Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Single-cell analysis with high frequency ultrasound
(USC Thesis Other)
Single-cell analysis with high frequency ultrasound
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
SINGLE-CELL ANALYSIS WITH HIGH FREQUENCY ULTRASOUND
by
Min Gon Kim
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillments of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOMEDICAL ENGINEERING)
August 2017
Copyright 2017 Min Gon Kim
ii
DEDICATION
With grateful thanks to my beloved parents, Kyunghee Ha and Yong Tae Kim
for their endless love and prayers, and unflagging support
iii
ACKNOWLEDGEMENTS
I would like to express my sincere gratitude to my advisor, Dr. K. Kirk Shung for
his thoughtful guidance, warm encouragement and unwavering support during my
doctoral studies. His mentorship was paramount in providing a well-rounded experience
consistent my long-term career goals. Also, he provided me precious opportunities to
inspire and enrich my growth as a Ph. D and kept consistent interest in the progress of my
thesis
I gratefully acknowledge my committee members: Dr. Robert H. Chow, Dr. Jesse
T. Yen, Dr. Megan L. McCain, and Dr. Keyue Shen for their agreement to serve as my
committee members and for taking the time and effort to evaluate my work. Their
insightful comments and suggestions considerably improved this dissertation.
It is always my pleasure to work with good people. I thank all of my colleagues in
NIH Resource Center on Medical Ultrasonic Transducer Technology. Morning coffee
meetings with Dr. Jin Ho Chang, Dr. Jinhyoung Park, Dr. Bong Jin Kang, and Dr.
Changhan Yoon were a source of joy and new ideas, as well as a chance to talk about my
concerns. I also thank all the former and current lab members, especially, Dr. Qifa Zhou,
Dr. Dong-Guk Paeng, Dr. Hyung Ham Kim, Dr. Jae Youn Hwang, Dr. Changyang Lee,
Dr. Hojong Choi, Dr. Fan Zheng, Jay Williams, Dr. Sangpil Yoon, Dr. Ruimin Chen,
Hayong Jung, Chi Woo Yoon, Hae Gyun Lim, Nestor E. Cabrrera-Munoz, Robert
Wodnicki, and Xiaoyang Chen.
It is a great pleasure to have the opportunity to sincerely thank all of my friends at
Los Angeles Onnuri Church, USC, and UCSD for their friendship, encouragement, and
iv
assistance during my graduate studies. Among them, Wonhyuk Choi, Hee Chur Yun,
Byongchan Kim, Myoung Sung Kim, Dong Oh Lee, Hong Bae, Dr. Jae Kyung Suh,
Youngin Kwon, Douglas Choi, Jaekwon Kim, Kwangwon Kim, Eun Joo Seo, Jun Hyung
Park, Young Chun Ahn, Chiho Choi, Yoo Ji Hwang, Dr. Joanna Jiyeon Lee, Myung Sung
Kang, Hakyung Chung, John K. Ryu, A Kung Woo, Dr. Seung Yeop Lee, Yae Sam Song,
and Hwana Lee deserve special thanks.
I would like to thank my lovely family, my parents, Kyunghee Ha and Yong Tae
Kim. I thank them for all their understanding, patience, and support. The completion of
this dissertation would not have been possible without my parents.
Lastly, I humbly thank God for guiding me to the place I am today.
v
TABLE OF CONTENTS
DEDICATION .................................................................................................................... ii
ACKNOWLEDGEMENTS ............................................................................................... iii
LIST OF TABLES ........................................................................................................... viii
LIST OF FIGURES ........................................................................................................... ix
ABSTRACT .................................................................................................................. xiv
CHAPTER 1 Introduction ..................................................................................................1
1.1 Background ................................................................................................. 1
1.2 High Frequency Ultrasound ........................................................................ 1
1.3 Objective of Research ................................................................................. 2
CHAPTER 2 Impedance Matching Network for High Frequency Ultrasonic
Transducer....................................................................................................4
2.1 Introduction ................................................................................................. 4
2.2 Materials and Methods ................................................................................ 8
2.2.1 Ultrasonic Transducers, Impedance Analyzer, Pulse-echo and Insertion
Loss Measurement System ........................................................................... 8
2.2.2 Design of Impedance Matching Network ................................................... 10
2.2.3 Validation of Impedance Matching Network ............................................. 14
2.3 Results ....................................................................................................... 16
2.3.1 Impedance Analysis Results ....................................................................... 16
2.3.2 Optimization of Impedance Matching Network ......................................... 19
2.3.3 Electrical Characteristics of Optimized Impedance Matching Network .... 21
2.3.4 Performance Validation for Cellular Applications ..................................... 23
2.4 Discussions ............................................................................................... 25
vi
CHAPTER 3 High Frequency All-in-one Front-end System for Multifunctional High
Frequency Ultrasound Imaging..................................................................28
3.1 Introduction ............................................................................................... 28
3.2 Materials and Methods .............................................................................. 29
3.2.1 Design of the Front-end System ................................................................. 29
3.2.2 Fabrication of the System and Measurement Setup .................................... 32
3.3 Results ....................................................................................................... 33
3.3.1 Electrical characteristics of the System ...................................................... 33
3.3.2 Performance Evaluation for Ultrasound Applications ................................ 34
3.4 Conclusion ................................................................................................ 34
CHAPTER 4 Acoustic Transfection with High Frequency Ultrasound ..........................36
4.1 Introduction ............................................................................................... 36
4.2 Materials and Methods .............................................................................. 39
4.2.1 Acoustic Transfection System and Controllable Treatment Conditions..... 39
4.2.2 Cell Culture ................................................................................................. 41
4.2.3 Live-cell Fluorescence Imaging .................................................................. 41
4.2.4 Intracellular Delivery Score for Optimal Treatment Conditions ................ 43
4.2.5 Intracellular Delivery of Macromolecules and Simultaneous Intracellular
Delivery of Two Molecules ........................................................................ 44
4.3 Results ....................................................................................................... 45
4.3.1 Time-based Intensity Changes of Propidium Iodide molecule ................... 45
4.3.2 Intracellular Delivery of Small-sized Molecule .......................................... 47
4.3.3 Optimal Treatment Conditions on Human Cancer Cell Lines indicated by
Intracellular Delivery Score ........................................................................ 49
4.3.4 Intracellular Delivery of Macromolecules with Optimal Treatment
Conditions ................................................................................................... 55
4.3.5 Simultaneous Intracellular Delivery of Two Molecules with Optimal
Treatment Conditions.................................................................................. 57
vii
4.4 Discussions ............................................................................................... 60
CHAPTER 5 Label-free Acoustic Sensing of a Single Cell Trapped by High
Frequency Ultrasound ................................................................................67
5.1 Introduction ............................................................................................... 67
5.2 Materials and Methods .............................................................................. 71
5.2.1 Two specifications for the Proposed Driving Condition for High Frequency
Ultrasound Mircrobeam .............................................................................. 71
5.2.2 High Frequency Ultrasound based Label-free Cell Separation System...... 72
5.2.3 Effects of High Frequency Ultrasound Microbeam in terms of Different
Driving Conditions...................................................................................... 73
5.2.4 Cell Preparation .......................................................................................... 74
5.2.5 Cell Viability Study and Statistical Analysis .............................................. 74
5.3 Results ....................................................................................................... 75
5.3.1 Characterization of High Frequency Ultrasound Microbeam with the
Proposed Driving Condition ....................................................................... 75
5.3.2 Size Determination of a Trapped Single particle ........................................ 78
5.3.3 Separation between Red Blood Cells and Cancer Cells ............................. 79
5.3.4 Cell Viability Study .................................................................................... 82
5.4 Discussions ............................................................................................... 83
CHAPTER 6 Summary and Future Works ......................................................................88
6.1 Summary ................................................................................................... 88
6.2 Future Works ............................................................................................ 89
6.2.1 A Strategy for Multifunctional Single-cell Analysis .................................. 89
6.2.2 Investigation and manipulation of cell membrane electrical impedance
during Acoustic Transfection ...................................................................... 90
BIBLOGRAPHY ...............................................................................................................92
viii
LIST OF TABLES
Table 2.1 Summary of the optimization of impedance matching network (IMN) for
ultrasonic transducer 1 (TR1) and ultrasonic transducer 2 (TR2) .....................................19
Table 4.1 Criterion for the intracellular delivery score (IDS) ...........................................44
Table 4.2 Summary of the optimal treatment conditions ...................................................55
ix
LIST OF FIGURES
Figure 2.1 Experimental setups for (a) pulse-echo and (b) insertion loss (IL)
measurement. .......................................................................................................................9
Figure 2.2 Schematic diagram of impedance matching network (IMN) for an
ultrasonic transducer. Zin represents input impedance seen from the excitation source
and should be equal to the complex conjugate of the Z1. Z1 and Z2 indicate the
electrical impedance of the excitation source and ultrasonic transducer, respectively.
(a) When R1 < R2, IMN is placed in shunt with R2 to lower R2 (b) When R1 > R2,
IMN is placed in series with R2 to boost R2. The component values and topology
such as capacitor (C)/inductor (L), C/C, L/L and L/C were dependent upon measured
real and imaginary values of electrical impedance of the ultrasonic transducer at the
target center frequency on a Smith chart. (c) The example of IMN design procedure.
The input electrical impedance of the ultrasonic transducer moved toward the center
of the Smith Chart by appropriate inductor and capacitor component values and
topology at the target center frequency. (d) Picture of IMN with ultrasonic transducer
1. The printed circuit board (PCB) was implemented with shunt-added capacitor of
100 pF and series-added inductor of 8 nH, and an overall size was 21 by 11 mm
2
. .........11
Figure 2.3 Experimental setup for the validation of impedance matching network
(IMN) and cell applications. Precise focal depth of the ultrasonic transducer 1 (TR1)
with IMN was adjusted by moving TR1, attached to a 3D linear translation/rotation
stage, while finding the optimized echo magnitude using a pulser/receiver and an
oscilloscope. After focusing TR1 with IMN, an acoustic pulse with peak-to-peak
voltage (V
pp
) and treatment time (T
t
) was applied to the surface of a plastic petridish
and HeLa cells. Tt was composed of the number of sine waves with the optimized
center frequency of TR1. The test and experiment were carried out comparing
tangible effects obtained by TR1 with and without IMN. .................................................16
Figure 2.4 Measured magnitude of electrical admittance (|Y|), phase angle on the left
column ( θ
z
), real (R) and imaginary (X) values on the right column of electrical
impedance of (a) transducer 1 (TR1) and (b) transducer 2 (TR2). The target center
frequency (TCF) of TR1 and TR2 were measured as 105 MHz and 150 MHz,
respectively, when |Y| was maximum value and θ
z
was zero as indicated with solid
arrows in left column in (a) and (b). At the target center frequency, measured R and X
were 1.5 Ω and 0.2 Ω for TR1 and 2.5 Ω and 0.1 Ω for TR2, respectively. .....................17
Figure 2.5 Pulse-echo waveform and echo spectrum of (a) ultrasonic transducer 1
(TR1) and (b) ultrasonic transducer 2 (TR2). Left column represents the pulse-echo
measurement without impedance matching network (IMN) and right column shows
the pulse-echo measurement with IMN. For TR1, the optimized center frequency
(OCF) was almost the same as the target center frequency (TCF, 105 MHz) and the
x
optimized echo magnitude (OEM) in the right column of (a) was increased by 65%
compared to the reference echo magnitude (REM) in the left column of (a). For TR2,
the optimized center frequency (OCF) was similar to the target center frequency (TCF,
150 MHz), measured by impedance analysis and the optimized echo magnitude
(OEM) in the right column of (b) was enhanced by 33% compared to the reference
echo magnitude (REM) in the left column of (b). .............................................................18
Figure 2.6 Measured electrical performance of ultrasonic transducers without and
with impedance matching network (IMN). (a) Magnitude of electrical impedance (|Z|),
(b) phase angle ( θ
z
) and (c) insertion loss (IL) of ultrasonic transducer 1 (TR1)
without and with IMN and ultrasonic transducer 2 (TR2) without and with IMN,
respectively. For TR1, |Z|, ( θ
z
) and IL were measured as 35 Ω, 85˚ and -21.5 dB at
the reference center frequency (RCF) without IMN, and 65 Ω, 76˚ and -13 dB at the
optimized center frequency (OCF) with IMN, respectively. For TR2, |Z|, ( θ
z
) and IL
were 32 Ω, 77˚ and -34 dB at the RCF without IMN, and 52 Ω, -40˚ and -29 dB at the
OCF with IMN, respectively..............................................................................................22
Figure 2.7 Circular dents on a 35 mm plastic petridish generated by acoustic pulses
from ultrasonic transducer 1 (TR1) (a) with impedance matching network (IMN) (b)
without IMN. Acoustic pulses were generated by the peak-to-peak voltage (V
pp
) from
40 V (left) to 70 V (right) with increments of 10 V and treatment time (T
t
) of 420 μs.
The scale bar indicates 20 µm. ...........................................................................................23
Figure 2.8 Cell responses to an acoustic pulse generated by ultrasonic transducer 1
(TR1) with impedance matching network (IMN) and TR1 without IMN. Peak-to-peak
voltage (V
pp
) and treatment time (T
t
) were 55 V and 420 μs, respectively. Arrows
indicate the area where the acoustic pulse was applied. The scale bars indicate 20 µm.
(a)-(c) TR1 with IMN (a) Bright-field image of a HeLa cell was acquired right after
treatment. The circular dent in the middle of the cell was generated by the acoustic
pulse. (b) Fluorescence image of the same cell stained with calcein was captured. (c)
Fluorescence image of the same cell stained with propidium iodide (PI) was acquired.
After 4 hours of incubation after treatment, the cell treated by TR1 with IMN was
dead. (d)-(f) TR1 without IMN (d) Bright-field image of a HeLa cell was taken right
after treatment. (e) Fluorescence image of the same cell stained with calcein was
taken. (f) Fluorescence image of the same cell stained with PI was captured. After 4
hours of incubation after treatment, the cell exposed to the treatment was alive. .............24
Figure 3.1 Proposed pulse generator (a) Schematics of pulse generator (b) Output
wave forms at the blue dotted position from A to F and the unit of the y-axis is Volts. ...30
Figure 3.2 High frequency mechanical ultrasound system setup. .....................................33
Figure 3.3 Measured output pulses from the proposed transmitter (a) Output pulses in
the time domain (b) Output pulses in the frequency domain (c) Two cycle pulse train
from the combined pulse generators. .................................................................................34
xi
Figure 3.4 Wire target study results (a) B-mode image and scale of both width and
depth is mm (b) Axial brightness profile and scale of x-axis is µm and y-axis is dB (c)
Lateral brightness profile and scale of x-axis is µm and y-axis is dB. ..............................34
Figure 4.1 The acoustic transfection system for intracellular delivery of membrane-
impermeable molecules into the cells. Fluorescence microscope was used to monitor
fluorescence changes on the targeted cells, and acoustic pulse generation system was
utilized to precisely focus ultrasonic transducer with impedance matching network
(IMN) and generate an acoustic pulses with controllable treatment conditions such as
peak-to-peak voltage (V
pp
), treatment time (T
t
) and number of cycles (#). .......................40
Figure 4.2 Intracellular delivery of PI inside the HeLa cell with the treatment of 23 V
/ 23 μs and 23 V / 30 μs, respectively (a)-(b) Time-based PI intensity changes.
Arrows show treated cells and scale bars indicate 40 μm..................................................46
Figure 4.3 Intracellular delivery of PI (a) Time-based PI intensity changes inside the
HeLa cell. (b) There was no PI intensity before treatment (0 sec) (c)-(e) Diffused and
slightly increased PI intensity caused by repeated treatment (3 sec, 63 sec and 123
sec). Arrows show treated cells and scale bars indicate 40 μm. ........................................47
Figure 4.4 Intracellular delivery of Alexa fluor 488-labeled of 3 kDa dextran with (a)-
(b) Treatment of 23 V / 16 μs (c)-(e) Treatment of 23 V / 23 μs. The first and second
columns indicate bright-field and fluorescence images of a HeLa cell after 0.5 hour of
treatment. The third and fourth columns show bright-field and fluorescence images of
the HeLa cell after 40 hours of treatment. Arrows show treated cell and scale bars
indicate 40 μm. ...................................................................................................................49
Figure 4.5 Percentage of delivery efficiency with different treatment conditions for (a)
HeLa (b) MCF-7 (c) MCF-10A and (d) MDA-MB231 cell lines. ....................................50
Figure 4.6 Cell membrane permeability with different treatment conditions for (a)
HeLa (b) MCF-7 (c) MCF-10A and (d) MDA-MB231 cell lines. ....................................51
Figure 4.7 Percentage of cell viability after 4 and 20 hours after acoustic transfection
with different treatment conditions for (a) HeLa (b) MCF-7 (c) MCF-10A and (d)
MDA-MB231 cell lines. ....................................................................................................53
Figure 4.8 Intracellular delivery graph (IDG) using the intracellular delivery scores
(IDS) after 4 and 20 hours of acoustic transfection for different treatment conditions.
