -28.3% 41.37 -28.2%
8MHz 320 51.34 44.48 -13.3% 40.35 -21.4% 38.94 -24.1% 36.37 -29.1%
8MHz 640 51.28 43.92 -14.4% 41.63 -18.8% 39.03 -23.9% 34.81 -32.1%
Table 3 shows that for each frequency setting, the longest pulse duration coincides with the largest total change in intensity. Furthermore, the largest intensity difference in total was accomplished when using 640 number of cycles and a single frequency of 5MHz. Another pattern that can be distinguished from column x2 in Table 3, corresponding to the intensity change after the fifth vaporization, is that for both 5MHZ single frequency and 5MHZ+8MHz dual frequency, the percentual change increases for an increasing number of cycles.
17
4. Discussion
The results achieved in this project gives some indications regarding how different parameters of the ultrasound wave affects the efficiency of vaporization of PFP droplets. Additionally, the experiments regarding the assessment of voltage, provided some valuable information of how the droplets can respond to vaporizing ultrasound. As noted from Figure, 7 and 8, the intensity can both be characterized by an inclining and declining behavior upon vaporization.
In section 3.1, this was explained by the disappearance of droplets, as seen in Figure 9, and the formation of microbubbles, illustrated in Figure 10. The formation of microbubbles coincides with increasing intensity. However, both Figure 11 and 12 are mainly characterized by a declining behavior, complemented by some increasing features. These results indicate that the vaporization sequence causes both vaporization and cavitation of droplets, whereby, each vaporization event will cause the intensity to decrease. Nonetheless, in some cases a microbubble persists, causing the intensity to increase and subsequently decrease when the bubble either ascends or cavitates.
In the results from the experiments investigating how the choice of frequency affects the vaporization threshold, illustrated in Figure 11 and Table 2, no sufficient consistency to form an adequate conclusion, is provided. Thereby, it is not possible to support either the use of dual frequencies or the use of higher frequencies to increase the efficiency of the vaporization based on the results provided in this project. These results contradict studies presented in Appendix 1, in which it has been concluded that increasing the frequency is a possible measure towards optimizing the vaporization [13]–[15]. Additionally, the results contradict a study in which the use of dual frequencies was found to increase the ADV efficiency [7].
There are several potential possible explanations for this inconsistency towards previous studies on the subject. Firstly, it is important to consider how the vaporization pulse was designed, Section 2.3.2, consisting of two separate transmissions, corresponding to every other transducer element, to allow for dual frequencies. This design may affect how the droplets are influenced by the vaporizing ultrasound wave, as there is uncertainty concerning how the waves will propagate and interact before hitting the droplet. Furthermore, this design stands in contrast to the design used in a previous study regarding dual frequencies, where a transducer composed by two circular arrays was utilized [7], which can explain the
differences in results. Secondly, previous studies on the subject have emphasized the importance of the droplet characteristics in the choice of frequency, as the superharmonic effect from using higher frequencies is much more pronounced for larger droplets than smaller droplets [14]. For example, one study showed that droplets tended to experience the superharmonic focusing effect when in the size range 6-10µm, whereas droplets in the size range 10-14µm, clearly experienced the focusing effect [14]. Considering the size distribution of the droplets used in this project, Figure 2, with a mean diameter of 2.67±0.6µm, the
droplets may be too small to experience any enhanced effect from increasing frequency, thereby posing a credible explanation for the contradicting behavior in this project towards previous studies.
Regarding the results from the second set of experiments, presented in Figure 12 and Table 3, the most explicit conclusion which can be drawn concerns the relation between the pulse duration of the ultrasound and vaporization efficiency, where longer pulse duration, i.e., a
18 higher number of cycles, corresponds to increasing vaporization of droplets. For all
frequencies, increasing the pulse duration coincided with increasing the percentual difference of intensity before and after vaporization. The only result contradicting this, is the data corresponding to dual frequencies of 5MHz and 8MHz, and a pulse length of 320 cycles, which resulted in a lower total percentual difference in comparison to the same frequency setting with 32 cycles. However, it can be noted from Figure 12, that the mean pixel intensity for 320 cycles experiences an increase followed by a decrease right at the end. As previously explained, such behavior most likely derives from the formation of one or several
microbubbles that do not cavitate immediately, thereby increasing the echogenicity in the ROI. Without this increase in intensity, it is likely that the result would be different, and would support the conclusion that longer pulse durations coincides with higher vaporization efficiency. This conclusion also agrees with the studies performed by Haworth and Kripfgans [16], and Lo et.al. [17], which are described more thoroughly in Appendix 1.