IDS is plotted on y-axis under six different V
pp
. Six different T
t
were applied at each
V
pp
. IDG after 4 and 20 hours was used to find optimal treatment conditions for (a)
HeLa (b) MCF-7 (c) MCF-10A and (d) MDA-MB231 cell lines. ....................................55
Figure 4.9 Intracellular delivery of 70 kDa dextran labeled with Oregon Green using
optimal treatment conditions. Left column indicates bright-field images before the
xii
acoustic transfection, and middle and right columns show bright-field and
fluorescence images after 0.5 hour of the acoustic transfection for (a) HeLa (b) MCF-
7 (c) MCF-10A and (d) MDA-MB231 cell lines. Arrows show acoustic transfected
cells and scale bars indicate 40 μm. ...................................................................................57
Figure 4.10 Simultaneous intracellular delivery of 70 kDa dextran labeled with
Oregon Green and propidium iodide (PI) under optimal treatment conditions. The
first and second columns indicate bright-field images before and after acoustic
transfection, respectively, and the third and fourth columns show fluorescence images
of 70 kDa dextran and PI after 0.5 hour of the acoustic transfection, respectively, and
the fifth columns overlapped images of the two different fluorescence images for (a)
HeLa (b) MCF-7 (c) MCF-10A and (d) MDA-MB231 cell lines. Arrows indicate
acoustic transfected cells and scale bars indicate 40 μm. ..................................................58
Figure 4.11 The intracellular delivery efficiency graph of 70 kDa dextran labeled with
Oregon Green under optimal treatment conditions. (a) Highest percentage of 70 kDa
dextran delivery efficiency is 89% (16/18) with the treatment condition of 24V / 30μs
for HeLa cell. (b) Highest percentage of 70 kDa dextran delivery efficiency of 83%
(5/6) and 83% (15/18) under treatment conditions are 24V / 30μs and 27V / 30μs,
respectively for MCF-7 cell. (c) Highest percentage of 70 kDa dextran delivery
efficiency is 83% (5/6) under the treatment condition of 27V / 18μs for MCF-10A
cell. (d) Highest percentage of 70 kDa dextran delivery efficiency of 72% (13/18)
under the treatment condition of 21V / 30μs for MDA-MB231 cell. ................................59
Figure 4.12 The simulation of acoustic pressure filed was performed by a commercial
finite element modeling software under ideal conditions treatment. .................................65
Figure 5.1 High frequency ultrasound based label-free cell separation system which
was used for individually measuring ultrasound backscattering coefficients of a
trapped single object on a mylar film. ...............................................................................72
Figure 5.2 Characterization of high frequency ultrasound microbeam with a proposed
driving condition. (a)-(d) An ability to trap and move a polystyrene microsphere
using the proposed monocycle ultrasound pulse with high PRF was confirmed by
taking time-based bright field images before and after influence of high frequency
ultrasound microbeam. White and red dashed circles were described as the initial and
moving location of the high frequency ultrasound beam, respectively. Scale bars
indicate 100 µm. (e-g) Acoustic trapping force and its effects were described by
distances between a trapped bead and its adjacent beads. Red dashed circles and red
arrow mean the location of the high frequency ultrasound beam and the distances
between a trapped bead and its adjacent beads. The impacts had the greatest in (e)
voltage: 10V / on time: 1μs / PRF: 1 kHz, followed by (f) voltage: 10V / on time: 100
ns / PRF: 10 kHz, and (g) voltage: 10V / on time: 6.7 ns / PRF: 167 kHz. (H-J)
Modified acoustic trapping force and its effects were demonstrated. There was very
similar effects after changing driving conditions (h) voltage: 3.8V / on time: 1μs /
xiii
PRF: 1 kHz, (i) voltage: 4.1V / on time: 100 ns / PRF: 10 kHz compared with
proposed driving condition (j) voltage: 10V / on time: 6.7 ns / PRF: 167 kHz. ................76
Figure 5.3 Size determination of a trapped micron-sized single particle (5 μm / 10 μm)
was performed. The ultrasound backscattering coefficient obtained by each
polystyrene microbead of -109.52 ± 0.75 dB (5 μm) and -98.84 ± 0.84 (10 μm) was
calculated and averaged (n=20). Also, this difference was strongly significant
according to statistical analysis (p-value < 0.01). ..............................................................79
Figure 5.4 A capability of trapping and moving each red blood cell (RBC) and normal
SV40 immortalized epithelial prostate (PNT1A) cell using the proposed monocycle
ultrasound pulse with high PRF was confirmed. The initial and moving locations of
the high frequency ultrasound beam are indicated as black and red dashed circles,
respectively. The scale bars indicate 20 µm. Manipulation of (A) RBC and (B)
PNT1A cell. .......................................................................................................................80
Figure 5.5 Size determination of a trapped cell (RBC / PNT1A cell) was studied. The
ultrasound backscattering coefficient obtained by each RBC and PNT1A cell of -
109.03 ± 0.77 dB and -106.74 ± 0.22 dB was calculated and averaged (n=16). Also,
statistical analysis shows statistical significance (p-value < 0.01). ...................................83
Figure 5.6 Cell viability study was performed, and normalized mean fluorescence
intensity was presented. Both normalized mean fluorescence intensity were slightly
decrease, but not significantly different between control condition and the proposed
driving condition for (a) RBCs (p-value: 0.29 > 0.05, n=10) and (b) PNT1A cells (p-
value: 0.15 > 0.05, n=10). ..................................................................................................83
Figure 6.1 Experimental setup consists of the ultrasonic transducer with IMN and
patch-clamp micropipette connected with EPC-9 patch-clamp amplifier with PULSE
software. .............................................................................................................................91
xiv
ABSTRACT
Single-cell analysis is an essential step to elucidate and understand cellular
behaviors and functional responses in a large cell population at the individual cell level.
The application of a high frequency ultrasound is a promising approach for single-cell
analysis, providing unique characteristics such as being easy-to-use and cost-effective as
well as having minimal effects on normal cell physiology and a label-free investigation
under high specificity and spatial resolution. This dissertation presents single-cell
analysis with high frequency ultrasound and covers single-cell analysis and custom-built
high frequency ultrasound electronics for single-cell analysis. For single-cell analysis, a
new transfection technique using a highly focused high frequency ultrasound was
proposed and demonstrated by live-cell fluorescence imaging of time-based intensity
changes for a fluorescent propidium iodide (PI) molecule and the cytoplasmic delivery of
3 kDa dextran. This transfection technique was developed further as a versatile and
adaptable transfection tool for the intensive investigation of treatment conditions between
different human cancer cell lines. This approach was verified by the intracellular delivery
of 70 kDa dextran and the simultaneous intracellular delivery of two molecules including
70 kDa dextran and PI into four human cancer cell lines using optimal treatment
conditions with high delivery efficiency with minimal cell membrane disruption. Also, a
label-free single-cell separation technique using the highly focused high frequency
ultrasound was presented and demonstrated by individually identifying the size of the
trapped single object based on the ultrasound backscattering coefficient. The proposed
strategy provides beneficial characteristics, including it being an easy-to-use, cost-
xv
effective, label-free investigation of physical properties with minimal effects on normal
cell physiology with a much more accurate analysis at the single-cell level. Moreover, for
custom-built high frequency ultrasound electronics for single-cell analysis, an impedance
matching network was developed to maximize energy transmission from the excitation
source to high frequency ultrasonic transducers for cell manipulation as well as to
achieve low-input parameters for the safe operation of ultrasonic transducers. Also, a
high frequency ultrasound all-in-one front-end system with a novel driving condition to
overcome the limitations of commercially available systems was designed and developed.
This system could be used for a label-free acoustic sensing method and for
multifunctional high frequency ultrasound imaging.
1
CHAPTER 1 Introduction
1.1 Background
Over the last ten years, there have been significant breakthroughs in cancer
research. However, cancer is still one of the leading causes of death despite remarkable
progress in understanding its biology and treatment. According to the World Cancer
Report 2014, there were 14 million new cancer cases, 8.2 million cancer deaths
worldwide, and 32.6 million people living with cancer in 2012 (World Cancer Report,
2014). With cancer being a leading cause of death, research for cancer treatments based
on understanding cellular properties can be our best defense against cancer (Alberts et al.,
2002; Lodish et al., 2000). However, conventional cellular analysis techniques could
provide averaged data which might result in misleading information. To address this
critical problem, a range of single-cell analysis techniques has been developed to
examine cellular properties and functional responses between individual cells at a single-
cell level. To achieve accurate information, the techniques should simultaneously satisfy
essential requirements such as being easy-to-use and cost-effective as well as minimal
effects on normal cell physiology and a high spatial resolution.
1.2 High Frequency Ultrasound
Medical ultrasounds have been widely used to image internal body structures due
to its unique characteristics of being non-invasive, real-time imaging, cost-effective, and
portable. However, to determine smaller and specific objects at the cellular level, the use
2
of higher frequencies is necessary to achieve a narrower and sharper resolution. Spatial
resolution refers to the ability to visualize two separate objects which are close together,
and it is further categorized into axial resolution (R
A
) and lateral resolution (R
L
). The
spatial resolution can be calculated by the following equations:
𝑅 𝐴 =
𝑐 2∙(𝐵𝑊
−6𝑑𝐵
)
(1.1)
𝑅 𝐿 = 𝑓 #
∙ 𝜆 (1.2)
where c is the sound velocity, 𝑓 #
is defined as the focal distance to the aperture
size of an ultrasonic transducer (f
#
= Z
f
/ D, Z
f
: the focal distance, D: the aperture size of
the ultrasonic transducer), λ is the wavelength defined as the ratio of the sound velocity to
the center frequency of the ultrasonic transducer (c / f
c
, f
c
: the center frequency of the
ultrasonic transducer), and BW
-6dB
is a -6 dB bandwidth of the ultrasonic transducer
(Foster et al., 2000). Thus, the use of a high frequency ultrasound can be applied to
single-cell analysis with significantly improved resolution.
1.3 Objective of Research
The goal of this research is to develop high frequency ultrasound electronics for
single-cell analysis and provide a new approach for single-cell analysis with high
frequency ultrasound. The application of a high frequency ultrasound is a promising
3
approach for single-cell analysis, providing unique characteristics such as being easy to
use and cost-effective as well as having minimal effects on normal cell physiology and a
label-free investigation under high specificity and spatial resolution. This thesis consists
of five chapters. Chapter 1 introduces general background and high frequency ultrasound.
Chapter 2 demonstrates the development of an impedance matching network for high
frequency ultrasonic transducers for cellular applications. Chapter 3 shows a high
frequency ultrasound all-in-one front-end system for multifunctional high frequency
ultrasound imaging. Chapter 4 describes acoustic transfection with high frequency
ultrasound. Chapter 5 presents a label-free acoustic sensing of a single cell trapped by
ultrasound microbeam. Chapter 6 summarizes all results in this thesis and discusses
potential future works.
4
CHAPTER 2 Impedance Matching Network for High
Frequency Ultrasonic Transducer
2.1 Introduction
An ultrasonic transducer/array is a crucial component of an ultrasonic scanner
such as those widely used in clinical diagnosis (Kim et al., 2013; Lockwood et al., 1996;
Merks et al., 2004; Paeng et al., 2004; Silverman et al., 2001; Yoon et al., 2015a; Yoon et
al., 2015b). There has been a growing attention in the frequency range higher than 100
MHz for applications in single cell analysis, acoustic trapping, and studying
mechanotransduction with improved lateral resolution (Hwang et al., 2014a; Hwang et al.,
2014b; Hwang et al., 2015; Lee et al., 2009; Lee et al., 2011). In these applications, a
tightly focused high frequency ultrasound beam is required (Lee et al., 2010). When
fabricating highly focused ultrasonic transducers, the transducers having a large aperture
are much easier to press-focus. A transducer with a large aperture also generates a
stronger pressure field under the same input parameters, such as voltage and pulse
duration. Therefore, to produce the same acoustic pressure field, a transducer with a large
aperture size is safer and more effective than a transducer with a small aperture size
because lower input voltage and shorter pulse duration are needed.
The performance of a transducer with a large aperture is generated only when the
electrical impedance between the ultrasonic transducer and a excitation source are
electrically matched. As the operating frequency approaches the resonance frequency of
the piezoelectric material, the magnitude of the acoustic pulse becomes larger because the
5
energy transfer condition is optimized (Ben-Yaakov et al., 2004). In other words, at the
resonance, the magnitude of the electrical impedance of the ultrasonic transducer
becomes optimized allowing maximal energy conversion. Energy conversion
optimization largely depends upon the input electrical impedance of the transducer which
in turn depends upon the dielectric constant of the piezoelectric material (Snook et al.,
2002). Electrical impedance mismatch between the ultrasonic transducer and the
excitation source results in a large energy reflection between them thereby wasting
energy, which leads to excessive power consumption when trying to achieve the same
acoustic pressure field (Bhuiyan et al., 2012; Cannata et al., 2003; Kong et al., 1995;
Stephen et al., 2006). Therefore, an impedance matching network (IMN) is necessary for
the optimization of power transfer and efficiency by minimizing reflections between the
ultrasonic transducer and the excitation source. This is particularly acute in applying
large aperture high frequency transducers, which have usually extremely low electrical
impedance at the resonance frequency, to a number of cellular applications given that
popular and highly sensitive piezoelectric materials typically have fairly high dielectric
constants, which are unfavorable for the design of large aperture devices (Hwang et al.,
2014a; Hwang et al., 2014b; Hwang et al., 2015; Lee et al., 2009; Lee et al., 2011).
A range of schemes of broadband IMN has been applied to ultrasound imaging
applications to achieve wide bandwidth because it enables to improve axial resolution
(Andersen et al., 1977; Augustine et al., 1979; Pozar, 1998; Shung, 2005; Youla, 1964).
Broadband IMN generates a spectral response that has a wide bandwidth with a gradual
peak and echo signal energy distributed over a broad range of frequencies (Duffer et al.,
6
1995). Therefore, broadband IMN may not be suitable in applications where maximum
intensity at the resonance frequency is required (Hoarau et al., 2007).
The commercially available transducer simulation software, PiezoCAD (Sonic
Concepts, Bothell, WA), has been widely used in transducer design; however, it is very
challenging to fabricate transducers in the frequency range higher than 100 MHz based
on PiezoCAD modeling. Since PiezoCAD modeling does not take into account the
effects of variations in thickness of the piezoelectric material at higher frequency and the
electromechanical effects resulting from press-focusing, PiezoCAD modeling is not
likely to show good agreement in terms of magnitude of admittance (|Y|), phase angle
( θ
z
), the real (R) and the imaginary (X) values compared with impedance analysis based
on the real fabricated transducer (Cannata et al., 2001; Cannata et al., 2003; Lam et al.,
2013; Herickhoff et al., 2010). For low frequency transducers, PiezoCAD modeling
agrees well with impedance analysis based on the real fabricated transducer because the
thickness variations and other aspects of the transducer design process have a more
negligible impact on the characteristics of the transducers. On the other hand, for high
frequency transducers, small variations have a direct influence on the characteristics of
the transducers. Therefore, a design method was selected for the impedance matching
network (IMN) using impedance analysis and the Smith chart to determine the
appropriate component values and topology of IMN instead of a design method based on
PiezoCAD modeling.
7
In this chapter, an L-type IMN for a large aperture high frequency transducer was
developed, which achieved a much sharper peak at the resonance frequency by
sacrificing bandwidth (Gonzalex, 1996), allowing a greater acoustic pulse at the resonant
frequency compared to that generated by an ultrasonic transducer without IMN. The L-
type IMN could be easy to tune the matching network with a reliable performance using
the combination of inductor and capacitor in the frequency range from kHz to GHz
depending on electrical impedance of the ultrasonic transducers compared to transmission
line IMN (Pozar, 1998). A detailed description of the design and optimization procedures
of an impedance matching network (IMN) for a large aperture high frequency transducer
is presented. The IMN design process follows five steps; (1) measuring maximum
magnitude of admittance (|Y|) and zero phase angle ( θ
z
) from impedance analysis and
estimating the resonance frequency that would be termed as the target center frequency of
the ultrasonic transducer; (2) determining the appropriate component values and topology
of IMN using the Smith chart based on the real (R) and the imaginary (X) values obtained
from the impedance analysis; (3) verification of the performance of the ultrasonic
transducer with and without IMN conducted by pulse-echo measurement; (4) optimizing
component values and topology of IMN by trial and error after comparing the target
center frequency and measured center frequency from pulse-echo measurement; (5)
determined component values and topology implemented on a Printed Circuit Board
(PCB), and integrated with the ultrasonic transducer. The optimized center frequency was
determined as a value similar to the target center frequency measured from impedance
analysis. At the same time, the optimized echo magnitude was set to a value at least 30%
8
bigger than the reference echo magnitude obtained by pulse-echo testing of the same
ultrasonic transducer without IMN. Under the same input parameters such as peak-to-
peak voltage (V
pp
) and treatment time (T
t
), the effects on the surface of a plastic petridish
and cell responses were investigated by using acoustic pressure field generated by
ultrasonic transducers with and without IMN.
2.2 Materials and Methods
2.2.1 Ultrasonic Transducers, Impedance Analyzer, Pulse-echo and Insertion
Loss Measurement System
Two ultrasonic transducers were designed, fabricated, and tested. Lithium niobate
(LiNbO3) single element ultrasonic transducers were fabricated with conventional
approaches (Lam et al., 2013). The aperture sizes of the first (TR1) and the second (TR2)
ultrasonic transducers were 4.3 mm and 2.6 mm, respectively. The f
number
of TR1 and
TR2 were 1.23 and 0.75, respectively. The designed center frequencies of TR1 and TR2
were 110 MHz and 150 MHz, respectively. The characteristics of the ultrasonic
transducer were measured by an impedance analyzer (Agilent E4991, Agilent
Technologies, Santa Clara, CA) which has an operating frequency range from 1 MHz to 3
GHz. The magnitude of electrical impedance (|Z|) and admittance (|Y|), phase angle ( θ
z
),
real (R) and imaginary (X) values of the electrical characteristics were measured.
9
To verify the performance of the impedance matching network (IMN), pulse-echo
and insertion loss (IL) measurement of the ultrasonic transducers with and without IMN
was conducted (Lam et al., 2013). For the pulse-echo measurement, an imaging system
Figure 2.1 Experimental setups for (a) pulse-echo and (b) insertion loss (IL) measurement.
was developed as shown in Figure 2.1(a). It was comprised of a pulser/receiver (5900PR,
Olympus NDT Inc., Waltham, MA), 12-bit analog to digital converter (ADC) with up to
2 GS/s sampling (CS122G1, Dynamics Signals LLC., Lockport, IL), a 3D linear
translation/rotation stage (ILS100HA, Newport, Irvine, CA), motion controller (ESP301-
3N, Newport, Irvine, CA), and a custom-built MATLAB (MathWorks Inc., Natick, MA)
program. An ultrasonic transducer with and without IMN was excited by the
pulser/receiver with 200 Hz pulse repetition frequency and the focal depth of the
ultrasonic transducer was adjusted by using a 3D linear translation/rotation stage with a
motion controller and a customized MATLAB program. The reflected echo signals from
a quartz target which was placed at the focal point were amplified by the pulser/receiver.
(a)
(b)
10
Received signals were then digitized with ADC at 1 GHz sampling rate. Digitized signals
were transferred to the customized MATLAB program to see pulse-echo responses. The
pulse-echo measurements were performed in the degassed/deionized water.
For the insertion loss (IL) measurement as presented in Figure 2.1(b), a sine burst
signal of 5V with 30 cycles produced by a function generator (AFG 3251, Tektronix,
Beaverton, Oregon) was used to excite ultrasonic transducers with and without IMN, and
the reflected echo signal from a quartz target placed at the focal distance was recorded in
an oscilloscope (TDS 5052, Tektronix, Beaverton, Oregon) set to 1 MΩ coupling. IL was
measured by voltage ratio of the echo signal to the sine burst signal, expressed in decibels
(dB) over a range of -6 dB bandwidth (BW). The measured value was then compensated
for the attenuation in water (0.0002 dB/mm/MHz
2
) (Cannata et al., 2001) and reflection
from the quartz target (1.9 dB).
2.2.2 Design of Impedance Matching Network
An impedance matching network (IMN) was used to transform the real value and
cancel the imaginary value of the input electrical impedance of an ultrasonic transducer.
The real value (R2) measured by impedance analysis was compared to the electrical
impedance of the excitation source (R1), usually 50 Ω, in order to determine whether
IMN was placed in shunt with R2 (lowers R2) or in series with R2 (boosts R2) as shown
in Figure 2.2(a) and 2.2(b), respectively. Once IMN was determined in shunt or in series
with R2, the component values and topology such as capacitor (C)/inductor (L), C/C, L/L
11
and L/C were selected by using the Smith chart, based on measured real and imaginary
values of the electrical impedance of the ultrasonic transducer. A series-added
Figure 2.2 Schematic diagram of impedance matching network (IMN) for an ultrasonic
transducer. Zin represents input impedance seen from the excitation source and should be
equal to the complex conjugate of the Z1. Z1 and Z2 indicate the electrical impedance of
the excitation source and ultrasonic transducer, respectively. (a) When R1 < R2, IMN is
(a) (b)
(c) (d)
12
placed in shunt with R2 to lower R2 (b) When R1 > R2, IMN is placed in series with R2
to boost R2. The component values and topology such as capacitor (C)/inductor (L), C/C,
L/L and L/C were dependent upon measured real and imaginary values of electrical
impedance of the ultrasonic transducer at the target center frequency on a Smith chart. (c)
The example of IMN design procedures. The input electrical impedance of the ultrasonic
transducer moved toward the center of the Smith Chart by proper inductor and capacitor
component values and topology at the target center frequency. (d) Picture of IMN with
ultrasonic transducer 1. The printed circuit board (PCB) was implemented with shunt-
added capacitor of 100 pF and series-added inductor of 8 nH, and an overall size was 21
by 11 mm
2
.
inductor (L) moved the impedance in an upward direction, which led to a clockwise
rotation along a constant circle of resistance, while a series-added capacitor (C) moved
the impedance in the downward direction, which led to a counter clockwise rotation
along a constant circle of resistance. A shunt-added inductor (L) moved the impedance
upward, which led to a counter clockwise rotation along a constant circle of conductance,
while a shunt-added capacitor (C) moved the impedance downward, which led to a
clockwise rotation along a constant circle of conductance (Pozar, 1998; Rhea, 2006). For
example, as presented in Figure 2(c), when measured real and imaginary values of
electrical impedance were 2.5 Ω and -5.5 Ω at target center frequency (TCF) of 130 MHz,
respectively, by selecting series-added inductor (L) of 20 nH, electrical impedance was
moved to a clockwise rotation along a constant circle of resistance, and real and
imaginary values were 2.5 Ω and 11.5 Ω at TCF, respectively. The electrical impedance
was then transformed a clockwise rotation along a constant circle of conductance by
shunt-added capacitor (C) of 100 pF, and real and imaginary values were 50 Ω and -0.2 Ω
at TCF, respectively. As a result, the input electrical impedance of the ultrasonic
13
transducer moved toward the center of the Smith Chart by selecting appropriate
component values and topology at the target center frequency.