The results presented in this project, provide some indications of how the vaporization efficiency can be improved by increasing the pulse duration. However, ADV is a highly complex phenomenon, where many different parameters influence the efficiency of the process. Despite this, ADV shows great potential for future use in a clinical setting, where it could play major role in e.g. cancer treatment through local drug delivery or gas
embolotherapy. Nonetheless, further studies and trials are necessary before such clinical translation can be performed.
19
5. Conclusion
The aim of this project was to investigate how the use of different frequencies and pulse durations affect the vaporization efficiency of PFP droplets using the Verasonics system. To achieve this, a Matlab script for imaging and vaporizing the droplets was developed, after which three sets of experiments were performed. Firstly, the vaporization effect of different voltages for imaging of the droplets was evaluated, whereby a setting of 15V was found to be suitable for the remaining experiments. Secondly, different settings of frequencies of the vaporization sequence was evaluated, including the use of both single and dual frequencies.
Lastly, the use of different frequencies was combined with varying number of cycles, to assess how the pulse duration affects the vaporization.
The results from the last two sets of experiments indicate that for the CNF-encapsulated PFP droplets used in these experiments, no substantial effect is achieved by using different frequency settings. However, the results regarding pulse duration, show clear indications of the benefit of using longer ultrasound pulses, conforming with previous studies on the subject.
Despite the complexity behind the ADV process, it shows great potential for future clinical use, where it may play a major role in the development of e.g. novel treatment options for cancer therapy. However, further studies and trials are required for complete mapping and understanding of the ADV phenomenon, and how it can be achieved in vivo without harming healthy tissues. When such understanding has been achieved, a clinical translation can be performed.
20
21
References
[1] A. L. Y. Kee and B. M. Teo, “Biomedical applications of acoustically responsive phase shift nanodroplets: Current status and future directions,” Ultrasonics Sonochemistry, vol. 56.
Elsevier B.V., pp. 37–45, 01-Sep-2019, doi: 10.1016/j.ultsonch.2019.03.024.
[2] H. Lea-Banks, M. A. O’Reilly, and K. Hynynen, “Ultrasound-responsive droplets for therapy: A review,” Journal of Controlled Release, vol. 293. Elsevier B.V., pp. 144–154, 10-Jan-2019, doi:
10.1016/j.jconrel.2018.11.028.
[3] P. S. Sheeran and P. A. Dayton, “Improving the Performance of Phase-Change Perfluorocarbon Droplets for Medical Ultrasonography: Current Progress, Challenges, and Prospects,” 2014, doi: 10.1155/2014/579684.
[4] A. A. Doinikov, A. Bouakaz, P. S. Sheeran, and P. A. Dayton, “Dynamics of volatile phase- change contrast agents: Theoretical model and experimental measurements,” in IEEE International Ultrasonics Symposium, IUS, 2014, pp. 2273–2276, doi:
10.1109/ULTSYM.2014.0566.
[5] C.-Y. Lin and W. G. Pitt, “Acoustic Droplet Vaporization in Biology and Medicine,” BioMed Research International, vol. 2013, p. 13, 2013, doi: 10.1155/2013/404361.
[6] S. Guo et al., “Lowering of acoustic droplet vaporization threshold via aggregation,” Applied Physics Letters, vol. 111, no. 25, p. 254102, Dec. 2017, doi: 10.1063/1.5005957.
[7] S. Xu et al., “Acoustic droplet vaporization and inertial cavitation thresholds and efficiencies of nanodroplets emulsions inside the focused region using a dual-frequency ring focused
ultrasound,” Ultrasonics Sonochemistry, vol. 48, pp. 532–537, Nov. 2018, doi:
10.1016/j.ultsonch.2018.07.020.