The target center frequency of the ultrasonic transducer with impedance matching
network (IMN) was selected because it was the frequency where the magnitude of
electrical admittance (|Y|) and the phase angle ( θ
z
) of the transducer was maximum and
zero, respectively (Kiamg et al., 2014). The design criterion was that the echo magnitude
of the ultrasonic transducer with impedance matching network (IMN) at the target
frequency should be at least 30% greater than the reference echo magnitude, acquired
from the pulse-echo test using the ultrasound transducer without IMN. When the design
criterion was not satisfied, the topology and component values of IMN should be
optimized according to the optimization approach. For the optimization approach, a
topology selection should be taken into consideration of availability of components and
frequency response related to applications; high-pass IMN (series-added capacitor/shunt-
added inductor) for blocking low frequency oscillating signals, and low-pass IMN
(series-added inductor/shunt-added capacitor) for filtering harmonic signals (Huang et al.,
2008; Rogers et al., 2003). Also, specifying component values was taken account of the
target center frequency sweep (20% bandwidth) and component values with a given
tolerance (Ilumoka et al., 1998). Since electrical impedance is changed as a function of
frequency on the Smith chart and initially determined component values does not take
account of the given tolerances, the component values and topology should be adjusted
by sweeping the target center frequency (20% bandwidth) and its tolerance ranges. The
14
optimization process of the topology and component values was repeated until the
criterion was satisfied.
The printed circuit board (PCB) was implemented on 0.062" fiberglass epoxy
laminate (FR-4) with a loss tangent of 0.0153 and dielectric constant of 4.5, with an
overall size of 21 by 11 mm
2
. IMN was made using the optimized component values and
topology, and then was connected to the ultrasonic transducer as shown in Figure 2.2(d).
2.2.3 Validation of Impedance Matching Network
Figure 2.3 illustrates the experimental arrangement which has been used for high
frequency ultrasound microbeam experiments. The experimental setup was composed of
transducer 1 (TR1) with and without IMN, a pulser/receiver, an oscilloscope (LC534,
LeCroy, Chestnut Ridge, NY), a 3D linear translation/rotation stage, a customized
LabVIEW (National Instruments, Austin, TX) program for focusing TR1, the function
generator and a 50 dB power amplifier (525LA, ENI, Rochester, NY) for the generation
of an acoustic pulse using TR1. An Epi-microscope (IX71, Olympus Corporation of the
Americas, Center Valley, PA) was integrated with the ultrasound system for cell imaging.
The ultrasonic transducer 1 (TR1) was accurately controlled by a 3D linear
translation/rotation stage to locate TR1’s focal point at the desired position by finding the
optimized echo magnitude. The focal point was positioned at the center of the
microscope’s field of view. Once focusing of TR1 with IMN was completed, TR1 with
IMN was excited by the treatment conditions of peak-to-peak voltage (V
pp
) and treatment
time (T
t
) as shown in Figure 2.3. T
t
was determined by the number of cycles of sine
15
waves with the optimized center frequency, which was found during the optimization
process of TR1 with IMN.
An acoustic pulse, generated by TR1 with and without IMN, with peak to peak
voltage (V
pp
) ranging from 40 V to 70 V and treatment time (T
t
) of 420 μs was applied on
the surface of a 35 mm plastic petridish. The effects of the applied acoustic pulse were
recorded by taking still shot images using the epi-fluorescence microscope.
Experiments on human cervical cancer cells (HeLa) due to the acoustic pulse
were further conducted. HeLa cells were grown in eagle's minimum essential medium
(EMEM) supplemented with 10% fetal bovine serum (FBS). The cells were plated on a
35 mm petridish and incubated for 2 days in 5% CO
2
at 37˚C. An acoustic pulse was
applied on cells with the treatment conditions of V
pp
of 55 V and T
t
of 420 μs.
Immediately after the treatment, cells were incubated with EMEM in 5% CO
2
at 37˚C for
4 hours. Cell viability was conducted with a LIVE/DEAD Cell Imaging kit (Life
Technologies Corp., Carlsbad, CA) according to the manufacturer’s instructions. Bright-
field and fluorescence images were taken with the epi-fluorescence microscope. In the
case of the live/dead assay, stained live cells have bright green fluorescence, while
stained dead cells have intense red fluorescence. In order to compare the performance of
the IMN, the experiment was repeated with the same ultrasonic transducer without IMN
under the same conditions and input parameters.
16
Figure 2.3 Experimental setup for the validation of impedance matching network (IMN)
and cell applications. Precise focal depth of the ultrasonic transducer 1 (TR1) with IMN
was adjusted by moving TR1, attached to a 3D linear translation/rotation stage, while
finding the optimized echo magnitude using a pulser/receiver and an oscilloscope. After
focusing TR1 with IMN, an acoustic pulse with peak-to-peak voltage (V
pp
) and treatment
time (T
t
) was applied to the surface of a plastic petridish and HeLa cells. Tt was
composed of the number of sine waves with the optimized center frequency of TR1. The
test and experiment were carried out comparing tangible effects obtained by TR1 with
and without IMN.
2.3 Results
2.3.1 Impedance Analysis Results
Figures 2.4(a) and 2.4(b) represent magnitude measurements of electrical
admittance (|Y|), phase angle ( θ
z
), real (R) and imaginary (X) values of the electrical
impedance of ultrasonic transducer 1 (TR1) and ultrasonic transducer 2 (TR2) based on
17
Figure 2.4 Measured magnitude of electrical admittance (|Y|), phase angle on the left
column ( θ
z
), real (R) and imaginary (X) values on the right column of electrical
impedance of (a) transducer 1 (TR1) and (b) transducer 2 (TR2). The target center
frequency (TCF) of TR1 and TR2 were measured as 105 MHz and 150 MHz,
respectively, when |Y| was maximum value and θ
z
was zero as indicated with solid
arrows in left column in (a) and (b). At the target center frequency, measured R and X
were 1.5 Ω and 0.2 Ω for TR1 and 2.5 Ω and 0.1 Ω for TR2, respectively.
impedance analysis, respectively. Left column in Figure 2.4 indicates |Y| and θ
z
measurements and right column shows R and X values of electrical impedance of TR1
and TR2. From |Y| and θ
z
measurements, the target center frequencies of TR1 and TR2
were estimated 105 MHz and 150 MHz, respectively. These target center frequencies are
denoted as solid arrows in Figures 2.4(a) and 2.4(b). At these target center
(a)
(b)
18
Figure 2.5 Pulse-echo waveform and echo spectrum of (a) ultrasonic transducer 1 (TR1)
and (b) ultrasonic transducer 2 (TR2). Left column represents the pulse-echo
measurement without impedance matching network (IMN) and right column shows the
pulse-echo measurement with IMN. For TR1, the optimized center frequency (OCF) was
almost the same as the target center frequency (TCF, 105 MHz) and the optimized echo
magnitude (OEM) in the right column of (a) was increased by 65% compared to the
reference echo magnitude (REM) in the left column of (a). For TR2, the optimized center
frequency (OCF) was similar to the target center frequency (TCF, 150 MHz), measured
by impedance analysis and the optimized echo magnitude (OEM) in the right column of
(b) was enhanced by 33% compared to the reference echo magnitude (REM) in the left
column of (b).
frequencies, measured R and X values are 1.5 Ω and 0.2 Ω for TR1 and 2.5 Ω and 0.1 Ω
for TR2, respectively. Since the real value of the two ultrasonic transducers at the target
center frequency was less than 50 Ω, impedance matching network (IMN) was placed in
(a)
(b)
19
series with the ultrasonic transducers to boost the real value of impedance of the two
ultrasonic transducers.
2.3.2 Optimization of Impedance Matching Network
The measured pulse-echo waveform and echo spectrum of ultrasonic transducer 1
(TR1) and ultrasonic transducer 2 (TR2) are presented in Figure 2.5(a) and 2.5(b),
respectively. Left column and right column in Figure 2.5(a) and 2.5(b) indicate the pulse-
echo measurements without impedance matching network (IMN) and with IMN,
respectively. Table 2.1 summarizes the optimization process of IMN.
Table 2.1 Summary of the optimization of impedance matching network (IMN) for
ultrasonic transducer 1 (TR1) and ultrasonic transducer 2 (TR2)
Impedance matching network
(IMN)
Transducer 1 (TR1) Transducer 2 (TR2)
without
IMN
# of Optimization without
IMN
# of Optimization
1
st
2
nd
… Final 1
st
2
nd
… Final
Target center frequency (MHz) 105 - - … - 150 - - … -
Reference center frequency (MHz) 113 - - … - 120 - - … -
Optimized center frequency (MHz) - 86 114 … 111 - 163 114 … 140
Reference echo magnitude (V) 2.0 - - … - 1.2 - - … -
Optimized echo magnitude (V) - 2.8 0.6 … 3.3 - 0.4 1.0 … 1.6
Inductor (nH) - 24 14 … 8 - 8 9 … 47
Capacitor (pF) - 146 740 … 100 - 45 90 … 25
Topology -
…
-
…
20
Left column in Figure 2.5(a) shows the reference center frequency and the
reference echo magnitude that were measured as 113 MHz and 2.0 V, respectively. For
TR1 with IMN, the combination of 146 pF/24 nH and topology of capacitor/inductor
were initially determined and tested. The center frequency of the pulse-echo response
was measured at 86 MHz, whereas echo magnitude was increased from 2.0 V to 2.8 V
compared to TR1 without IMN. Since the center frequency was too low, the second
optimization process was conducted with updated component values of 14 nH/740 pF and
topology of inductor/capacitor. The center frequency was 114 MHz, but the received
echo magnitude was decreased to 0.6 V compared to TR1 without IMN. After the
optimization process was repeated, the optimized component values and topology of IMN
were 100 pF/8 nH and capacitor/inductor, respectively. The optimized echo magnitude of
TR1 with IMN was 65% greater than TR1 without IMN while the optimized center
frequency of TR1 with IMN was 111 MHz which was similar to the target center
frequency (105 MHz) measured by an impedance analysis as shown in the right column
of Figure 2.5(a).
Left column in Figure 2.5(b) indicates the reference center frequency and the
reference echo magnitude which were measured as 120 MHz and 1.2 V from pulse-echo
testing, respectively. For TR2 with IMN, the combination of 8 nH/45 pF and topology of
inductor/capacitor were initially selected by using the Smith chart. The center frequency
of the pulse-echo measurement was improved to 163 MHz, whereas echo magnitude was
decreased from 1.2 V to 0.4 V. Since the echo magnitude was too low, it was necessary
for further optimization. A second optimization process was conducted with updated
21
component values of 90 pF/9 nH and topology of capacitor/inductor. The center
frequency and the received echo magnitude were simultaneously decreased. After the
optimization process was repeated, the optimized component values and topology of IMN
were 47 nH/25 pF and inductor/capacitor, respectively. The optimized center frequency
of TR2 with IMN was 140 MHz which was similar to the target center frequency (150
MHz) measured by impedance analysis, and the optimized echo magnitude of TR2 with
IMN was improved by 33% compared to the reference echo magnitude as shown in the
right column of Figure 2.5(b).
2.3.3 Electrical Characteristics of Optimized Impedance Matching Network
Figure 2.6(a)-(c) indicates electrical performance of two ultrasonic transducers
without and with optimized impedance matching network (IMN). The electrical
properties of the cable are 6 inch (153 mm) length, 50 Ω impedance, insertion loss of
0.05 dB and return loss of 37 dB in the frequency range from 100 MHz to 200 MHz. Left
column in Figure 2.6(a) and 2.6(b) represents |Z| and θ
z
measurements that are 35 Ω and
85˚ for TR1 without IMN, and 32 Ω and 77˚ for TR2 without IMN at the reference center
frequency (RCF), respectively. Right column in Figure 2.6(a) and 2.6(b) shows |Z| and θ
z
measurements of 65 Ω and 76˚ for TR1 with IMN, and 52 Ω and -40˚ for TR2 with IMN
at the optimized center frequency (OCF), respectively. After compensation for the
attenuation caused by water and reflection from the quartz target, the insertion loss (IL)
values over a range of -6 dB bandwidth (BW) are shown in Figure 2.6(c). Left column in
Figure 2.6(c) indicates IL is calculated as -21.5 dB for TR1 without IMN at the RCF
22
Figure 2.6 Measured electrical performance of ultrasonic transducers without and with
impedance matching network (IMN). (a) Magnitude of electrical impedance (|Z|), (b)
phase angle ( θ
z
) and (c) insertion loss (IL) of ultrasonic transducer 1 (TR1) without and
with IMN and ultrasonic transducer 2 (TR2) without and with IMN, respectively. For
TR1, |Z|, ( θ
z
) and IL were measured as 35 Ω, 85˚ and -21.5 dB at the reference center
frequency (RCF) without IMN, and 65 Ω, 76˚ and -13 dB at the optimized center
frequency (OCF) with IMN, respectively. For TR2, |Z|, ( θ
z
) and IL were 32 Ω, 77˚ and -
34 dB at the RCF without IMN, and 52 Ω, -40˚ and -29 dB at the OCF with IMN,
respectively.
(a)
(b)
(c)
23
and -13 dB for TR1 with IMN at the OCF, respectively. Right column in Figure 2.6(c)
shows IL is measured as -34 dB for TR2 without IMN at the RCF and -29 dB for TR2
with IMN at the OCF, respectively.
2.3.4 Performance Validation for Cellular Applications
Figure 2.7(a) demonstrates noticeable circular dents, produced by ultrasonic
transducer 1 (TR1) with impedance matching network (IMN). In contrast, the acoustic
pulse generated by TR1 without IMN did not produce recognizable circular dents as
shown in Figure 2.7(b). The locations of the acoustic pulse are indicated as dashed circles
in Figure 2.7(a)-(b). The scale bars indicate 20 µm.
Figure 2.7 Circular dents on a 35 mm plastic petridish generated by acoustic pulses from
ultrasonic transducer 1 (TR1) (a) with impedance matching network (IMN) (b) without
IMN. Acoustic pulses were generated by the peak-to-peak voltage (V
pp
) from 40 V (left)
to 70 V (right) with increments of 10 V and treatment time (T
t
) of 420 μs. The scale bar
indicates 20 µm.
Bright-field and fluorescence images of a HeLa cell treated by ultrasonic
transducer 1 (TR1) with impedance matching network (IMN) and without IMN are
(b)
(a)
24
shown in Figure 2.8(a)-(c) and Figure 2.8(d)-(f), respectively. Figure 2.8(a) shows a
circular dent which was found in the middle of a HeLa cell right after one acoustic pulse
was applied. The cell treated by TR1 with IMN was dead after 4 hours of the treatment
because red fluorescence from propidium iodide was clearly seen in Figure 2.8(c). On the
other hand, the cell treated by TR1 without IMN was alive after 4 hours of the treatment
because green fluorescence from calcein dye in the cell was observed as shown in Figure
2.8(e).
Figure 2.8 Cell responses to an acoustic pulse generated by ultrasonic transducer 1 (TR1)
with impedance matching network (IMN) and TR1 without IMN. Peak-to-peak voltage
(V
pp
) and treatment time (T
t
) were 55 V and 420 μs, respectively. Arrows indicate the
area where the acoustic pulse was applied. The scale bars indicate 20 µm. (a)-(c) TR1
with IMN (a) Bright-field image of a HeLa cell was acquired right after treatment. The
circular dent in the middle of the cell was generated by the acoustic pulse. (b)
Fluorescence image of the same cell stained with calcein was captured. (c) Fluorescence
image of the same cell stained with propidium iodide (PI) was acquired. After 4 hours of
incubation after treatment, the cell treated by TR1 with IMN was dead. (d)-(f) TR1
without IMN (d) Bright-field image of a HeLa cell was taken right after treatment. (e)
Fluorescence image of the same cell stained with calcein was taken. (f) Fluorescence
(a) (b) (c)
(d) (e)
(f)
25
image of the same cell stained with PI was captured. After 4 hours of incubation after
treatment, the cell exposed to the treatment was alive.
2.4 Discussions
To demonstrate the effectiveness of an impedance matching network (IMN) in
improving energy transfer of a large aperture high frequency ultrasonic transducer, two
ultrasonic transducers were tested. After optimization of IMN, the optimized center
frequency was similar to the target center frequency measured by an impedance analysis.
At the same time, the optimized echo magnitude was improved by at least 30% compared
to the reference echo magnitude. For ultrasonic transducer 1 (TR1), the optimized center
frequency of TR1 with IMN was 111 MHz, which was similar to the target center
frequency, and the optimized echo magnitude of TR1 with IMN was increased from 2 V
to 3.3 V (65%). Insertion loss was measured as -21.5 dB for TR1 without IMN at the
reference center frequency, -13 dB for TR1 with IMN at the target and optimized center
frequency, respectively. For ultrasonic transducer 2 (TR2), the optimized center
frequency of TR2 with IMN was almost the same as the target center frequency, and the
optimized echo magnitude of TR2 with IMN was improved from 1.2 V to 1.6 V (33%).
Insertion loss was -34 dB for TR2 without IMN at the reference center frequency, -18 dB
and -29 dB for TR2 with IMN at the target and optimized center frequency, respectively.
These measured insertion loss results demonstrate a great improvement on sensitivity,
which is essential for high frequency ultrasonic transducers.
26
There were some remarkable differences between simulation results obtained by
using the Smith chart and experimental results acquired by pulse-echo measurement.
Most notably, non-ideal properties such as equivalent series resistance (ESR) were not
taken into consideration in simulation. However, ESR is a very important characteristic
and should be considered in practice. As the frequency increases, non-ideal inductor and
capacitor values were simply modeled as an ideal inductor and capacitor in series with
ESR (Li et al., 1998; Piwowarska et al., 2006). Additionally, circuit board performance
such as trace characteristics might be attributed to the experimental results, which was
not taken into account for the simulation. Moreover, even if the above mentioned
characteristics were taken into consideration in simulation, designed component values
might be slightly different from the commercially available component values with some
tolerance. Therefore, the repeated optimization process was required to obtain the desired
values of optimized center frequency and optimized echo magnitude.
The tested ultrasonic transducer 1 (TR1) with IMN was able to generate clear
circular dents on the plastic petridish; however, TR1 without IMN did not produce
recognizable circular dents with the same peak-to-peak voltage (V
pp
) and treatment time
(T
t
) as shown in Figure 2.7. These experimental results indicate that the ultrasonic
transducer with IMN can effectively generate a stronger pressure field at its focal area
without damaging the ultrasonic transducer. An acoustic pulse generated by the ultrasonic
transducer with IMN under lower input voltage and/or lower PRF will be capable of
inducing similar effects to the ultrasonic transducer without IMN. Therefore, cell
27
responses could be controlled by adjusting treatment conditions, i.e., peak-to-peak
voltage (V
pp
) and treatment time (T
t
), according to the purposes of the applications.