[8] K. Loskutova, D. Grishenkov, and M. Ghorbani, “Review on Acoustic Droplet Vaporization in Ultrasound Diagnostics and Therapeutics,” 2019, doi: 10.1155/2019/9480193.
[9] R. Daigle, “Sequence Programming Tutorial,” no. April, 2012.
[10] R. Daigle, “Sequence Programming Manual,” 2012.
[11] ATS Laboratories, “Peripheral Vascular Doppler Flow Phantom, Models ATS 524 &525.”
[12] B. Starkoff, “Ultrasound physical principles in today’s technology,” Australasian Journal of Ultrasound in Medicine, vol. 17, no. 1, pp. 4–10, Feb. 2014, doi: 10.1002/j.2205-
0140.2014.tb00086.x.
[13] O. D. Kripfgans, J. B. Fowlkes, D. L. Miller, O. P. Eldevik, and P. L. Carson, “Acoustic droplet vaporization for therapeutic and diagnostic applications,” Ultrasound in Medicine and Biology, vol. 26, no. 7, pp. 1177–1189, Sep. 2000, doi: 10.1016/S0301-5629(00)00262-3.
[14] O. Shpak, M. Verweij, H. J. Vos, N. de Jong, D. Lohse, and M. Versluis, “Acoustic droplet vaporization is initiated by superharmonic focusing,” Proceedings of the National Academy of Sciences of the United States of America, vol. 111, no. 5, pp. 1697–1702, 2014, doi:
10.1073/pnas.1312171111.
22 [15] D. S. Li, O. D. Kripfgans, M. L. Fabiilli, J. Brian Fowlkes, and J. L. Bull, “Initial nucleation site
formation due to acoustic droplet vaporization,” Applied Physics Letters, vol. 104, no. 6, p.
063703, Oct. 2014, doi: 10.1063/1.4864110.
[16] K. J. Haworth and O. D. Kripfgans, “Initial growth and coalescence of acoustically vaporized perfluorocarbon microdroplets,” in Proceedings - IEEE Ultrasonics Symposium, 2008, pp. 623–
626, doi: 10.1109/ULTSYM.2008.0149.
[17] A. H. Lo, O. D. Kripfgans, P. L. Carson, E. D. Rothman, and J. B. Fowlkes, “Acoustic droplet vaporization threshold: Effects of pulse duration and contrast agent,” IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, vol. 54, no. 5, pp. 933–945, May 2007, doi:
10.1109/TUFFC.2007.339.
1
Appendix 1: State of the Art
This appendix aims at providing information concerning acoustic droplet vaporization, and what parameters that are relevant for achieving an efficient phase transition of droplets.
Firstly, the phenomenon ADV will be introduced and explained. Secondly, there is a section presenting the current knowledge concerning what parameters influence the efficiency of the vaporization and how these parameters can be manipulated to optimize the phase transition.
Lastly, a conclusion will summarize the presented information and provide information regarding which parameters have been chosen to be studied in this project.
A.1. Acoustic Droplet Vaporization
Acoustic droplet vaporization (ADV) is a phenomenon in which a droplet with a liquid core undergoes a phase change, whereby it turns into a microbubble with a gas core [1]. In this chapter, the ADV process will be explained in more detail, including information regarding how it can be achieved using ultrasound. Furthermore, there is a section for droplets and microbubbles respectively, which aims at providing knowledge of necessary characteristics of droplets and how microbubbles respond to ultrasound.
1.1 Vaporization
For vaporization of the liquid inside a droplet to occur, the vapor pressure of the liquid within the droplet, needs to exceed the surrounding pressure [2]. The vapor pressure is defined as “the pressure of the specified gas in equilibrium with its own liquid in a closed system at a specified temperature” [1], and it rises with the temperature of the liquid [2]. The temperature at which the vapor pressure is 1atm, is the boiling point of the liquid.
Temperatures exceeding the boiling point, can cause the liquid to undergo a phase change to gas [2]. However, the droplet can also remain in a liquid phase when the boiling point is reached, as it also experiences La Place pressure, which is a pressure resulting from the surface tension between to immiscible liquids [2]. Thereby, the total local pressure imposing on the droplet is composed by both the surrounding pressure, and the La Place pressure. For successful vaporization of the droplet, the vapor pressure needs to exceed the total local pressure imposing on the droplet [1].
1.2. Nucleation and Subpressurization
The state in which a droplet remains in a liquid phase, despite the temperature exceeding the boiling point of the liquid, is known as the droplet being in a superheated state [2]. In this state, a nucleation event is required to initiate the phase change [2]. The required nucleation can be triggered using ultrasound pulses [3], as these causes the pressure to increase and decrease in a cycle. The increased pressure phase is known as compression whereas the
2 decreased pressure phase is known as rarefaction. When the peak negative pressure, i.e. the pressure in the rarefactional phase, is sufficiently large, the vapor pressure of the droplet can exceed the total local pressure, figure 1, allowing the formation of a gas, a state called subpressurization [2].
Figure 1: Illustration of how an ultrasonic wave induces subpressurization and causes a phase change. The gas phase forms when the local pressure wave (blue line), e.g. induced by ultrasound, causes the pressure to drop below the black line, which symbolizes the vapor pressure of a perfluorocarbon (PFC) droplet minus the La Place pressure imposing on the droplet [1].
However, it is not certain that a vaporization takes place despite current subpressurization.
The probability for vaporization depends on both the time of constant subpressurization and the magnitude of the subpressurization [1]. Thereby, the efficiency of ADV is related to ultrasound parameters such as the frequency, the peak negative pressure and the pulse duration [2].
1.3 Droplets
The droplets used for ADV are also called phase change contrast agents (PCCAs) [4]. In medical applications, it is important that PCCAs are non-toxic and immiscible in water [2].
If PCCAs are immiscible in water, they will not diffuse, whereby they can remain stable in the bloodstream for a longer period of time [1]. Another desirable trait of PCCAs is that they should have a boiling point close to the body temperature, to limit the pressure amplitude required for initiating a phase change, while ensuring that the droplets remain in a liquid phase in the blood stream until ultrasound exposure [5]. As perfluorocarbons (PFCs), fulfill these necessary characteristics, they are commonly used in ADV applications. The most used PFC in ADV is perfluoropentane (PFP), which has a boiling point of 29.2℃ [2]. For
stabilization purposes, droplets are often encapsulated by a lipid or polymer shell [6]. The shell increases the La Place pressure imposing on the droplet, whereby the energy required to initiate a phase change is increased [6]. Thereby, droplets composed by e.g. PFP, can
3 remain stable in the blood stream despite the body temperature exceeding the boiling point of PFP. Stabilization of droplets can also be achieved by encapsulation using solid particles, e.g. cellulose nanofibers [7].
1.4 Microbubbles
Upon vaporization of a PCCA, a microbubble with a gas core will be formed. Microbubbles have many different applications in ultrasound. For example, microbubbles can be utilized as ultrasound contrast agents. There are two main properties that makes microbubbles suitable for contrast agent applications. Firstly, microbubbles scatter ultrasound efficiently. Secondly, microbubbles respond non-linearly to ultrasound [8]. The nonlinear response derives from an asymmetrical compression and expansion of the microbubbles [9]. The asymmetry is caused by the limited compression of the bubbles with increasing acoustic pressures, whereas the expansion during the rarefactional phase does not experience this limitation, resulting in oscillations where the expansion is greater than the compression [10], figure 2. The nonlinear response of microbubbles causes them to scatter ultrasound at the transmitted frequency and its harmonics, where the second harmonic is most significant [11]. The harmonic frequencies can be exploited in imaging modalities to achieve contrast enhancement [8].
When exposed to sufficiently high amplitudes of acoustic pressures, the microbubbles can undergo inertial cavitation, whereby the high-pressure amplitude causes them to rapidly grow in volume to a point where they implode [12], figure 2. This process allows
microbubbles to be utilized for other purposes than contrast enhancement, e.g. for achieving drug delivery [13] and to quantitively measure the blood perfusion of organs using
replenishment kinetics upon microbubble destruction [14].