28
CHAPTER 3 High Frequency All-in-one Front-end System for
Multifunctional High Frequency Ultrasound Imaging
3.1 Introduction
The application of ultrasound imaging to biological samples having micro scale
structures requires an imaging system at higher resolution than conventional clinical
ultrasound imaging (Liang et al., 2003). For example, although the recovery of the nerve
bundle was recently identified in the longitudinal study, the recovery of each axon could
not be figured out due to the limited resolution (Kuffler., 2010). For studying the axon
regeneration, the higher resolution is required to detect the axon fiber. A lot of efforts
have been made to enhance the spatial resolution with the higher imaging frequency for
visualizing either cellular structures or their mechanical properties because the spatial
resolution is proportional to the imaging frequency (Kim et al., 2013). Typically, the
combination of benchtop function generator and power amplifier has been utilized in the
frequency range over 100 MHz; however, the approach suffers from the critical noise
issue coming from the power amplifier and overall bulky. To address the above problems,
a custom-built all-in-one front-end system performing the frequency range higher than
100 MHz was designed and developed. The system could generate a frequency tunable
monocycle pulse with center frequencies of 120 MHz, 250 MHz, and 455 MHz, the
bandwidths of 68-182 MHz, 170-330 MHz, and 310-600 MHz, and peak-to-peak
amplitudes of 70 V
pp
, 60 V
pp
, and 30 V
pp
respectively. Also, a couple of the pulse
29
generators could be combined and generate multicycle pulses and pseudo phase
modulated pulses, which can provides better visualization of cross-sectional imaging.
3.2 Materials and Methods
3.2.1 Design of the Front-end System
The proposed system consists of a pulse forming network and an amplifier
utilizing radio frequency Laterally Diffused Metal Oxide Semiconductor (RF LDMOS)
technology. Figure 3.1 shows the schematics and descriptions of the signal at each stage
from A to F.
The pulse forming network included a timing circuit (R2, C3), a speed-up
capacitor (C2), and switching bipolar transistor (BJT1). When a trigger signal consisting
of a 5 V rectangular pulse was received at the input port, the voltage at the base of BJT2
(NE46100, California Eastern Laboratories, CA) reached the threshold level to turn on.
For sharpening the edges of the pulse generated, a speed up capacitor C2 was located
before the base of BJT2. Most transistors have a delay time in turning on and storage
time in shutting off. These two parameters can be adjusted to minimize the transitional
time of BJT2 and eventually the pulse width. The storage time depends on the amount of
current injected into the base of BJT2 and the additional current to maintain the saturation
condition. The delay time can be shortened by draining the base current quickly with a
strong reverse bias. The network of R1 and C2 accomplished the above described
30
purpose. During the voltage rise, R1 was shorted by C2, minimizing the time to reach 5 V
by injecting the current into the base of BJT2 and reducing the supply current for holding
the voltage.
Figure 3.1 Proposed pulse generator (a) Schematics of pulse generator (b) Output wave
forms at the blue dotted position from A to F and the unit of the y-axis is Volts.
C2 was then shorted out during the falling edge of the input signal. The values of R1 and
C2 were set to 100 Ω and 47 pF by assuming the rise time needs to be at least two times
(a)
(b)
31
faster than the targeted output frequency of 100 MHz. When the trigger pulse reached
point A in Figure 3.1, the capacitor in the timing network (C3) started to be charged.
Once the voltage on C3 approached the threshold voltage for turning on BJT1 (NE46100,
California Eastern Laboratories, CA), BJT1 drained the current from point C in Figure
3.1 to ground, and BJT2 was shut off. Therefore, the combination of the timing circuit
and speed-up capacitor allows the generation of a short pulse out of the low frequency
trigger signal. The on time of BJT1 was dependent on the RC constant shown in the
equation below and could be controlled by changing the resistance value of R2 (Oldham
et al., 2004).
V
C3
= V
B
∙ (1 - 𝑒 −
𝑡𝑖𝑚𝑒 𝑅 2
∙𝐶 3
) (3.1)
where V
C3
is the base voltage, V
B
is the voltage at point B in Figure 3.1 and time is
the on-time of BJT1. The values of R2 and C3 for generating the output pulse over 100
MHz were 500 Ω and 4 pF. The following stage of the proposed system was an amplifier
utilizing RF LDMOS (BLF571, NXP Semiconductors, Eindhoven, Netherlands) which
has a narrow channel length and a heavily doped n-type region for obtaining high output
current and minimal on-resistance (Sung et al., 2011). The output of the pulse forming
network at point D in Figure 3.1 was a pulse having a negative voltage peak which turns
on MOSFET1. The output of the amplifier was frequency tuned by using a high pass
filter formed by L1 and C5 set to 100 nH and 18 pF for the 100 MHz pulse output. The
32
output impedance of the amplifier was matched with 50 Ω by placing a 50 Ohm resistor
in parallel.
3.2.2 Fabrication of the System and Measurement Setup
The proposed system was implemented on a 0.062" FR-4 substrate with a
dielectric constant of 4.5 and its overall size was 90 × 55 mm
2
. The measured pulses at
the output of the designed system were recorded with a digital phosphor oscilloscope
(TDS 5052, Tektronix, Beaverton, Oregon).
The trigger pulse for the input to the pulse generator was generated by a function
generator (AFG 3252, Tektronix, Beaverton, OR), and the pulse repetition frequency was
20 kHz. For the application to ultrasound imaging, a linear mechanical scanner called a
biomicroscope with a 140 MHz single element lithium niobate single element transducer
having bandwidth of 40 % and f
number
of 0.75 was used (Yoon et al., 2015). The
transducer, a receive amplifier (5900PR, Olympus NDT Inc., Waltham, MA) with 37 dB
gain, and the developed system were configured as the T network shown in Figure 3.2.
The surface of the transducer was submerged into de-ionized water and translated over
the 2.5 µm tungsten wires placed at the focal distance of 1.9 mm.
33
Figure 3.2 High frequency mechanical ultrasound system setup.
3.3 Results
3.3.1 Electrical characteristics of the System
Figure 3.3(a) shows the output pulse waveforms and Figure 3.3(b) shows
frequency responses in the frequency domain. The center frequencies of the provided
signals were 125 MHz, 250 MHz, and 455 MHz and the -6 dB bandwidths were 68-182
MHz, 170-330 MHz, and 310-600 MHz with peak-to-peak voltage levels of 70 V
pp
, 60
V
pp
and 30 V
pp
, respectively. The pulse repetition frequency (PRF) could be increased up
to 10 MHz and 6 W output power was estimated when generating a 100 MHz monocycle
with 70 V
pp
. Two cycled pulses could be generated by combining two identical pulse
generators. Figure 3.3(c) shows two cycle bipolar pulses with peak-to-peak amplitude of
30 V
pp
, center frequency of 285 MHz, and -6 dB bandwidth of 220-354 MHz.
34
Figure 3.3 Measured output pulses from the proposed transmitter (a) Output pulses in the
time domain (b) Output pulses in the frequency domain (c) Two cycle pulse train from
the combined pulse generators.
3.3.2 Performance Evaluation for Ultrasound Applications
Figure 3.4 shows the wire target measurement results. The measured axial and
lateral resolutions were 17 µm and 7 µm, respectively which were close to the theoretical
values of 22 µm and 7 µm.
Figure 3.4 Wire target study results (a) B-mode image and scale of both width and depth
is mm (b) Axial brightness profile and scale of x-axis is µm and y-axis is dB (c) Lateral
brightness profile and scale of x-axis is µm and y-axis is dB.
3.4 Conclusion
A frequency tunable front-end system for high frequency (>100MHz) ultrasound
imaging was presented. The implemented pulse generator was integrated with a
(a)
(b)
(c)
(a) (b) (c)
35
biomicroscope to assess its resolution which was shown to be comparable to cell size.
Also, it was shown that it was possible to generate high pulse repetition pulse trains by
combining two pulse generators. The number of cycles could be increased by using
additional pulse generators.
(a) (b)
36
CHAPTER 4 Acoustic Transfection with High Frequency
Ultrasound
4.1 Introduction
The intracellular delivery of cell membrane-impermeable molecules into the
cytoplasm of cells has been one of the most attractive areas of research in biomedicine
(Chen 2010; Glover et al., 2005; Kim et al., 2009; Leader et al., 2008; Shi et al., 2015).
Controlling cellular functions and structures through delivering exogenous therapeutic or
genetic materials into cells enables the exploration for visualization of specific cellular
structures (Herce et al., 2013), analysis of cell mechanisms (Weill et al., 2008), and
treatment of genetic diseases (Nabel et al., 1990). In these applications, investigations
into unique expression profiles between individual cells at the single-cell level have had
an increased emphasis on elucidation of potentially undetected particular cellular
functions and structures of individual cells, and cell-to-cell interaction within a cell
population (Chattopadhyay et al. 2014; Hoppe et al. 2014; Wang et al. 2009). A range of
techniques such as virus-mediated transfection (Tanaka et al., 1997), lipid-mediated
transfection (Felgner et al., 1987), microinjection (Meacham et al., 2014; Mehier-
Humbert et al., 2005), electroporation (Escoffre et al., 2007; Mehier-Humbert et al., 2005;
Palanker et al., 2006), and optical transfection (Mehier-Humbert et al., 2005; Watanabe et
al., 2007; Zeira et al., 2006) to deliver foreign molecules into cells has been developed.
Virus-mediated transfection is highly efficient, while viral vectors possess a limited
packaging capacity and sometimes cannot be specifically integrated into the target cell.
37
Lipid-mediated transfection has relatively low toxicity; however, transfection efficiency
largely relies on cell types and culture conditions. Microinjection is straightforward and
efficient, while this technique requires direct perforation of cell membranes, which may
lead to physical damage to the cells. Electroporation and optical transfection have
relatively high transfection efficiency; however, these methods may cause irreversible
membrane damage under the influence of electrical fields and shorter wavelengths of
laser radiation, respectively. Therefore, there is still a need for the development of
approaches which are capable of simultaneously satisfying the requirements of high
transfection efficiency, minimal cytotoxicity, and single-cell selectivity independent of
cell types and transfer molecules (Antkowial et al., 2013; Kawamura et al., 2009; Kim et
al., 2010; Kollmannsperger et al., 2016; Nishikawa et al., 2008; Sharei et al., 2013).
Acoustic transfection with high frequency ultrasound was developed in our
laboratory as a new transfection method for delivering membrane-impermeable
molecules into the cytoplasm of cells at ultrasound frequencies higher than 150 MHz
(Yoon et al., 2015; Yoon et al., 2016a; Yoon et al., 2016b). A tightly focused high
frequency ultrasound beam physically stretches cell lipid bilayer on plasma membrane,
which generates transient and reversible holes on cell membranes. The focused
ultrasound beam was on only for a single ultrasound pulse of a few microseconds within
an area of approximately 10 μm in diameter, which might result in the formation of a
physical pathway for introducing membrane-impermeable molecules into cell cytoplasm
and passive diffusion driven by the concentration gradient across transiently generated
holes. An impedance matching network was also developed to optimize excitation
38
frequency of transducers (Kim et al., 2016). In this chapter, the concept of acoustic-
transfection technique was demonstrated by the intracellular delivery of propidium iodide
(PI) molecule, and 3 kDa dextran labeled with Alexa 488 into HeLa cells at single-cell
level. The intracellular delivery was confirmed by live-cell fluorescence imaging of the
time-based intensity changes of delivered molecules.
Additional experiments were necessary to determine the optimal acoustic
transfection conditions, which enables the acoustic transfection method to be a versatile
and viable transfection tool. Optimal treatment conditions for the acoustic transfection
technique across different cell types and molecules were necessary to achieve high
delivery efficiency and high cell membrane permeability with minimizing membrane
disruption. A detailed description of how optimal treatment conditions were determined
towards achieving high delivery efficiency with low cytotoxicity by studying treatment
conditions of acoustic transfection using high frequency ultrasound on four human cancer
cell lines; human cervical cancer cell (HeLa), Michigan cancer foundation-10A (MCF-
10A), Michigan cancer foundation-7 (MCF-7), and M.D. Anderson-metastatic breast 231
(MDA-MB231). To find optimal treatment conditions, a criterion termed "intracellular
delivery score (IDS)" was proposed and calculated from results—obtained on four human
cancer cell lines—of (1) the delivery efficiency, (2) a cell membrane permeability
study—which was measured by fluorescence intensity of PI after acoustic transfection on
target cells—and (3) a cell viability study after 4 and 20 hours of acoustic transfection
under different treatment conditions and a control condition. Adjustable parameters for
acoustic transfection were peak-to-peak voltage (V
pp
), treatment time (T
t
), and number of
39
cycles. Acoustic transfection was carried out at 6 different V
pp
at each 6 different T
t
with
1 cycle. Live-cell fluorescence imaging of acoustic transfected cells was performed. The
fluorescence intensity changes of PI inside treated cells under different treatment
conditions were compared to those of background fluorescence intensity to quantify and
calculate delivery efficiency and cell membrane permeability based on IDS. The cell
viability study—the results of which were needed to compute IDS—was performed with
a LIVE/DEAD Cell Imaging kit after 4 and 20 hours following the acoustic transfection.
The IDS was plotted with respect to six values of V
pp
at each of the six values of T
t
in
intracellular delivery graph (IDG) for easier viewing. The optimal treatment conditions
was chosen if IDS is larger than 9 points in IDG, which indicates high delivery efficiency
and high cell membrane permeability with minimum effects on cells. The intracellular
delivery of macromolecule and simultaneous intracellular delivery of two molecules with
optimal treatment conditions were successfully achieved.
4.2 Materials and Methods
4.2.1 Acoustic Transfection System and Controllable Treatment Conditions
The acoustic transfection system consists of fluorescence microscope and acoustic
pulse generation system as shown in Figure 4.1. A fluorescence microscope (Leica DMI
4000B, Germany) was used to monitor fluorescence intensity changes of the acoustic
40
transfected cells, and was integrated with the tightly focused high frequency ultrasonic
transducer with impedance matching network (IMN). A three dimensional linear
Figure 4.1 The acoustic transfection system for intracellular delivery of membrane-
impermeable molecules into the cells. Fluorescence microscope was used to monitor
fluorescence changes on the targeted cells, and acoustic pulse generation system was
utilized to precisely focus ultrasonic transducer with impedance matching network (IMN)
and generate an acoustic pulses with controllable treatment conditions such as peak-to-
peak voltage (V
pp
), treatment time (T
t
) and number of cycles (#).
translation/rotation stage controlled by a customized LabVIEW program was used to
accurately place the location of the transducer during the experiments. The focus of the
transducer was co-aligned with the focus of an objective lens of the microscope using a
pulser/receiver and an oscilloscope. Acoustic pulses were generated by a function
generator and a 50 dB power amplifier with controllable treatment conditions including
41
peak-to-peak voltage (V
pp
) and treatment time (T
t
). The values of V
pp
were 12V, 15V,
18V, 21V, 24V and 27V with 6 different T
t
of 6µs, 12µs, 18µs, 30µs, 60µs and 90µs with
1 cycle.
4.2.2 Cell Culture
HeLa, MCF-10A, MCF-7, MDA-MB231 cells were used in these studies. HeLa
cells were cultured in eagle's minimum essential medium (EMEM) supplemented with 10%
fetal bovine serum (FBS). MCF-7 and MDA-MB231 cells were cultured in dulbecco's
modified eagle's medium (DMEM) supplemented with 10% FBS and 1% penicillin
streptomycin. MCF-10A cells were cultured in mammary epithelial cell growth medium
(MEGM Bullet Kit, Lonza, Basel, Switzerland) supplemented with 100 ng/ml cholera
toxin (Sigma-Aldrich, St. Louis, MO). These cells were incubated in a humidified
atmosphere of 5% CO
2
at 37˚C and routinely sub-cultured in T25 vented-top culture
flasks from two to three times per week. Cells were seeded in 35 mm grid petridishes
(ibidi, Martinsried, Germany) with 2 mL of complete culture medium and incubated 36
hours in a humidified atmosphere.
4.2.3 Live-cell Fluorescence Imaging
The effects of applied acoustic pulses on cells were recorded by using microscope
imaging software (Leica LAS AF, Buffalo Grove, IL). The cell monolayer on the
prepared petridishes was gently washed twice with 2 ml of phosphate-buffered saline
42
(PBS), and 50 µM of the propidium iodide (PI) solution was added to the cell monolayer.
Live-cell fluorescence imaging was performed to quantify time-resolved PI fluorescence
intensity and compare the intensity between (1) before acoustic transfection which was
related to PI(0) at 0 second and (2) after 5 minutes of acoustic transfection which was
related to PI(∞) at steady-state. To extract quantitative information in a region of interest
(ROI) in an acoustic transfected single cell, imaging processing was performed from live-
cell fluorescence images by subtraction of a background region (ROB) next to ROI,
previously described in references (McCloy et al. 2014; Ophir et al. 2013). Then, to
directly check PI intensity comparison of individual cells among different human cancer
cell lines even if the cells did not share similar physical characteristics including cell size
the effects of physical characteristics of cell size were averaged. The number of acoustic
transfected cells at each treatment condition was more than 5.
PI
averaged intensity
=
ROI A
ROB F ROI A ROI ID ROB F ROI A ROI ID
_
))] 0 ( _ _ ) 0 ( _ ( )) ( _ _ ) ( _ [(
(4.1)
where Integrated density (ID) is the sum of the fluorescence intensity values of
selected regions, area (A) is the selected regions in square pixels, mean fluorescence ( F )
is the average fluorescence intensity value of selected regions, ID_ROI(∞) is the
integrated density in ROI at steady-state, ID_ROI(0) is the integrated density in ROI at 0
second, A_ROI is the area in ROI, F _ROB (∞) is the mean fluorescence in ROB at
steady-state, and F _ROB (0) is the mean fluorescence in ROB at 0 second. For the cell
viability study, the effects of treatment conditions and a control condition (0V / 0μs) on
four human cancer cell lines were systemically investigated. After acoustic pulses were
43
applied to the cells on the prepared petridishes, the monolayer was washed twice with 2
ml of PBS, and incubated with 2 ml fresh cell culture medium in a humidified
atmosphere for 4 and 20 hours. Before performing live-cell fluorescence imaging, the
cells were washed twice with 2 ml of PBS and stained with a LIVE/DEAD Cell Imaging
kit (Life Technologies Corp., Carlsbad, CA) according to the manufacturer’s instructions.
Numbers of acoustic-transfected cells at each treatment condition were more than 6.
4.2.4 Intracellular Delivery Score for Optimal Treatment Conditions
Table 4.1 gives the proposed criterion for intracellular delivery score (IDS) that
takes into consideration of delivery efficiency and cell membrane permeability in % out
of more than 185 cells for each cell line, and viability after 4 and 20 hours of acoustic-
transfection in % out of more than 222 cells for each cell line. The percentage of delivery
efficiency was calculated as the ratio of the number of delivered cells to the total number
of the acoustic transfected cells. The minimum PI intensity for counting the number of
delivered cells was 0.01 arbitrary units (a.u.) of the averaged PI intensity. The cell
membrane permeability was calculated and categorized according to the amount of the
averaged PI intensity. The percentage of cell viability was calculated as the ratio of the
number of live cells to the total number of the acoustic-transfected cells. The final IDS
was computed using a sum of the calculated values on the percentage of delivery
efficiency, cell membrane permeability, and cell viability according to the criterion
defined for the IDS. The intracellular delivery graph (IDG) was generated with respect to
different V
pp
at each of different T
t
to clearly observe the effect on cells of different
44
treatment conditions. The optimal treatment conditions were selected when IDS was
above 9 on IDG.