Figure 2: Illustration of the nonlinear response of a microbubble when exposed to an ultrasonic wave, eventually causing cavitation [15].
4
A.2. Optimizing Phase Transition
For the potential use of ADV in a clinical setting, it is of great importance to achieve an efficient phase transition to minimize the number of droplets required for a sufficient amount of microbubbles [16]. This chapter aims at presenting the current knowledge of how the phase transition of droplets can be optimized by reducing the pressure threshold. First, approaches aiming at modifying ultrasound parameters to increase the conversion efficiency will be presented. Second, two more complicated procedures to optimize the phase transition will be explained; the use of standing waves to induce aggregation of the droplets, and the use of a photoacoustic approach, utilizing both laser and ultrasound simultaneously to achieve vaporization.
2.1 Pressure Threshold
For successful vaporization of a droplet, the acoustic pressure must exceed a certain
threshold [17], which as previously described, is related to the different pressures acting on the droplet. Thereby, the probability of ADV events to occur, increases with increasing pressure amplitudes [2], which makes it possible to optimize the conversion by increasing the peak negative pressure. However, one current problem with ADV is that the pressure amplitudes required for achieving vaporization of droplets corresponds to mechanical indices that exceeds the limit that has been approved by FDA for diagnostic ultrasound [18].
Therefore, it is desirable to lower the pressure threshold required to induce ADV to avoid negative effects from the high pressures [17].
There are many different factors that influence the ADV threshold, and they can be categorized into three different groups: environmental factors, droplet properties and ultrasound parameters [17]. Environmental factors include ambient pressure and
temperature. Regarding droplet properties, both droplet size and the boiling point of the liquid core affects the ADV threshold [17]. Lastly, ultrasound parameters that influence the ADV threshold include transmitting frequency, pulse repetition frequency and pulse duration [17]. This project will mainly focus on the ultrasound parameters as these can easily be controlled and manipulated by using a programmable ultrasound system. Moreover, the two additional approaches, utilizing standing waves and photoacoustics respectively, are
included as they use relatively conventional equipment, whereby it would be possible to somewhat replicate the experimental setup using simple means.
2.2 Frequency
According to nucleation theory, the probability of a successful ADV event should decrease with increasing frequency, as increasing frequency causes the subpressurization time
window to become shorter [2]. Furthermore, a higher frequency decreases the probability of a nucleation event inside a droplet as it gives a smaller focal zone [19]. Thereby, it would be expected that the ADV threshold increases with increasing frequency. However, in year 2000, Kripfgans et al. published a study that indicates the opposite relationship between
5 frequency and ADV threshold, as the results show that using higher frequency significantly reduced the ADV threshold [20]. In the study, the ADV threshold was reported to be four times higher for a frequency of 1.5MHz than for 4MHz [20], figure 3.
Figure 3: The pressure threshold as a function of frequency. The concentration denotes the concentration of albumin used in the production of dodecafluoropentane droplets. Reprinted from [20], © [2000] with permission from Elsevier.
This result, contradicting nucleation theory, has later been explained by a study performed by Shpak et al. concerning superharmonic focusing [21]. According to Shpak et al. the ultrasound wave will propagate nonlinearly, resulting in harmonics of the transmitting ultrasound frequency. The harmonics, which have a wavelength in the same size order as the droplets, generate a focusing effect when hitting a spherical shaped droplet. Furthermore, the focusing effect is a result of the differences in acoustic impedance between the droplet and the surrounding. The resulting superharmonic focusing will generate a highly confined pressure amplification within the droplet, thereby increasing the probability of nucleation in that spot[21].
The significance of the focusing effect of the spherical droplets has also been emphasized in a published article by Li et al., in which the results indicate that acoustic lensing within droplets generates local peak negative pressures of greater amplitude, resulting in the formation of nucleation events [22]. Both Li et al. and Shpak et al. also emphasizes the importance of droplet size, where the focusing effect is diminished if the droplet is too small [21], [22]. In contrast, larger droplets will experience a more pronounced focusing effect, thereby achieving a higher amplification of pressure and resulting in a lowered threshold [21]. Moreover, Shpak et al. describes the importance of choice of frequency depending on the droplet size, as smaller droplets require higher frequencies to achieve the local peak negative pressure required for ADV initiation. However, higher frequencies cause higher attenuation, whereby the focusing effect risk being diminished if too high frequencies are used [21].
6 In regard to the studies mentioned above [20]–[22], it can be concluded that a possible measure to optimize the vaporization of droplets is to increase the frequency of the transmitted ultrasound waves. However, there is a delicate balance between getting the desired increased conversion efficiency and risk of loosing this effect due to increased attenuation, which should not be neglected.
Dual frequencies
Another approach of lowering the ADV threshold that is related to transmitting frequency, has been proposed in a study by Xu et al., in which the utilization of dual frequency
ultrasound to lower the ADV threshold has been investigated [17]. Xu et al. suggests that by using two transmitting frequencies simultaneously, it may be possible to lower the ADV threshold by inducing an increase of the peak negative pressure. Moreover, they investigated this phenomenon by using a custom-made ring transducer consisting of two spherical arrays.
The outer and inner transducer array had a transmitting frequency of 1.1MHz and 5MHz respectively. The results from the study indicates that transmitting ultrasound with dual frequencies resulted in the lowest ADV threshold in comparison with using a single frequency of 1.1MHz or 5MHz separately. The most significant change was noted in the comparison between using a single frequency of 1.1MHz and using dual frequencies, where the threshold was found to be 4.5 times higher when only using the lower frequency [17].
This can also be considered an implication of the previously mentioned relation between ADV threshold and frequency, namely that the threshold decreases with increasing frequencies. Lastly, the study concluded that using a dual-frequency transmission both resulted in the lowest ADV threshold and the highest ADV efficiency [17].
2.3 Pulse Repetition Frequency
Pulse Repetition Frequency (PRF) is another acoustic parameter that influences the ADV threshold [17]. The current available literature on the subject indicates that increasing the PRF results in a lowering of the pressure threshold [23], [24]. Seda and Harmon performed a study where the frequency, acoustic pressure and pulse duration were kept constant whereas for the PRF, values between 10Hz and 1000Hz were used [24]. The results from the study illustrated that using higher PRFs resulted in significant vaporization of droplets [24].
Furthermore, Fabiilli et al. found that the ADV threshold was lowered when using a PRF high enough for allowing a fluid volume to be exposed to several ultrasound pulses [23]. In the experimental setup used by Fabiilli et al., a PRF of at least 25Hz was required for allowing a fluid volume to pass by the ultrasound exposure multiple times [23]. As experimental setups vary, it is important to adjust the PRF depending on the setup used.
2.4 Pulse duration
In 2008, Haworth and Kripfgans published a study in which the impact of different pulse durations on the vaporization process of dodecafluoropentane was investigated [25].
7 Haworth and Kripfgans used ultrasound bursts with varying number of cycles, thereby varying the pulse duration, and utilized high-speed photography to study how the droplets were affected by the transmitted bursts. It was concluded that the choice of acoustic pulse duration affected the vaporization process. When only two cycle pulses were used, the droplets did not experience a complete vaporization, whereas ultrasound with thirteen cycles, resulted in complete vaporization in all of the droplets [25].
The impact of pulse duration on ADV has also been investigated by Lo et al., who examined how the pressure threshold varies for different pulse durations [26]. Lo et al. found that for some pulse durations, a change in threshold could be observed, figure 4. When using short pulse durations, up to 1ms, the threshold was constant. Pulse durations longer than 1ms resulted in a drop of the ADV threshold. However, eventually a plateau was reached
whereby no change in threshold was observed when increasing pulse duration further. Lo et al. also emphasize that despite the positive effect on the vaporization threshold, longer pulse durations may be problematic in a clinical setting as it can induce heat deposition in tissue [26].