Table 4.1 Criterion for the intracellular delivery score (IDS)
Delivery efficiency Permeability Viability
90% ≤ D +5 15 ≤ P +5 90% ≤ V -0
70% ≤ D < 90% +4 10 ≤ P < 15 +4 70% ≤ V < 90% -2
50% ≤ D < 70% +3 5 ≤ P < 10 +3 50% ≤ V < 70% -4
30% ≤ D < 50% +2 1 ≤ P < 5 +2 30% ≤ V < 50% -8
10% ≤ D < 30% +1 0.01 ≤ P < 1 +1 10% ≤ V < 30% -9
D < 10% +0 P < 0.01 +0 V < 10% -10
4.2.5 Intracellular Delivery of Macromolecules and Simultaneous
Intracellular Delivery of Two Molecules
Now, the optimal treatment conditions was utilized for the intracellular delivery
of a 70 kDa dextran labeled with Oregon Green and simultaneous intracellular delivery of
two molecules (70 kDa dextran and PI) into four kinds of human cancer cell lines. 35 µM
of 70 kDa dextran solution for 70 kDa dextran was used for intracellular delivery and 35
µM of 70 kDa dextran solution and 50 µM of the PI solution were applied for
simultaneous intracellular delivery to the cell monolayer. Bright-field images were
45
acquired before the acoustic-transfection, and acoustic pulses were applied to the targeted
cells. After acoustic transfection, the monolayer was incubated in a humidified
atmosphere for 30 minutes, and the monolayer was washed twice with 2 ml of PBS, and
bright-field and fluorescence images were taken to track and observe the presence of
macromolecules in the acoustic transfected cells. The percentage of 70 kDa dextran
delivery efficiency was further calculated as the ratio of the number of delivered cells to
the total number of the acoustic transfected cells for each cell line under different optimal
treatment conditions. Numbers of acoustic transfected cells under each treatment
condition were more than 5.
4.3 Results
4.3.1 Time-based Intensity Changes of Propidium Iodide molecule
Figure 4.2(a) and 4.2(b) illustrates intracellular delivery of PI with the treatment of
23 V / 23 μs and 23 V / 30 μs, respectively. The first columns of Figure 4.2 illustrate
time-based intensity changes of PI inside a HeLa cell (Van Wamel et al., 2006). Figure
4.2(a) shows that there was no PI intensity at 0 sec, the maximum PI intensity at 35 sec
and diffused and slightly decreased PI intensity at 250 sec, respectively. Figure 4.2(b)
indicates that there was no PI intensity at 0 sec, and after treatment, rapidly increased PI
intensity and then reached the plateau of PI intensity were observed.
46
Figure 4.2 Intracellular delivery of PI inside the HeLa cell with the treatment of 23 V / 23
μs and 23 V / 30 μs, respectively (a)-(b) Time-based PI intensity changes. Arrows show
treated cells and scale bars indicate 40 μm.
Figure 4.3 illustrates intracellular delivery of PI with the repeated treatment (23 V
/ 3 μs). Figure 4.3(a) illustrates time-dependent PI intensity changes inside the treated
HeLa cell. Figure 4.3(b)-(e) suggest that there was no PI intensity at 0 sec before the
treatment, and acoustic pulses were applied at 3 sec, 63 sec and 123 sec, resulted in three
increases of PI intensity in the targeted cell, respectively.
(b)
(a)
47
Figure 4.3 Intracellular delivery of PI (a) Time-based PI intensity changes inside the
HeLa cell. (b) There was no PI intensity before treatment (0 sec) (c)-(e) Diffused and
slightly increased PI intensity caused by repeated treatment (3 sec, 63 sec and 123 sec).
Arrows show treated cells and scale bars indicate 40 μm.
4.3.2 Intracellular Delivery of Small-sized Molecule
Figure 4.4 (a)-(b) and 4.4(c)-(e) represent intracellular delivery of 3 kDa dextran
labeled with alexa fluor 488 using the treatment of 23 V / 16 μs and 23 V / 23 μs,
respectively. The first and second columns of Figure 4.4 show bright-field and
fluorescence images of a HeLa cell 0.5 hour after the treatment. In the vicinity of
untreated HeLa cells did not indicate any green fluorescence, while diffused green
fluorescence was observed in the targeted HeLa cell. The third and fourth columns of
Figure 4.4 illustrate bright-field and fluorescence images of a HeLa cell after 40 hours of
the treatment. Daughter cells were observed and had the same fluorescence signal, which
suggesting that the treated cells had divided after 40 hours and were functioning well
without a significant defect as shown in fourth columns of Figure 4.4.
(a)
(b)
(c) (d)
(e)
48
(c)
(a)
(b)
49
Figure 4.4 Intracellular delivery of Alexa fluor 488-labeled of 3 kDa dextran with (a)-(b)
Treatment of 23 V / 16 μs (c)-(e) Treatment of 23 V / 23 μs. The first and second
columns indicate bright-field and fluorescence images of a HeLa cell after 0.5 hour of
treatment. The third and fourth columns show bright-field and fluorescence images of the
HeLa cell after 40 hours of the treatment. Arrows show treated cell and scale bars
indicate 40 μm.
4.3.3 Optimal Treatment Conditions on Human Cancer Cell Lines indicated
by Intracellular Delivery Score
A detailed quantitative study was carried out to find optimal treatment conditions
towards high delivery efficient with low cytotoxicity by controlling treatment conditions
of acoustic transfection. Figure 4.5 indicates the effects of different treatment conditions
(d)
(e)
50
Figure 4.5 Percentage of delivery efficiency with different treatment conditions for (a)
HeLa (b) MCF-7 (c) MCF-10A and (d) MDA-MB231 cell lines.
on four human cancer lines for delivery efficiency. Percentage of delivery efficiency with
different treatment conditions for (a) HeLa, (b) MCF-7, (c) MCF-10A, and (d) MDA-
MB231 cell lines. As higher Vpp and longer Tt, the percentage of delivery efficiency
continued to increase. To be specific, the percentage of delivery efficiency reached 100%
under treatment conditions of 18V / 60μs - 90μs, 21V / 18μs - 90μs, 24V / 12μs - 90μs
and 27V / 12μs - 90μs, while treatment conditions of 12V / 6μs - 30μs did not have
influence on the percentage of delivery efficiency. Figure 4.6 demonstrates the effects of
(a)
(b)
(c) (d)
51
Figure 4.6 Cell membrane permeability with different treatment conditions for (a) HeLa
(b) MCF-7 (c) MCF-10A and (d) MDA-MB231 cell lines.
different treatment conditions on four human cancer cell lines for cell membrane
permeability. Cell membrane permeability with different treatment conditions for (a)
HeLa, (b) MCF-7, (c) MCF-10A, and (d) MDA-MB231 cell lines. Acoustic transfected
cells had higher PI uptake with higher Vpp and longer Tt. It was found that treatment
conditions of 12V / 6μs - 90μs and 15V / 6μs - 60μs did not have significant impact on
cell permeability. Figure 4.7 shows the effects of different treatment conditions on four
human cancer cell lines for cell viability after 4 and 20 hours of the acoustic transfection.
Percentage of cell viability after 4 and 20 hours after acoustic transfection with different
(a)
(b)
(c)
(d)
52
(a)
(b)
(c)
(d)
53
Figure 4.7 Percentage of cell viability after 4 and 20 hours after acoustic transfection with
different treatment conditions for (a) HeLa (b) MCF-7 (c) MCF-10A and (d) MDA-
MB231 cell lines.
treatment conditions for (a) HeLa, (b) MCF-7, (c) MCF-10A, and (d) MDA-MB231 cell
lines. The percentage of cell viability after 4 and 20 hours of the acoustic transfection
decreased with treatment conditions of higher Vpp and longer Tt. It was seen that the
percentage of cell viability after 4 and 20 hours of the acoustic transfection was
dramatically decreased when acoustic transfected cells were exposed to treatment
conditions of 21V / 90μs, 24V / 60μs - 90μs and 27V / 60μs - 90μs, while the most other
treatment conditions did not affect cell viability. The intracellular delivery scores (IDS)
for HeLa, MCF-7, MCF-10A, and MDA-MB231 cells were calculated for varying
treatment conditions. Figure 4.8 shows the intracellular delivery graph (IDG) after 4 and
20 hours of the acoustic transfection using the IDS on HeLa (Figure 4.8(a)), MCF-7
(Figure 4.8(b)), MCF-10A (Figure 4.8(c)) and MDA-MB231 (Figure 4.8(d)) cells for
easier viewing. Table 4.2 gives optimal treatment conditions which yields the IDS larger
than 9 for the human cancer cell lines studied. According to IDS at 4 and 20 hours,
optimal treatment conditions for each cell line were chosen as shown in table 4.2. These
optimal treatment conditions were selected if cell viability was bigger than 90%.
54
(a)
(b)
(c)
(d)
55
Figure 4.8 Intracellular delivery graph (IDG) using the intracellular delivery scores (IDS)
after 4 and 20 hours of acoustic transfection for different treatment conditions. IDS is
plotted on y-axis under six different V
pp
. Six different T
t
were applied at each V
pp
. IDG
after 4 and 20 hours was used to find optimal treatment conditions for (a) HeLa (b) MCF-
7 (c) MCF-10A and (d) MDA-MB231 cell lines.
Table 4.2 Summary of the optimal treatment conditions
HeLa MCF-7
MCF-10A
MDA-MB231
V
pp
(V)
T
t
(µs)
IDS
V
pp
(V)
T
t
(µs)
IDS
V
pp
(V)
T
t
(µs)
IDS
V
pp
(V)
T
t
(µs)
IDS
4
h
20
h
4
h
20
h
4
h
20
h
4
h
20
h
18 60 9 9
18 90 10 10 21 60 9 9 18 60 10 10
18 90 9 9
21 60 10 10 24 30 9 9 21 30 9 9
21 60 10 10
24 30 9 9 27 18 9 9 24 18 10 10
24 30 9 9
27 18 9 9 27 30 9 9 24 30 10 10
27 30 10 10
27 30 10 10 - - - - 27 12 9 9
- - - -
- - - - - - - - 27 18 10 10
4.3.4 Intracellular Delivery of Macromolecules with Optimal Treatment
Conditions
The intracellular delivery of 70 kDa dextran labeled with Oregon Green into
HeLa, MCF-7, MCF-10A, and MDA-MB231 cells under optimal treatment conditions is
demonstrated in Figure 4.9. Left columns in Figure 4.9 indicate bright-field images
before the acoustic transfection, and middle and right columns in Figure 4.9 show bright-
field and fluorescence images after 0.5 hour of the acoustic transfection, respectively.
56
(a)
(b)
(c)
(d)
57
Figure 4.9 Intracellular delivery of 70 kDa dextran labeled with Oregon Green using
optimal treatment conditions. Left column indicates bright-field images before the
acoustic-transfection, and middle and right columns show bright-field and fluorescence
images after 0.5 hour of the acoustic transfection for (a) HeLa (b) MCF-7 (c) MCF-10A
and (d) MDA-MB231 cell lines. Arrows show acoustic transfected cells and scale bars
indicate 40 μm.
Diffused green fluorescence was observed in the targeted cells while neighboring cells
have no green fluorescence.
4.3.5 Simultaneous Intracellular Delivery of Two Molecules with Optimal
Treatment Conditions
Simultaneous intracellular delivery of 70 kDa dextran labeled with Oregon Green
and propidium iodide (PI) into HeLa, MCF-7, MCF-10A, and MDA-MB231 cells under
optimal treatment conditions are shown in Figure 4.10. The first and second columns in
Figure 4.10 indicate bright-field images before and after acoustic transfection,
respectively, and the third and fourth columns in Figure 4.10 show fluorescence images
of 70 kDa dextran labeled with Oregon Green and PI after 0.5 hour of the acoustic
transfection, respectively, and the fifth columns in Figure 4.10 show overlapped images
of the two different fluorescence images. It can be seen that acoustic transfected cells in
the nucleus region emit red fluorescence due to the binding of PI molecules with nucleic
acids, while the cytoplasmic staining of 70 kDa dextran labeled with Oregon Green is
clearly observed.
58
Figure 4.10 Simultaneous intracellular delivery of 70 kDa dextran labeled with Oregon
Green and propidium iodide (PI) under optimal treatment conditions. The first and
second columns indicate bright-field images before and after acoustic transfection,
respectively, and the third and fourth columns show fluorescence images of 70 kDa
dextran and PI after 0.5 hour of the acoustic transfection, respectively, and the fifth
columns overlapped images of the two different fluorescence images for (a) HeLa (b)
MCF-7 (c) MCF-10A and (d) MDA-MB231 cell lines. Arrows indicate acoustic
transfected cells and scale bars indicate 40 μm.
Figure 4.11 illustrate intracellular delivery efficiency of 70 kDa dextran labeled with
Oregon Green into HeLa, MCF-7, MCF-10A, and MDA-MB231 cells under optimal
treatment conditions, respectively. Therefore, the optimal treatment conditions for
(d)
(c)
(b)
(a)
59
macromolecules (≤ 70 kDa) were those that achieved highest percentage of 70 kDa
dextran delivery efficiency. Highest percentages of 70 kDa dextran delivery efficiency
were 89% (16/18) with the treatment condition of 24V / 30μs for HeLa cell, 83% (5/6)
and 83% (15/18) with the treatment conditions are 24V / 30μs and 27V / 30μs,
respectively for MCF-7 cell, 83% (5/6) with the treatment condition of 27V / 18μs for
MCF-10A cell, 72% (13/18) with the treatment condition of 21V / 30μs for MDA-
MB231 cell.
Figure 4.11 The intracellular delivery efficiency graph of 70 kDa dextran labeled with
Oregon Green under optimal treatment conditions. (a) Highest percentage of 70 kDa
dextran delivery efficiency is 89% (16/18) with the treatment condition of 24V / 30μs for
HeLa cell. (b) Highest percentage of 70 kDa dextran delivery efficiency of 83% (5/6) and
83% (15/18) under treatment conditions are 24V / 30μs and 27V / 30μs, respectively for
MCF-7 cell. (c) Highest percentage of 70 kDa dextran delivery efficiency is 83% (5/6)
(a) (b)
(c) (d)
60
under the treatment condition of 27V / 18μs for MCF-10A cell. (d) Highest percentage of
70 kDa dextran delivery efficiency of 72% (13/18) under the treatment condition of 21V /
30μs for MDA-MB231 cell.
4.4 Discussions
In this chapter, the concept of acoustic transfection with high frequency
ultrasound was confirmed by intracellular delivery of small-sized molecules with
capabilities of low cytotoxicity and remote targeting at single-cell level. The investigation
of treatment conditions was further conducted to find optimized ultrasound exposure
levels that allow the most efficient and safe delivery using the acoustic transfection.
Results on the intracellular delivery of a macromolecule such as 70 kDa dextran into four
human cancer cell lines were acquired to determine whether acoustic transfection is
dependent on cell types and molecules to be transfected (Fan et al., 2013; Guzman et al.,
2001; Karshafian et al., 2009; Zeira et al., 2003). In order to shed light on these
unresolved issues, the optimal treatment conditions of acoustic transfection were
determined, and utilized for the intracellular delivery of 70 kDa dextran labeled with
Oregon Green and simultaneous intracellular delivery of two molecules namely, 70 kDa
and propidium iodide (PI) to assess the acoustic transfection as a method for efficient
delivery of therapeutic or genetic materials. As described in Table 4.1, three parameters
i.e., delivery efficiency, cell membrane permeability, and cell viability after 4 and 20
hours of the acoustic transfection were considered to determine the intracellular delivery
score (IDS). The most crucial determinant factor among the three parameters is the cell
viability. According to previously published studies, it has been reported that highest
61
percentage of 70 kDa dextran delivery efficiency was from 30% to 72% and at the same
time, cell viability was less than 90% (Larina et al., 2005; Sharei et al., 2013; Song et al.,
2015). In contrast, Figure 4.11 shows highest percentage of 70 kDa dextran efficiency
ranged from 72% to 89% and at the same time, cell viability was bigger than 90%.
Therefore, the proposed criterion for the intracellular delivery score (IDS) enables the
determination of optimal treatment conditions in various molecular size ranges from
small-molecules to macromolecules by satisfying high delivery efficiency with minimal
cell membrane disruption.
In this approach, the hypothesis was that a tightly focused high frequency
ultrasound beam was possible to disrupt the structural integrity of cell membranes and
perhaps generate transient and reversible holes on cell membranes with merely the dose
of a single ultrasound pulse of a few microsecond duration focused into an area of
approximately 10 μm in diameter. The ultrasound energy may result in the formation of a
physical pathway for introducing membrane-impermeable molecules into cell cytoplasm
and passive diffusion driven by the concentration gradient across transiently generated
holes. The hypothesis was further supported by intensive investigation of treatment
conditions between different human cancer cell lines, and intracellular delivery of a
molecule from small-molecules to macromolecules and simultaneous intracellular
delivery of two molecules.
The dynamic behaviors of cell plasma membrane correlated with various
treatment conditions were classified into four processes based on the intensive
62
investigation of treatment conditions. First, the effects of smallest V
pp
and/or shortest T
t
were observed for lowest percentage of delivery efficiency, lowest cell membrane
permeability, and highest percentage of cell viability, suggesting that the treatment
conditions were not sufficient to induce significant disruption of cell membrane and
deliver membrane-impermeable molecules into the cytoplasm of cells. Second, a slight
increase in V
pp
and/or T
t
resulted in disturbance of cell membrane and perhaps generation
of small-scale transient holes on the cell membrane, diffusion-driven transport, and
immediate sealing of the holes in the ruptured membrane, which allowed intracellular
delivery of exogenous small-molecules such as propidium iodide (PI) and 3 kDa dextran.
Third, as V
pp
and/or T
t
were increased up to the threshold point of cell death from the
second treatment conditions, increased disruption of cell membrane, passive diffusion
across transiently generated holes, and sealing of the transient holes in the disturbed cell
membrane were confirmed by intracellular delivery of small-molecules, macromolecules
such as PI, 3 kDa dextran, and 70 kDa dextran, and even two molecules including PI and
70 kDa dextran. Fourth, the cell responses exposed to highest V
pp
and/or longest T
t
exhibited remarkably high percentage of delivery efficiency, high cell membrane
permeability; however, lowest percentage of cell viability. It was postulated that these
treatment conditions were sufficient to disrupt cell membrane and to prevent transient
holes on the membrane from sealing themselves quickly, leading to permanent opening
of the holes on the cell membrane and irreversible membrane damage. Also, it can be
estimated that 70 kDa dextran is bigger than the cut-off size of the nuclear membrane
from the results in Figure 4.9 and 4.10. Acoustic transfected cells in the nucleus region
63
emitted red fluorescence with binding of PI to nucleic acids; on the other hand, the cells
were mostly stained with 70 kDa dextran in cytoplasm region rather than nucleus region.
A small amount of diffusion-driven 70 kDa dextran transport in the nucleus region was
likely to have limited access to the nucleus region due to its molecular size compared
with the nuclear membrane size. These results are in good agreement with previous
published studies (Guzman et al., 2002; Larina et al., 2005).
Cell responses of different human cancer cell lines exposed to the acoustic-
transfection were found to be different, suggesting that optimal treatment conditions may
be cell dependent (Antkowial et al., 2013; Larina et al., 2005; Mthunzi et al., 2010). For
example, for the delivery efficiency of 70 kDa dextran, the optimal treatment conditions
were 24V / 30μs, which resulted in 89% delivery for HeLa cell and 83% for MCF-7 cell
but only 57% for MCF-10A cell and 50% for MDA-MB231 cell. Some possibilities were
highlighted why different human cancer cell lines showed different responses to the same
treatment condition based on the hypothesis of acoustic transfection. Since the observed
behavior was the result of the creation of a physical pathway in which transient and
reversible holes were generated on the cell membranes, different cell responses might be
attributed to considerably different physical and mechanical properties of human cancer
cell lines including cell membrane recovery time related to elasticity of cell membranes
and diameters of transiently generated holes on the surface of cell membranes determined
by the membrane tensile strength. Therefore, to verify this hypothesis it will be necessary
to extend this study to design and develop a system to allow simultaneous acoustic
transfection and membrane electrical impedance measurements to determine how the
64
acoustic transfection influences membrane dynamics of different cancer cell lines in the
future.
There was a limitation of the present study. Measurement of acoustic pressure
field for high frequency ultrasound in the frequency range higher than 60 MHz cannot be
performed by using currently existing technology (Bleeker et al., 2000; Nagle et al., 2013;
Umchid et al., 2009). As a result, treatment conditions represented by V
pp
and T
t
could
not directly be converted to acoustic pressure, intensity, and energy, which has been a
controversial issue. Therefore, the acoustic pressure generated by these tightly focused
high frequency ultrasonic transducers was estimated with a commercial finite element
modeling software (PZFlex, Cupertino, CA), and a quantitatively simulated values of
acoustic pressure filed at the focus of the transducer were plotted for treatment conditions
of Vpp (12V, 15V, 18V, 21V, 24V, and 27V) and Tt (6µs, 12µs, 18µs, 30µs, 60µs, and
90µs) in Figure 4.12. It was found that the maximum pressures estimated under ideal
conditions at the focal depth of the ultrasonic transducer were approximately 0.86 MPa
(12V / 6μs - 90μs), 1 MPa (15V / 6μs - 90μs), 1.28 MPa (18V / 6μs - 90μs), 1.5 MPa
(21V / 6μs - 90μs), 1.7 MPa (24V / 6μs - 90μs), and 2 MPa (27V / 6μs - 90μs),
respectively. Mechanical index (MI) was further estimated 0.15 which was calculated by
dividing simulated values of maximum acoustic pressure field (2 MPa) by the square root
of the center frequency (182 MHz) of the high frequency ultrasound beam. The MI of
0.15 was much lower than the maximum permitted value for the MI of 1.9 under the
Food and Drug Administration (FDA) rules (FDA, 1997). However, either the current
device will have to be improved or new devices developed to allow experimental
65
Figure 4.12 The simulation of acoustic pressure filed was performed by a commercial
finite element modeling software under ideal conditions treatment.
measurement of acoustic pressure, intensity, and energy for high frequency ultrasound in
the range above 60 MHz.
Controlling cell functions by efficiently and specifically introducing therapeutic
or genetic materials into the targeted single cells with minimal effects on normal cell
physiology is extremely useful for investigating induction of programmed cell death of
cancer cells which is referred to as apoptosis and mapping of cellular signaling pathways
(Elmore et al., 2007; Fesus et al., 1991; Matsushita et al., 2000). In these applications, the
capability of single-cell targeting without significantly affecting surrounding cells is
preferred. Since the signal pathways underlying apoptosis and intercellular interactions
among a cell in apoptosis and its adjacent cells are still poorly understood, careful
66
measurements of intracellular delivery of molecules including p53 tumor suppressor
protein and Ca
2+
may shed more light on extracellular and intracellular cell signaling
pathways. Once the extracellular and intracellular signal pathways are precisely known,
appropriate strategies on apoptosis-targeted therapies may be formulated and
subsequently translated to clinical medicine for the treatment of numerous human
diseases such as cancer.
67
CHAPTER 5 Label-free Acoustic Sensing of a Single Cell
Trapped by High Frequency Ultrasound
5.1 Introduction
Cellular analysis is a critical stage in research and clinical applications
(Chattopadhyay et al., 2014; Del Fattore et al., 2006; Hence et al., 2013; Kawamura et al.,
2009; Kawamura et al., 2014; Ramser et al., 2010). Investigating cellular properties and
functional responses along with precise manipulation techniques plays a pivotal role in
great improvements with respect to gaining insight into molecular dynamic studies in
living cells (Atienza tl al., 2006; Guo et al., 2015), potentially undetected cell signaling
pathway and networks (Palomba et al., 2014; van Unen et al., 2010), discovering gene
expression profiles (Chen et al., 2005; Rue et al., 2015; Sanchez-Freire et al., 2012), and
discovery and development of new drug (Atienzar et al., 2011; Heath et al., 2010; Piret et
al., 2016). For example, recent study has discovered unique gene expression patterns by
controlling human metastatic breast cancer cells, measuring and identifying gene
expression differences (Lawson et al., 2015). However, conventional cellular analysis
techniques have been shown to have a serious drawback because these approaches could
hardly discern subpopulations in a heterogeneous population, which might yield
misleading information (Gross et al., 2015; Hu et al., 2016; Wang et al., 2010; Yin et al.,
2012).
To address the critical problem, a range of single-cell analysis techniques have
been developed to investigate cellular behaviors with deeper information between
68
individual cells at single-cell level. Fluorescent-activated cell sorting (FACS) and
Magnetic-Activated Cell Sorting (MACS) have been widely used for a quantitative
approach for the determination of cells of interest. FACS provides rapid and reliable
information in a homogeneous subpopulation within a heterogeneous population;
however it is overall bulky, expensive, labor-intensive and time-consuming procedures,
and may influence on the normal cell physiology and cellular functions by tagging a
fluorescence probe for the detection of cells of interest (Cho et al., 2010; Schoell et al.,
1999). MACS offers high-throughput screening, accurate information, and cost-effective,
while specific markers of interest to visually recognize targeted cells and labor-intensive
and time-consuming sample preparations are required (Miltenyi et al., 2009; Said et al.,
2010). Also, it may be likely to interfere with cellular behaviors in the middle of labelling
with antibodies on the cell membrane surface. Therefore, label-free single-cell analysis
techniques on the basis of intrinsic physical properties, such as cell size, shape,
compressibility, and polarizability have been attracting attention with minimal effects on
normal cell physiology and cell functions by managing the complexity of sample
preparation and analysis procedures (Grossett et al., 2009; Kumar et al., 2010; Xu et al.,
2016). Microfluidic system provides label-free single-cell analysis, high-throughput, and
low device fabrication; however, this method suffers from unexpected negative effects on
cell behaviors and responses caused by uncoordinated shear stress, and clogging along
with geometrical microstructures (Crowley et al., 2005; Yamada et al., 2004) due to
complicated design. Besides, microfluidic system with a variety of physical sources
including dielectrophoretic forces (Huang et al., 2002; VanDelinder et al., 2006), laser
69
radiations (Liu et al., 1996; Neuman et al., 1999), and standing surface acoustic waves
(Ding et al., 2012; Falou et al., 2010) has been developed to utilize advantages of each
physical source on top of strong points of microfluidic system; however, these physical
sources in the microfluidic technique are restricted to complex microelectrode fabrication,
expensive laser setups, complicated alignments of standing surface acoustic waves and
laser radiations, and difficult-to-handle with the physical sources which may cause
irreversible membrane damage.
An acoustic-based single-cell analysis method using a single element ultrasonic
transducer has been recently demonstrated by theoretical and experimental approach with
simple, cost-effective, label-free investigation of physical properties by manipulating
particles and cells (Falou et al., 2010; Lee et al., 2011; Lee et al., 2015). However,
previous approaches suffer from some practical limitations; for example, in order to
obtain and measure accurate information, microinjection system as a manipulation
module was integrated with ultrasound measurement system; however, the microinjection
technique requires special care in handling with a micropipette on cell membrane (Falou
et al., 2010). Moreover, wavelength of ultrasound microbeam used in previous
approaches was much larger than cell sizes in suspension, which is much less likely to
achieve accurate cellular physical properties. Therefore, a systematic approach in solving
the problems was needed to individually control micron-sized particles and single cells,
and at the same time, analyze physical properties from the high frequency ultrasound
microbeam whose wavelength is a close parallel with the size of the cells.
70
In this chapter, a rational strategy was proposed for separating microscopic
objects capable of individually identifying the size of the trapped single object based on
ultrasound backscattering coefficient through a monocycle ultrasound pulse of a few
nanosecond periods within a wavelength of 10 μm without integrating with other devices
and changing any experimental settings. This approach utilizes a highly focused high
frequency ultrasound microbeam generated by a tightly focused high frequency ultrasonic
transducer with custom-built impedance matching network, and a custom-built all-in-one
front-end system having the ability to produce high pulse repetition frequency (PRF)
monocycle bipolar pulse as a transmitter module, and capability to discern low-level
acoustic signals from noise signals by improving signal to noise ratios (SNR) as a
receiver module. The proposed strategy provides beneficial characteristics, including it
being an easy-to-use, cost-effective, label-free investigation of physical properties with
minimal effects on normal cell physiology with a much more accurate analysis at the
single-cell level. To formulate the objective, the ability to manipulate micron-sized
polystyrene particles by using the monocycle ultrasound pulse with high PRF was tested,
and then trapping force and the effects induced by high frequency ultrasound microbeam
with three driving conditions in the burst length range from long pulses to one pulse
while maintaining the same duty factor were investigated. After the manipulation
capability in the proposed driving condition was verified, the rational strategy was used
for the size determination of a trapped polystyrene particle in accordance with ultrasound
backscattering coefficient, and functionally differentiating between red blood cells (RBCs)
71
and cancer cells (PNT1A) by independently measurement based on ultrasound
backscattering coefficient with visual confirmation of controlling objects.
5.2 Materials and Methods
5.2.1 Two specifications for the Proposed Driving Condition for High
Frequency Ultrasound Mircrobeam
A high frequency ultrasound microbeam with a proposed driving condition was
proposed in terms of two specifications to prevent reverberation and interference between
transmitted signals and reflected echo signals. The two specifications were related with
on time and pulse repetition frequency (PRF).
On time < time
2way
(=
2∙𝑑 𝑐 ) < PRF (5.1)
For example, the tested high frequency ultrasonic transducer had the focal length
of 1.95 mm, and time
2way
was calculated as 2.6 μs. In order to minimize reverberation and
interference between transmitted pulses and reflected echo signals, on time and PRF of
high frequency ultrasound microbeam should be shorter and longer than 2.6 μs,
respectively. Therefore, the driving condition was determined as the shortest on time of
6.7 ns and PRF of 6.7 μs, which was duty factor of 0.1 %.
72
5.2.2 High Frequency Ultrasound based Label-free Cell Separation System
Figure 5.1 demonstrates high frequency ultrasound based label-free cell separation
system comprised of an inverted fluorescence microscope and acoustic cell separation
system. The inverted fluorescence microscope (IX71, Olympus, Center Valley, PA) was
used to acquire time-resolved bright-field images for visual confirmation of trapping a
particle and moving along the direction of the high frequency ultrasound microbeam. A
three dimensional linear translation / rotation stage controlled by a customized LabVIEW
Figure 5.1 High frequency ultrasound based label-free cell separation system which was
used for individually measuring ultrasound backscattering coefficients of a trapped single
object on a mylar film.
73
(National Instruments, Austin, TX) program was used to precisely control the location of
the transducer with impedance matching network (IMN) during the experiments. The
custom-built front-end system and an oscilloscope were used to accurately focus the
transducer with IMN, and generate a monocycle bipolar pulse with high PRF (on time:
6.7 ns / PRF: 167 kHz) for individually measuring ultrasound backscattering coefficients
of a trapped single object on a mylar film.
5.2.3 Effects of High Frequency Ultrasound Microbeam in terms of
Different Driving Conditions
To study the effects of high frequency ultrasound microbeam induced by different
driving conditions, three driving conditions in the burst length range from long pulses (on
time: 1μs / PRF: 1 ms) to one pulse (on time: 6.7 ns / PRF: 167 kHz) while maintaining
the same duty factor was selected. The high frequency ultrasound microbeam was applied
on polystyrene bead (50 μm) monolayer, and after 3 minutes, a trapped bead was moved
along the direction of the high frequency ultrasound microbeam. Bright-field images
were acquired to track the influences on the particles before and after acoustic trapping.
Also, as a quantitative approach, acoustic trapping force measurement was performed by
following previously published study (Lee et al., 2010; Lim et al., 2016). Acoustic
trapping force for 1.5 μm (n=4) and 2 μm displacement (n=4) from the center of high
frequency ultrasound microbeam was measured and compared in terms of three different
driving conditions.
74
5.2.4 Cell Preparation
Fresh human whole blood were obtained by a volunteer and utilized according to
an institutional review board (IRB) protocol (UP-16-00713). Whole blood was
centrifuged with phosphate-buffered saline (PBS) at 500 × g for 10 minutes to separate
red blood cells (RBCs) from white blood cells (WBCs) with platelets, blood plasma, and
PBS. After gently removing supernatant, RBCs were resuspended with PBS, and then the
cells with PBS were once again centrifuged and resuspended with a mixed solution of
PBS and alsever’s solution.
Normal SV40 immortalized epithelial prostate (PNT1A) cells were cultured in
RPMI 1640 supplemented with 10% fetal bovine serum (FBS) and incubated in a
humidified 5% CO
2
at 37˚C incubator. The PNT1A cells were gently washed twice with
PBS and dispensed with TrypLE solution in the incubator for 5 minutes, and centrifuged
at 150 × g for 5 minutes. After gently discarding supernatant, PNT1A cells were
resuspended with PBS, and then the cells with PBS were once again centrifuged and
resuspended with the PBS.
5.2.5 Cell Viability Study and Statistical Analysis
Viability of RBCs and PNT1A cells was performed to investigate effects of the
proposed driving condition (voltage: 50 V / on time: 6.7 ns / PRF: 167 kHz / duty factor:
0.1 %) compared to control condition (voltage: 0 V / on time: 0 ns / PRF: 0 kHz / duty
75
factor: 0 %). The cells were washed twice with PBS, and stained with a membrane-
permeable live-cell labeling dye (Calcein, AM, Thermo Fisher Scientific, Waltham, MA)
by a previously reported approach (Lam et al., 2016). Live-cell fluorescence imaging was
acquired to compare the fluorescence level before and after 30 minutes of the exposure to
high frequency ultrasound microbeam with the control and proposed driving condition.
Statistical analysis was used to (1) size determination of a trapped micron-sized
single particle (n=20), (2) separation between red blood cells and cancer cells (n=16), and
(3) cell viability study (n=10) using a two-tailed paired t-test.
5.3 Results
5.3.1 Characterization of High Frequency Ultrasound Microbeam with the
Proposed Driving Condition
A capability to manipulate polystyrene microspheres using the proposed
monocycle ultrasound pulse with high PRF was tested. Figures 5.2(a)-(d) describe time-
resolved bright-field images before and after influence of high frequency ultrasound
microbeam. The initial and moving location of the high frequency ultrasound beam are
indicated as white and red dashed circles, respectively, and the scale bars indicate 100
µm. It was found that high frequency ultrasound microbeam driven by the monocycle
ultrasound pulse with high PRF could certainly trap a particle and move along the
76
Figure 5.2 Characterization of high frequency ultrasound microbeam with a proposed
driving condition. (a)-(d) An ability to trap and move a polystyrene microsphere using the
proposed monocycle ultrasound pulse with high PRF was confirmed by taking time-
based bright field images before and after influence of high frequency ultrasound
microbeam. White and red dashed circles were described as the initial and moving
location of the high frequency ultrasound beam, respectively. Scale bars indicate 100 µm.
(e-g) Acoustic trapping force and its effects were described by distances between a
trapped bead and its adjacent beads. Red dashed circles and red arrow mean the location
of the high frequency ultrasound beam and the distances between a trapped bead and its
adjacent beads. The impacts had the greatest in (e) voltage: 10V / on time: 1μs / PRF: 1
kHz, followed by (f) voltage: 10V / on time: 100 ns / PRF: 10 kHz, and (g) voltage: 10V /
on time: 6.7 ns / PRF: 167 kHz. (H-J) Modified acoustic trapping force and its effects
were demonstrated. There was very similar effects after changing driving conditions (h)
voltage: 3.8V / on time: 1μs / PRF: 1 kHz, (i) voltage: 4.1V / on time: 100 ns / PRF: 10
kHz compared with proposed driving condition (j) voltage: 10V / on time: 6.7 ns / PRF:
167 kHz.
(a) (b) (c) (d)
(e)
(f) (g)
(h) (i) (j)
77
direction of the high frequency ultrasound microbeam. In addition to observation of
trapping phenomenon, acoustic trapping force was measured as 5.2 ± 0.43 nN at 1.5 μm
displacement from the center of ultrasound microbeam (n=4) as a quantitative approach.
After probing acoustic trapping performance, a comparative study was further conducted
with regard to acoustic trapping force and performance, which resulted from three
different driving conditions, such as on time: 1μs / PRF: 1 kHz in Figure 5.2(e), on time:
100 ns / PRF: 10 kHz in Figure 5.2(f), and on time: 6.7 ns / PRF: 167 kHz in Figure
5.2(g). The location of the high frequency ultrasound beam and the distances between a
trapped bead and its adjacent beads are illustrated as red dashed circles and red arrow,
respectively. It was seen that the impacts of high frequency ultrasound microbeam
linking with distances between a trapped bead and its adjacent beads had the greatest in
the first driving condition (on time: 1μs / PRF: 1 kHz) in Figure 5.2(e), followed by the
second driving condition (on time: 100 ns / PRF: 10 kHz) in Figure 5.2(f), and the third
driving condition (on time: 6.7 ns / PRF: 167 kH) in Figure 5.2(g). Also, we measured
acoustic trapping force generated from each driving condition, and the first (Figure 5.2(e))
and second (Figure 5.2(f)) driving condition exceeded 50 nN compared to 5.2 ± 0.43 nN
in the third (Figure 5.2(g)) driving condition (n=4). To find the same acoustic trapping
force of 4.8 ± 0.59 nN at 2 μm displacement from the center of high frequency ultrasound
microbeam (n=4), input voltage of the first and second driving conditions was adjusted
from 10 V to 3.8 V, 4.1 V, respectively while maintaining on time, PRF, and duty factor.
It was found that quite similar effects connected with distances between a trapped bead
78
and its surrounding beads produced by revised driving conditions in Figure 5.2(h) and
Figure 5.2(i) and proposed driving condition in Figure 5.2(j) were examined.
5.3.2 Size Determination of a Trapped Single particle
Next, the proposed rational strategy capable of individually identifying the size of the
trapped micron-sized polystyrene single microsphere based on ultrasound backscattering
coefficient was performed and validated. Size difference between 5 μm and 10 μm was
determined by different levels of ultrasound backscattering coefficient as presented in
Figure 5.3. The calculated and averaged ultrasound backscattering coefficient obtained
from each polystyrene microbead in a diameter of 5 μm and 10 μm were -109.52 ± 0.75
dB and -98.84 ± 0.84, respectively (n=20). This result suggests that the smaller particle
had a lower level of ultrasound backscattering coefficient, and bigger particle had a high
level of ultrasound backscattering coefficient, which is in good agreement with previous
published studies (Lee et al., 2011). Also, this difference was strongly significant (p-
value < 0.01), which can be explained by the characteristic of high specificity.
79
Figure 5.3 Size determination of a trapped micron-sized single particle (5 μm / 10 μm)
was performed. The ultrasound backscattering coefficient obtained by each polystyrene
microbead of -109.52 ± 0.75 dB (5 μm) and -98.84 ± 0.84 (10 μm) was calculated and
averaged (n=20). Also, this difference was strongly significant according to statistical
analysis (p-value < 0.01).