Figure 4: Pressure threshold for ADV as a function of pulse duration. The bars denote the standard error of the mean. (*) Denotes the use of a 1000 tone burst, each burst having a duration of 20µs, in contrast to the remaining datapoints corresponding to single tone bursts of varying pulse duration [26] © [2007] IEEE.
Another concern regarding the use of long pulse duration in ADV has been reported by Sheeran et al., who have investigated the activation threshold of decafluorobutane using ultra-high-speed microscopy [27]. In their work, Sheeran et al. reported that when using longer pulse durations, it causes bubbles produced early in the pulse to be further affected by the rest of the pulse, which can result in either coalescence or destruction of the
8 microbubbles. Furthermore, Sheeran et al. emphasize that this effect can either be desired or unacceptable, depending on the application of ADV. For contrast imaging purposes, in which it is important to preserve as many bubbles as possible, such effects need to be diminished. However, in applications such as vessel occlusion, coalescence may instead be desired. Overall, there is a delicate balance that needs to be considered when increasing the pulse length, to get the maximum amount of bubbles while minimizing other effects [27].
2.5 Aggregation
In contrast to the previously mentioned approaches to lower the ADV threshold, which are directly related to the manipulation of ultrasound parameters, Guo et al. have reported an alternative approach to reach the same result [18]. Guo et al. proposes the use of additional ultrasound standing waves to generate aggregations of nanodroplets, resulting in collections of nanodroplets that are micron-sized. The study is partly based on the current knowledge regarding the relationship between ADV threshold and droplet size, where larger droplets have a lower ADV threshold. However, droplets with a larger diameter also vaporize into larger microbubbles, which may limit the oscillating response of the microbubbles when exposed to ultrasound, thereby limiting the use of large microbubbles for enhanced contrast imaging [18]. By forming aggregations of nanosized droplets, Guo et al. suggests that it is possible to achieve the same lowering of ADV threshold as for larger droplets, whereas the resulting microbubbles are within the desired size range for contrast enhancing applications [18].
In their study, Guo et al. used two ultrasound transducers, operating at 1MHz, in a water tank to produce the standing waves. Furthermore, a third transducer was used to deliver the ultrasonic waves to initiate vaporization of the droplets. Guo et al. exposed the droplets to standing waves for fourteen minutes to get the desired aggregation of droplets. Moreover, the size of the aggregation was controlled by manipulating the concentration of droplets. In their experiments, Guo et al. found that the ADV threshold was significantly lowered for the aggregations in comparison to the disperse nanodroplets. The authors also suggest an
explanation of this result, based on the previously explained phenomena of superharmonic focusing. As previously mentioned, larger droplets experience a more pronounced effect of superharmonic focusing. Guo et al. concludes that by forming aggregations of droplets, it is possible to induce the same enhanced focusing effect as for large droplets, thereby resulting in a lowering of the ADV threshold [18].
9
2.6 Photoacoustics
An additional alternative approach to avoid the high acoustic pressure amplitudes required for ADV, has been reported by Strohm et al. [28]. In their work, Strohm et al. proposes a solution where vaporization of droplets is achieved using laser irradiation instead of
ultrasound, a process known as optical droplet vaporization (ODV). Furthermore, Strohm et al. mentions additional benefits of implementing ODV instead of ADV. Firstly, ODV avoids the limitations of penetration depth, which is experienced when using ultrasound. As
previously mentioned, ultrasound suffers from attenuation, especially when utilizing higher frequencies. By using laser irradiation with wavelengths corresponding to near-infrared light, such attenuation is diminished, whereby it is possible to achieve a better penetration depth in tissue in comparison to ultrasound. Secondly, ultrasound will experience a great amount of reflection when striking a boundary between tissue and air, making it impossible to penetrate e.g. the lungs. Laser light does not experience this problem and can thereby be delivered to locations that is unreachable to ultrasound [28].
To successfully vaporize a PFC droplet using laser irradiation, the droplet must absorb the incident light [28]. Upon light absorption, the optical energy converts into thermal energy, resulting in a local and temporary rise in temperature that can vaporize the droplets [29].