5.3.3 Separation between Red Blood Cells and Cancer Cells
This approach further used to separate between red blood cells (RBCs) and
normal SV40 immortalized epithelial prostate (PNT1A) cells. To check whether the
ultrasound backscattering coefficient accurately returned from the targeted single cell,
each RBC and PNT1A cell was trapped and moved along the direction of the high
80
Figure 5.4 A capability of trapping and moving each red blood cell (RBC) and normal
SV40 immortalized epithelial prostate (PNT1A) cell using the proposed monocycle
ultrasound pulse with high PRF was confirmed. The initial and moving locations of the
high frequency ultrasound beam are indicated as black and red dashed circles,
respectively. The scale bars indicate 20 µm. Manipulation of (A) RBC and (B) PNT1A
cell.
Frequency ultrasound microbeam as presented in Figure 5.4(a)-(h) and Figure 5.4(i)-(p),
respectively. The initial and moving location of the high frequency ultrasound beam are
(a) (b) (c) (d)
(e)
(f)
(g)
(h)
(i) (j) (k) (l)
(m) (n) (o) (p)
81
indicated as black and red dashed circles, respectively, and the scale bars indicate 20 µm.
After confirming the ability to manipulate cells using the proposed driving condition, cell
separation between RBCs and PNT1A cells was conducted. Figure 5.5 shows diameter
Figure 5.5 Size determination of a trapped cell (RBC / PNT1A cell) was studied. The
ultrasound backscattering coefficient obtained by each RBC and PNT1A cell of -109.03
± 0.77 dB and -106.74 ± 0.22 dB was calculated and averaged (n=16). Also, statistical
analysis shows statistical significance (p-value < 0.01).
differences between RBCs (6 μm – 8 μm) and PNT1A cells (9 μm – 12 μm) resulted in
different levels of ultrasound backscattering coefficient. The calculated and averaged
ultrasound backscattering coefficient from each RBC and PNT1A cell were -109.03 ±
82
0.77 dB and -106.74 ± 0.22 dB, respectively (n=16), and statistical significance (p-value
< 0.01) was confirmed by values of ultrasound backscattering coefficient. Therefore, the
results clearly demonstrate that our label-free separating approach not only allows for
identification of the size of an object, but also enables for the manipulation of the object
in suspension at single-cell level.
5.3.4 Cell Viability Study
Cell viability test was performed to verify this approach is safe to RBCs and
PNT1A cells. Figure 5.6 demonstrates statistical analysis of RBCs and PNT1A cells
based on normalized mean fluorescence intensity obtained from fluorescence level before
and after acoustic trapping in terms of a control condition and a proposed driving
condition. The normalized mean fluorescence intensity obtained from RBC was 0.98 ±
0.02 for the control condition and 0.97 ± 0.01 for the proposed driving condition,
respectively. Also, the normalized mean fluorescence intensity obtained from PNT1A
cell was 0.96 ± 0.02 for the control condition and 0.94 ± 0.02 for the proposed driving
condition, respectively. It was found that there was slightly decrease in normalized mean
fluorescence intensity, but no significantly different between control condition and
proposed driving condition for RBCs (p-value: 0.29 > 0.05, n=10) and PNT1A cells (p-
value: 0.15 > 0.05, n=10). Based on the cell viability results, this approach manipulates
every single cell without compromising cell viability under the influence of high
frequency ultrasound microbeam with the proposed driving condition.
83
Figure 5.6 Cell viability study was performed, and normalized mean fluorescence
intensity was presented. Both normalized mean fluorescence intensity were slightly
decrease, but not significantly different between control condition and the proposed
driving condition for (a) RBCs (p-value: 0.29 > 0.05, n=10) and (b) PNT1A cells (p-
value: 0.15 > 0.05, n=10).
5.4 Discussions
In this chapter, a tightly focused high frequency ultrasound microbeam based
label-free cell separation approach was developed, providing unique characteristics such
as being easy-to-use and cost-effective as well as high specificity without compromising
normal cell physiology at single-cell level. To shed light on a rational strategy for the
separation of microscopic objects, an approach for independently measuring ultrasound
backscattering coefficient of the trapped the individual single object was adopted. In
order to reach the objective, a custom-built all-in-one front-end system was developed to
trap a polystyrene bead and move along the direction of the high frequency ultrasound
microbeam, and tightly focused high frequency ultrasonic transducer and an impedance
matching network were fabricated to generate very narrow and sharp high frequency
(a)
(b)
84
ultrasound microbeam. After trapping performance was confirmed, size determination of
a trapped micron-sized particle and separation between red blood cells (RBCs) and
normal SV40 immortalized epithelial prostate (PNT1A) cells were successfully
conducted.
To develop a high frequency ultrasound microbeam based label-free cell
separation approach as viable and adaptable separation method, it will be necessary to
extend to elucidate how different physical properties such as shape and elasticity on top
of size influences the levels of ultrasound backscattering coefficient. For example,
according to the results, the proposed approach might separate between polystyrene
microbeads (6 µm - 8 µm) and RBCs (6 µm - 8 µm) based on estimated ultrasound
backscattering coefficient from -104 dB to -107 dB for microbeads and measured
ultrasound backscattering coefficient from -108.5 dB to -110 dB for RBCs even though
they have very similar size. However, the capability to distinguish between various
objects having several physical properties should be confirmed in the near future.
Theoretically, acoustic trapping with tightly focused high frequency ultrasound
microbeam was demonstrated by the change of the momentum of incident beams in the
interaction between a gradient force and a scattering force (Lam et al., 2013; Lee et al.,
2009; Lee et al., 2010; Li et al., 2014; Yoon et al., 2014). Trapping performance can be
stable when the gradient force pulling a particle towards the beam focus exceeds the
scattering force pushing the particle away from the focus in the direction of the incident
beam. Three models for acoustic force calculation were explained by a correlation
between the diameter of the trapped particle (D) and wavelength of used ultrasound beam
85
(Ashkin, 1992; Lam et al., 2016; Nieminen et al., 2007). First, when the diameter of the
trapped particle (D) is greater than the wavelength of used ultrasound beam (λ), i.e. (D >
λ), this condition referred to as the Mie regime. Second, when the diameter of the trapped
particle (D) is smaller than the wavelength of used ultrasound beam (λ), i.e. (D < λ), this
condition referred to as the Rayleigh regime. Third, when the diameter of the trapped
particle (D) is very similar to wavelength of used ultrasound beam (λ), i.e. (D ≈ λ), this
condition referred to as the intermediate Mie regime. In terms of the size of a trapped
single object and wavelength of high frequency ultrasound microbeam (10 μm), this
approach simultaneously worked well with Mie regime (microspheres: 5 μm) / RBCs: 6
μm – 8 μm), Rayleigh regime (PNT1A cells: 12 μm) / microspheres: 50 μm), and
intermediate Mie regime (microspheres: 10 μm) / PNT1A cells: 9 μm – 11 μm). These
three regimes could provide an intuitive understanding of the physical principles of
acoustic trapping; however, it will be necessary to extend our study for the quantitative
and qualitative descriptions in the near future.
A comparative study on the influences of three driving conditions in the burst
length range from long pulses to one pulse while maintaining the same duty factor was
conducted as presented in Figure 5.2(e)-(g). Longer burst pulses (on time: 1μs / PRF: 1
kHz) contributed to higher impacts linked with distances between a trapped bead and its
adjacent beads in Figure 5.2(e), and stronger acoustic trapping force. On the other hands,
proposed shorter burst pulses (on time: 6.7 ns / PRF: 167 kHz) generated lower
influences related to distances between a trapped particle and its surrounding particles in
Figure 5.2(g), and weaker acoustic trapping force. It can be estimated that the effects of
86
longer burst pulses (on time: 1μs / PRF: 1 kHz) on a targeted cell with neighboring cells
are likely to be at least ten times stronger than effects of the proposed shorter burst pulses
(on time: 6.7 ns / PRF: 167 kHz) based on measurement of acoustic trapping force if the
same number of acoustic pulses during the same time was applied. Therefore, the
proposed driving condition has significant advantages in specifically controlling cell
functions and responses with the capability of exclusively localizable single-cell targeting
without significantly affecting surrounding cells.
In this approach, separation between 5 μm and 10 μm polystyrene microspheres
was successfully conducted by simultaneously investigating size of particles, and at the
same time, trapping the particles from the exposure to high frequency ultrasound
microbeam with the proposed driving conditions. Figure 5.3 shows separating
performance was clearly shown by the statistical analysis based on the calculated and
averaged ultrasound backscattering coefficients of two different polystyrene particles. To
further demonstrate the ability to separate cells using high frequency ultrasound
microbeam, two different cells such as red blood cells (RBCs) and normal SV40
immortalized epithelial prostate (PNT1A) cells were tested. Figure 5.5 shows that
differentiating between RBCs (6 μm – 8 μm) and PNT1A cells (9 μm – 12 μm) was
successfully achieved based on different levels of ultrasound backscattering coefficients
and extremely significant from the statistical analysis. From the cell viability studies,
every single cell could be manipulated under the influence of high frequency ultrasound
microbeam with the proposed driving condition without compromising cell viability.
These results suggest that the tightly focused high frequency ultrasound microbeam based
87
label-free cell separation approach was able to specifically discriminate between different
sized single-cells without affecting normal cell physiology.
This label-free approach using tightly focused high frequency ultrasound
microbeam with safe and simple controllable driving conditions can contribute to a good
stepping stone for investigations into unique expression profiles between individual cells
at the single-cell level. For example, manipulation of intercellular communications
between an engineered single cell and adjacent cells provides an insight into potentially
undetected cellular signal pathway and responses. Thus, the proposed approach enables
to create rational strategies capable of simultaneously performing acoustic trapping of a
suspended single-cell, measuring the physical property of the trapped cell based on
ultrasonic backscattering coefficients, acoustic transfection of the trapped cell, and
acoustic manipulation of the transfected and trapped cell. Once the rational strategies are
systematically developed, these approaches may be formulated and subsequently
translated to clinical medicine for the treatment of numerous human diseases such as
neurodegenerative diseases and inflammatory diseases.
88
CHAPTER 6 Summary and Future Works
6.1 Summary
This dissertation presents single-cell analysis with high frequency ultrasound, and
covers custom-built high frequency ultrasound electronics for single-cell analysis and
single-cell analysis.
The utilization of impedance matching network for high frequency ultrasonic
transducers enables to maximize energy transmission from the excitation source to the
ultrasonic transducers and improve sensitivity of the ultrasonic transducers. Designed
impedance matching network integrated with the ultrasonic transducer can effectively
generate a stronger acoustic pressure field at its focal area compared to an acoustic
pressure field produced by the ultrasonic transducer without impedance matching
network, and thus cell responses can be controlled by adjusting treatment conditions
according to the purposes of the applications.
A high frequency ultrasound system can be used to image smaller objects like
axon in the nerve bundles or single-cell structures. The measured axial and lateral
resolution of the wire phantom images were 17 µm and 7 µm respectively, which is close
to single-cell size, and can be much more improved by using frequencies up to 600 MHz..
Also, the system can be employed in ultrasound biomicroscopy for generating ultrahigh
frequency (> 300 MHz) coded excitation, which provides better visualization of cross-
sectional imaging.
89
A new transfection technique is proposed with the hypothesis that a tightly
focused high frequency ultrasound physical stretches cell lipid bilayer on cell membrane
and generate transient and reversible holes on the membrane. The concept of the
proposed approach was confirmed by the intracellular delivery of small-sized molecules,
macromolecules, and even two molecules. By determining the optimal treatment
conditions, this approach can be developed as a versatile and viable transfection tool to
transport molecules and particles across cell membranes in a safe and efficient manner.
The highly focused high frequency ultrasound beam generated by high pulse
repetition frequency monocycle bipolar pulse can be applied to separating microscopic
objects with unique characteristics, including it being an easy-to-use, cost-effective,
label-free investigation of physical properties with minimal effects on normal cell
physiology with a much more accurate analysis at the single-cell level. Separating two
different micron-sized particles, and differentiating red blood cells and cancer cells by
individually identifying the size of the trapped single-object based on ultrasound
backscattering coefficient were successfully achieved.
6.2 Future Works
6.2.1 A Strategy for Multifunctional Single-cell Analysis
The ultimate goal is to create rational strategies for single-cell analysis with high
frequency ultrasound capable of simultaneously performing acoustic trapping of
90
suspended single-cells, measuring the cellular properties of the trapped cells based on
ultrasonic backscattering, acoustic transfection of the cells, and acoustic tweezers of the
transfected cells to elucidate and understand cellular behaviors between individual cells
and intercellular communications between an engineered cell and adjacent cells.
6.2.2 Investigation and manipulation of cell membrane electrical impedance
during Acoustic Transfection
In addition to observation of the intracellular delivery of various sized molecules,
as a quantitative approach, the measurement of cell membrane electrical impedance
during acoustic transfection is crucial for a better understanding of the hypothesis of
acoustic transfection. Figure 6.1 illustrates the preliminary experimental setup composed
of the ultrasonic transducer with impedance matching network and patch-clamp
micropipette connected with EPC-9 patch-clamp amplifier with PULSE software (HEKA
Elektronik, Lambrecht, Germany). By examining current-voltage responses and electrical
impedance changes (Cm, Gm and Gs) of cell membrane during patch-clamp recording of
acoustic transfected cell, the hypothesis of acoustic transfection can be proved, and
further manipulation of cellular functions and mechanisms through creating a physical
pathway in which transient and reversible holes were generated on the cell membranes
can be achieved by controlling depolarizing current and action potential using acoustic
transfection.
91
Figure 6.1 Experimental setup consists of the ultrasonic transducer with IMN and patch-
clamp micropipette connected with EPC-9 patch-clamp amplifier with PULSE software.
92
BIBLOGRAPHY
Aerts, J. T., Louis, K. R., Crandall, S. R., Govindaiah G., Cox, C. L., and J. V. Sweedler.
(2014). Patch clamp electrophysiology and capillary electrophoresis-mass spectrometry
metabolomics for single cell characterization, Anal. Chem, 86, 3203-3208.
Anderson, B. (2007). Echocardiography: the normal examination and echocardiographic
measurements, MGA Graphics.
Andersen, J., and L. Wilkins. (1977) The Design of Optimum Lumped Broadband
Equalizers for Ultrasonic Transducers, IEEE Ultrasonics Symposium, 422-427.
Antkowiak, M., Torres-Mapa, M.L., Stevenson, D.J., Dholakia, K., and F. J. Gunn-
Moore. (2013) Femtosecond optical transfection of individual mammalian cells, Nat
Protoc, 8, p. 1216-1233.
Ashkin, A., 1992, Forces of a single-beam gradient laser trap on a dielectric sphere in the
ray optics regime, Biophys J, 61, 569-582.
Atienza, J. M., Yu, N., Kirstein, S. L., Xi, B., Wang, X., Xu, X., and Y. A. Abassi. (1992).
Forces of a single-beam gradient laser trap on a dielectric sphere in the ray optics regime,
Assay Drug Dev Technol, 4, 569-582.
Augustine, L. J., and J. Andersen. (1979). An algorithm for the design of transformerless
broadband equalizers of ultrasonic transducers, The Journal of the Acoustical Society of
America, 66, 629-635.
Aydin, N., L. Fan, and D. H. Evans. (1994). Quadrature-to-directional format conversion
of Doppler signals using digital methods, Physiol Meas, 15, 181-199.
Bakkers, J. (2011). Zebrafish as a model to study cardiac development and human cardiac
disease, Cardiovasc Res, 91, 279-288.
Ben-Yaakov, and S. Krihely. (2004) Modeling and driving piezoelectric resonant blade
elements, Applied Power Electronics Conference and Exposition, 3, 1733-1739.
Bergmann, O., R. D. Bhardwaj, S. Bernard, S. Zdunek, F. Barnabé-Heider, S. Walsh, J.
Zupicich, K. Alkass, B. A. Buchholz, H. Druid, S. Jovinge, and J. Frisén. (2009)
Evidence for cardiomyocyte renewal in humans, Science, 324, 98-102.
Bersell, K., S. Arab, B. Haring, and B. Kühn. (2009). Neuregulin1/ErbB4 signaling
induces cardiomyocyte proliferation and repair of heart injury, Cell, 138, 257-270.
Bhuiyan, R., Islam, M. and H. Huang. (2012) Wireless excitation and electrical
impedance matching of piezoelectric wafer active sensors, ASME Conference on Smart
Materials, Adaptive Structures and Intelligent Systems, 2, 271-276.
93
Brown, J. A., F. S. Foster, A. Needles, E. Cherin, and G. R. Lockwood. (2007)
Fabrication and performance of a 40-MHz linear array based on a 1-3 composite with
geometric elevation focusing, IEEE Trans Ultrason Ferroelectr Freq Control, 54, 1888-
1894.
Cannata, J.M., Ritter, T.A., Chen, W, and K. K. Shung. (2001). Design of focused single
element (50-100 MHz) transducers using lithium niobate, Medical Imaging, Proc. SPIE,
4325, 28-35.
Cannata, J.M., Ritter, T.A., Chen, W.H, Silverman, R.H., and K. K. Shung. (2003).
Design of efficient, broadband single-element (20-80 MHz) ultrasonic transducers for
medical imaging applications, IEEE Transactions on Ultrasonics, Ferroelectrics and
Frequency Control, 50, 1548-1557.
Cannata, J. M., J. A. Williams, Q. Zhou, T. A. Ritter, and K. K. Shung. (2006).
Development of a 35-MHz piezo-composite ultrasound array for medical imaging, IEEE
Trans Ultrason Ferroelectr Freq Control, 53, 224-236.
Cannata, J. M., J. A. Williams, L. Zhang, C. H. Hu, and K. K. Shung. (2011). A high-
frequency linear ultrasonic array utilizing an interdigitally bonded 2-2 piezo-composite:
IEEE Trans Ultrason Ferroelectr Freq Control, 58, 2202-2212.
Chablais, F., J. Veit, G. Rainer, and A. Jaźwińska. (2011). The zebrafish heart
regenerates after cryoinjury-induced myocardial infarction, BMC Dev Biol, 11, 21.
Chattopadhyay, P. K., Gierahn, T. M., Roederer, M. and J. C. Love. (2014). Single-cell
technologies for monitoring immune systems. Nat. Immunol., 15, 128-135.
Chen, R., N. E. Cabrera-Munoz, K. H. Lam, H. S. Hsu, F. Zheng, Q. Zhou, and K. K.
Shung (2014) PMN-PT single-crystal high-frequency kerfless phased array, IEEE Trans
Ultrason Ferroelectr Freq Control, 61, 1033-1041.
Chen, X., and S. T. C. Wong. (2005). Automated dynamic cellular analysis in high
throughput drug screens, Circuits and Systems, IEEE International Symposium on, 4229-
4232.
Chen, Z. (2010). Small-molecule delivery by nanoparticles for anticancer therapy, Trends
Mol. Med., 16, 594-602.
Cho, S. H., Godin, J. M., Chen, C. H., Qiao, W., Lee, H., and Y. H. Lo. (2010). Review
Article: Recent advancements in optofluidic flow cytometer, Biomicrofluidics, 4, 4300.
Choi, H., Kim, M. G., and K. K. Shung. (2012). New MOSFET-based expander for high
frequency ultrasound systems, Ultrasonics Symposium (IUS), IEEE International, 623-
626.
94
Choi, H., Kim, M. G., Cummins, T. M., Hwang, J. Y., and K. K. Shung. (2014). Power
MOSFET-diode-based limiter for high-frequency ultrasound systems, Ultrasonic
Imaging, 36, 317-330..