However, natural PFC droplets do not absorb light of the wavelengths used for ADV efficiently, making droplet modification necessary [28]–[30]. In their study, Strohm et al modified the PFC droplets by incorporating infrared absorbing nanoparticles. They also suggest that other optical absorbing materials can be utilized, such as gold nanoparticles [28]. By using the modified PFC droplets, Strohm et al. successfully vaporized the droplets into microbubbles upon laser irradiation. Moreover, Strohm et al. used ultrasound to image the photoacoustic signal generated when irradiating a droplet with laser [28]. The
photoacoustic signal is generated through thermal expansion of the droplet upon laser irradiation, resulting in acoustic waves [29], illustrated in figure 5, which thereby can be measured using an ultrasound transducer. In another study by Strohm et al., it is suggested that two different types of absorbing particles can be introduced in the droplets, allowing for vaporization and photoacoustic imaging at two different wavelengths [30].
Figure 5: Illustration of the generation of a photoacoustic signal upon thermal expansion due to laser excitation. The photoacoustic signal is enhanced when the laser triggers a phase change of the medium [31].
10 In their work, Strohm et al. only achieved vaporization of droplets using optical means.
However, there are also studies in which vaporization has been achieved using laser and ultrasound simultaneously [29], [32], [33]. Feng et al. utilized the simultaneous use of laser and ultrasound to reduce the threshold of phase transition [33]. Moreover, Feng et al.
emphasize problems concerning insecurities with spatial and temporal control that are present in ADV, which were evident in their experiments when solely ADV was utilized.
When including a focused pulsed laser simultaneously with ultrasound, they found that vaporization only occurred at the region of the laser focus, and only when the laser was activated, thereby overcoming the issues with spatial and temporal control. Despite other references emphasizing the importance of introducing light absorbing particles in the droplet, Feng et al. used perfluoropentane with a bovine serum albumin encapsulation, without further introduced particles. No motivation for this choice is provided in the article. Finally, Feng et al. found that a lowering of the energy threshold for phase change was achieved when using laser and ultrasound simultaneously, making it a safer alternative approach in future therapeutic applications [33].
Liu et al. have further investigated how laser and ultrasound can be combined to lower the vaporization threshold [32]. In contrast to Feng et al., who only used perfluoropentane with a bovine serum albumin encapsulation [33], Liu et al. used human serum albumin
encapsulated perfluoropentane with gold nanoparticles [32]. Furthermore, Liu et al.
specifically focused on the use of synchronized laser and ultrasound in sonoporation applications. Sonoporation is a process in which the permeability of cellmembranes are transiently increased, for example by inducing cavitation in contrast agents using acoustics [32]. In their study, Liu et al. investigated if the vaporization could be made more efficient, i.e. optimizing sonoporation effects, when synchronizing ultrasound with pulsed laser
irradiation. Synchronization of ultrasound and laser was achieved by applying the laser in the peak negative pressure phase of the ultrasound waves. The results show that by
synchronizing ADV and ODV, it is possible to achieve vaporization and cavitation using laser energies and ultrasound pressures that are below the FDA approved safety limits.
Moreover, the pulse repetition frequency of the pulsed laser was found to be critical for successful vaporization. When using lower laser fluence, which corresponds to the energy delivered by unit area, and pressure amplitudes, a higher PRF was required for vaporization.
Finally, Liu et al. concluded that the synchronization ADV and ODV is beneficial for initiating vaporization and cavitation in sonoporation applications, as it reduces the pressure and energy thresholds necessary when using ADV and ODV separately.
A.3. Conclusion
Acoustic droplet vaporization refers to the process where an acoustic wave induces a phase change in a droplet with a liquid core, resulting in a gas-filled microbubble. For achieving a sufficient amount of produced microbubbles while minimizing the number of required droplets, it is important to optimize the vaporization process. One approach of such
optimization includes increasing the acoustic pressure. However, higher pressure amplitudes may induce harm to tissue when used in a clinical setting, whereby supplementary
approaches are necessary. Therefore, there is extensive material in scientific literature aiming at finding such alternative approaches.