Christopher, D. A., P. N. Burns, J. Armstrong, and F. S. Foster. (1996). A high-frequency
continuous-wave Doppler ultrasound system for the detection of blood flow in the
microcirculation, Ultrasound Med Biol, 22, p. 1191-1203.
Christopher, D. A., P. N. Burns, B. G. Starkoski, and F. S. Foster. (1997). A high-
frequency pulsed-wave Doppler ultrasound system for the detection and imaging of blood
flow in the microcirculation, Ultrasound Med Biol, 23, 997-1015.
Crowley, T. A., and V. Pizziconi. (2005). Isolation of plasma from whole blood using
planar microfilters for lab-on-a-chip applications, Lab Chip, 5, 922-929.
Del Fattore, A., Peruzzi, B., Peruzzi, B., Rucci, N., Recchia, I., Cappariello, A., Longo,
M., Fortunati, D., Ballanti, P., Iacobini, M., Luciani, M., Devito, R., Pinto, R., Caniglia,
M., Lanino, E., Messina, C., Cesaro, S., Letizia, C., Bianchini, G., Fryssira, H.,
Grabowski, P., Shaw, N., Bishop, N., Hughes, D., Kapur, R. P., Datta, H. K., Taranta, A.,
Fornari, R., Migliaccio, S., and A. Teti. (2006). Clinical, genetic, and cellular analysis of
49 osteopetrotic patients: implications for diagnosis and treatmentm, J Med Genet, 43,
315-325.
Deramo, V. A., G. K. Shah, C. R. Baumal, M. S. Fineman, Z. M. Corrêa, W. E. Benson,
C. J. Rapuano, E. J. Cohen, and J. J. Augsburger. (1999). Ultrasound biomicroscopy as a
tool for detecting and localizing occult foreign bodies after ocular trauma,
Ophthalmology, 106, 301-305.
Ding, X., Lin, S. C., Kiraly, B., Yue, H., Li, S., Chiang, I. K., Shi, J., Benkovic, S. J., and
T. J. Huang. (2012). On-chip manipulation of single microparticles, cells, and organisms
using surface acoustic waves, Proc. Natl. Acad. Sci, 109, 11105-11109.
Ding, X., Peng, Z., Lin S-C. S., Geri, M., Li, S., Li, P., Chen, Y., Dao, M., Suresh, S., and
T. J. Huang. (2014). Cell separation using tilted-angle standing surface acoustic waves,
Proc. Natl. Acad. Sci, 111, 12992-12997.
Duffer, C. E., amd C. P. Burger. (1995). Narrow Band Laser Ultrasonic NDE, Review of
Progress in Quantitative Nondestructive Evaluation, 15, 593-600.
Escoffre, J. M., Hubert, M., Dean, D. S., Rols, M. P., and C. Favard. (2007). Membrane
perturbation by an external electric field: a mechanism to permit molecular uptake, Eur
Biophys J, 36, 973-983.
Falou, O., Rui, M., El Kaffas, A., Kumaradas, J. C. and M. C. Kolios. (2010). The
measurement of ultrasound scattering from individual micron-sized objects and its
application in single cell scattering, J Acoust Soc Am, 128, 894-902.
95
Fan, Z., Kumon, R. E., Park, J., and C. X. Deng. (2010). Intracellular delivery and
calcium transients generated in sonoporation facilitated by microbubbles, J. Control.
Release, 142, 31-39.
Fan, Z., Liu, H., Mayer, M., and C. X. Deng. (2012) Spatiotemporally controlled single
cell sonoporation, Proc Natl Acad Sci U S A, 109, 16486-16491.
Fan, Z., Liu, H., Mayer, M., C. X. Deng. (2013). Improving ultrasound gene transfection
efficiency by controlling ultrasound excitation of microbubbles, J. Control. Release, 170,
401-413.
Foster, F. S., C. J. Pavlin, K. A. Harasiewicz, D. A. Christopher, and D. H. Turnbull.
(2000). Advances in ultrasound biomicroscopy, Ultrasound Med Biol, 26, 1-27.
Foster, F. S., M. Y. Zhang, Y. Q. Zhou, G. Liu, J. Mehi, E. Cherin, K. A. Harasiewicz, B.
G. Starkoski, L. Zan, D. A. Knapik, and S. L. Adamson. (2002). A new ultrasound
instrument for in vivo microimaging of mice, Ultrasound Med Biol, 28, 1165-1172.
Gao, Y., Li, W. J. and D. Pappas. (2013). Recent advances in microfluidic cell
separations, Analyst, 138, 4714-4721.
Garcí a-Feijoó, J., J. L. Hernández-Matamoros, C. Méndez-Hernández, A. Castillo-
Gómez, C. Lázaro, T. Martí n, R. Cuiña-Sardiña, and J. Garcí a-Sánchez. (2003).
Ultrasound biomicroscopy of silicone posterior chamber phakic intraocular lens for
myopia, J Cataract Refract Surg, 29, 1932-1939.
Glover, D. J., Lipps, H. J., and Jans, D. A. (2005) Towards safe, non-viral therapeutic
gene expression in humans, Nat Rev Genet, 6, 299-310
Gonzalez, G. (1996). Microwave Transistor Amplifier Analysis and Design 2nd edition,
New Jersey: Prentice Hall.
González-Rosa, J. M., V. Martí n, M. Peralta, M. Torres, and N. Mercader. (2011).
Extensive scar formation and regression during heart regeneration after cryoinjury in
zebrafish, Development, 138, 1663-1674.
Gossett, DR., Weaver, W. M., Mach, A. J., Hur, S. C., Tse, H. T., Lee, W., Amini, H.,
and D. Di Carlo. (2010). Label-free cell separation and sorting in microfluidic systems,
Analytical and bioanalytical chemistry, 397, 3249-3267.
Grego-Bessa, J., L. Luna-Zurita, G. del Monte, V. Bolós, P. Melgar, A. Arandilla, A. N.
Garratt, H. Zang, Y. S. Mukouyama, H. Chen, W. Shou, E. Ballestar, M. Esteller, A.
Rojas, J. M. Pérez-Pomares, and J. L. de la Pompa. (2007). Notch signaling is essential
for ventricular chamber development, Dev Cell, 12, 415-29.
Gross, A., Schoendube, J., Zimmermann, S., Steeb, M., Zengerle, R., and P. Koltay.
(2015). Technologies for single-cell isolation, Int J Mol Sci. 16, 16897-16919.
96
Guo, F., Li, P., French, J. B., Mao, Z., Zhao, H., Li, S., Nama, N., Fick, J. R., Benkovic,
S. J., and T. J. Huang. (2015). Controlling cell–cell interactions using surface acoustic
waves, Proc. Natl. Acad. Sci, 112, 43-11248.
Guzman, H. R., Nguyen, D. X., Khan, S., and M. R. Prausnitz. (2001). Ultrasound-
mediated disruption of cell membranes. I. Quantification of molecular uptake and cell
viability, J Acoust Soc Am, 110, 588-596.
Guzman, H. R., Nguyen, D. X., McNamara, A. J., and M. R. Prausnitz. (2002).
Equilibrium loading of cells with macromolecules by ultrasound: effects of molecular
size and acoustic energy, J Pharm Sci, 91, 1693-1701.
Hacein-Bey-Abina, S., Le Deist, F., Carlier, F., Bouneaud, C., Hue, C., De Villartay, J. P.,
Thrasher, A. J., Wulffraat, N., Sorensen, R., Dupuis-Girod, S., Fischer, A., Davies, E. G.,
Kuis, W., Leiva, L., and M. Cavazzana-Calvo. (2002). Sustained correction of X-linked
severe combined immunodeficiency by ex vivo gene therapy, N Engl J Med, 346, 1185-
1193.
Hahn, C., and M. A. Schwartz. (2008). The role of cellular adaptation to mechanical
forces in atherosclerosis, Arterioscler Thromb Vasc Biol, 28, 2101-2107.
Heath J. R., Ribas A., and P. S. Mischel. (2016). Single-cell analysis tools for drug
discovery and development, Nat Rev Drug Discov., 15, 204-216.
Herce, H. D., Deng, W., Helma, J., Leonhardt, H. and M. C. Cardoso. Visualization and
targeted disruption of protein interactions in living cells, Nat. Commun, 4, 2660.
Herickhoff, C. D., Grant, G. A., Britz, G. W., and S. W. Smith. (2010). Dual-mode
intracranial catheter integrating 3-D ultrasound imaging and hyperthermia for neuro-
oncology: Feasibility study, IEEE Transactions on Ultrasonics, Ferroelectrics and
Frequency Control, 57, 2572-2584.
Ho, Y. L., Y. W. Shau, H. J. Tsai, L. C. Lin, P. J. Huang, and F. J. Hsieh. (2002).
Assessment of zebrafish cardiac performance using Doppler echocardiography and power
angiography, Ultrasound Med Biol, 28, 1137-1143.
Hoarau, C., Bailly, P.E., Arnould, J. D., Ferrari, P., and P. and P. Xavier. (2007). A RF
tunable impedance matching network with a complete design and measurement
methodology, Microwave Conference, 751-754.
Hozumi, N., R. Yamashita, C. K. Lee, M. Nagao, K. Kobayashi, Y. Saijo, M. Tanaka, N.
Tanaka, and S. Ohtsuki. (2004). Time-frequency analysis for pulse driven ultrasonic
microscopy for biological tissue characterization, Ultrasonics, 42, 717-22.
Hsieh, P. C., V. F. Segers, M. E. Davis, C. MacGillivray, J. Gannon, J. D. Molkentin, J.
Robbins, and R. T. Lee. (2007). Evidence from a genetic fate-mapping study that stem
cells refresh adult mammalian cardiomyocytes after injury, Nat Med, 13, 970-974.
97
Hu, C., L. Zhang, J. M. Cannata, J. Yen, and K. K. Shung. (2011). Development of a 64
channel ultrasonic high frequency linear array imaging system, Ultrasonics, 51, 953-959.
Hu, P., Zhang, W., Xin, H. and G. Deng. (2016). Single cell isolation and analysis, Front.
Cell Dev. Biol, 4, 116.
Huang, Y., and K. Boyle. (2008). Antennas: From Theory to Practice, New York: Wiley.
Huang, Y., Joo, S., Duhon, M., Heller, M., Wallace, B., and X. Xu. (2002).
Dielectrophoretic cell separation and gene expression profiling on microelectronic chip
arrays, Anal Chem, 74, 3362-3371.
Hunt, J. W., Worthington, A. E., Xuan, A., Kolios, M. C., Czarnota, G. J., and M. D.
Sherar. (2002). A model based upon pseudo regular spacing of cells combined with the
randomisation of the nuclei can explain the significant changes in high-frequency
ultrasound signals during apoptosis. Ultrasound Med. Biol. 28, 217-226.
Hwang, J. Y., Lee, N .S., Lee, C., Lam, K. H., Kim, H. H., Woo, J., Lin, M. Y., Kisler, K.,
Choi, H., Zhou, Q., Chow, R. H., and K. K. Shung. (2013). Investigating contactless high
frequency ultrasound microbeam stimulation for determination of invasion potential of
breast cancer cells, Biotechnol Bioeng, 110, 2697-2705.
Hwang, J. Y., Lim, H. G., Yoon, C. W., Lam, K. H., Yoon, S., Lee, C., Chiu, C. T., Kang,
B. J., and K. K. Shung. (2014). Non-contact high-frequency ultrasound microbeam
stimulation for studying mechanotransduction in human umbilical vein endothelial cells,
Ultrasound in Medicine and Biology, 40, 2172-2182.
Hwang, J. Y., Lee, C., Lam, K. H., Kim, H. H., Lee, J., and K. K. Shung. (2014). Cell
membrane deformation induced by a fibronectin-coated polystyrene microbead in a 200-
MHz acoustic trap, IEEE Transactions on Ultrasonics, Ferroelectrics and Frequency
Control, 61, 399-406.
Hwang, J. Y., Yoon, C. W., Lim, H. G., Park, J., Yoon, S., Lee, J., and K. K. Shung.
(2015). Acoustic tweezers for studying intracellular calcium signaling in SKBR-3 human
breast cancer cells, Ultrasonics, 63, 94-101.
Ilumoka, A., and R. Spence. (1988). Parameter tolerance design for electrical circuits,
Quality and Reliability Engineering International, 4, 87-94.
Jensen, J. A.. (1996). Estimation of blood velocities using ultrasound: a signal processing
approach New York, NY, Cambridge University Press.
Karshafian, R., Bevan, P.D., Williams, R., Samac, S., Burns, P.N., 2009, Sonoporation by
ultrasound-activated microbubble contrast agents: effect of acoustic exposure parameters
on cell membrane permeability and cell viability, Ultrasound Med Biol, 35, 847-860.
98
Kawamura, T., Suzuki, J., Wang, Y. V., Menendez, S., Morera, L. B., Raya, A., Wahl, G.
M., and J. C. Izpisúa Belmonte. (2009). Linking the p53 tumour suppressor pathway to
somatic cell reprogramming, Nature, 460, 1140-1144.
Kikuchi, K., J. E. Holdway, A. A. Werdich, R. M. Anderson, Y. Fang, G. F. Egnaczyk, T.
Evans, C. A. Macrae, D. Y. Stainier, and K. D. Poss, 2010, Primary contribution to
zebrafish heart regeneration by gata4(+) cardiomyocytes, Nature, 464, 601-605.
Kim, D., Kim C.H., Moon J. I., Chung, Y. G., Chang M. Y., Han, B. S., Ko. S., Yang, E.,
Cha, K. Y., Lanza, R., and K. S. Kim. (2009). Generation of human induced pluripotent
stem cells by direct delivery of reprogramming proteins, Cell Stem Cell, v. 4, p. 472-476.
Kim, D. Y., D. Z. Reinstein, R. H. Silverman, D. J. Najafi, S. C. Belmont, A. P. Hatsis, G.
W. Rozakis, and D. J. Coleman, 1998, Very high frequency ultrasound analysis of a new
phakic posterior chamber intraocular lens in situ: Am J Ophthalmol, v. 125, p. 725-9.
Kim, H. H., J. M. Cannata, J. A. Williams, J. H. Chang, and K. K. Shung, 2009,
Fabrication of 20 MHz Convex Array Transducers for High Frequency Ophthalmic
Imaging: Ultrasonics Symposium (IUS), 2009 IEEE International, p. 1130-1133.
Kim, H. H., C.-H. Hu, J. Park, B. J. Kang, J. A. Williams, J. M. Cannata, and K. K.
Shung, 2010, Characterization and Evaluation of High Frequency Convex Array
Transducers: Ultrasonics Symposium (IUS), 2010 IEEE International, p. 650-653.
Kim, M. G., H. Choi, H. H. Kim, and K. K. Shung, 2013, Bipolar pulse generator for
very high frequency (> 100 MHz) ultrasound applications, Ultrasonics Symposium (IUS),
2013 IEEE International, 1567-1570.
Kim, M. G., Yoon, S., Kim, H. H., and K. K. Shung. (2016). Impedance matching
network for high frequency ultrasonic transducer for cellular applications, Ultrasonics, 65,
258-267.
Kim, T.K, and J. H. Eberwine. (2010). Mammalian cell transfection: the present and the
future, Anal Bioanal Chem, 397, 3173-3178.
Kong, C.S. (1995). A general maximum power transfer theorem, IEEE Transactions on
Education, 38, 296-298.
Kuang, Y., Jin, Y., Cochran, S., and Z. Huang. (2014). Resonance tracking and vibration
stablilization for high power ultrasonic transducers, Ultrasonics, 54, 187-194.
Kuffler D. P. (2010). Ultrasound imaging of regenerating rat sciatic nerves in situ,
Journal of Neuroscience Methods, 188, 276-279.
Kumar, A. and A. Bhardwaj. (2008). Methods in cell separation for biomedical
application: cryogels as a new tool, Biomed Mater, 3, 034008.
99
Lam, K. H., Hsu, H.-S., Li, Y., Lee, C., Lin, A., Zhou, Q., Kim, E. S., and K. K. Shung.
(2013). Ultrahigh frequency lensless ultrasonic transducers for acoustic tweezers
application, Biotechnology and Bioengineering, 110, 881-886.
Lam, K. H., Ji, H.., Zheng, F., Ren, W., Zhou, Q., and K. K. Shung. (2013). Development
of lead-free single-element ultrahigh frequency (170-320 MHz) ultrasonic transducers,
Ultrasonics, 53, 1033-1038.
Lam, K. H., Li, Y., Li, Y., Lim, H. G., Zhou, Q., and K. K. Shung. (2016).
Multifunctional single beam acoustic tweezer for non-invasive cell/organism
manipulation and tissue imaging, Scientific Reports, 6, 37554.
Larina, I. V., Evers, B. M., and R. O. Esenaliev. (2005). Optimal drug and gene delivery
in cancer cells by ultrasound-induced cavitation, Anticancer Res, 25, 149-156.
Lawson, D. A., Bhakta, N. R., Kessenbrock, K., Prummel, K. D., Yu, Y., Takai, K., Zhou,
A., Eyob, H., Balakrishnan, S., Wang, C. Y., Yaswen, P., Goga, A., and Z. Werb. (2015).
Single-cell analysis reveals a stem-cell program in human metastatic breast cancer cells,
Nature, 526, 131-135.
Lassau, N., A. Spatz, M. F. Avril, A. Tardivon, A. Margulis, G. Mamelle, D. Vanel, and J.
Leclere. (1997). Value of high-frequency US for preoperative assessment of skin tumors,
Radiographics, 17, 1559-1565.
Leader, B., Baca, Q. J. and D. E. Golan. (2008). Protein therapeutics: A summary and
pharmacological classification, Nat Rev Drug Discov, 7, 21-39.
Lee, C., Jung, H., Lam, K. H., Yoon, C., and K. K. Shung. (2015). Ultrasonic scattering
measurements of a live single cell at 86 MHz, IEEE Trans Ultrason Ferroelectr Freq
Control, 62, 1968-78.
Lee, J., Teh, S. Y., Lee, A., Kim, H. H., Lee, C., and K. K. Shung. (2009). Single beam
acoustic trapping, Applied Physics Letters, 95, 1-3.
Lee, J., Teh, S. Y., Lee, A., Kim, H. H., Lee, C., and K. K. Shung. (2010). Transverse
acoustic trapping using a gaussian focused ultrasound, Ultrasound in Medicine and
Biology, 36, 350-355.
Lee, J., Lee, C., Kim, H. H., Jakob, A., Lemor, R., Teh, S.-Y., Lee, A., and K. K. Shung.
(2011). Targeted cell immobilization by ultrasound microbeam, Biotechnology and
Bioengineering, 108, 1643-1650.
Li, Y., Lee, C., Chen, R., Zhou, Q., and K. K. Shung. (2014). A feasibility study of in
vivo applications of single beam acoustic tweezers, Applied Physics Letters, 105, 173701.
100
Li, Y. L., Figueroa, D. G., Rodriguez, J. P., Huang, L., Liao, J. C., M. Taniguch, Canner,
J., and T. Kondo. (1998). A new technique for high frequency characterization of
capacitors, Electronic Components & Technology Conference, 1384-1390.
Liang, H. D., and M. J. Blomley. (2003). The role of ultrasound in molecular imaging, Br.
J. Radiol., 76,140-150.
Lien, C. L., M. R. Harrison, T. L. Tuan, and V. A. Starnes. (2012). Heart repair and
regeneration: recent insights from zebrafish studies, Wound Repair Regen, 20, 638-646.
Lim, H. G., Li, Y., Lin, M. Y., Yoon, C., Lee, C., Jung, H., Chow, R. H., and K. K.
Shung. (2016). Calibration of trapping force on cell-size objects from ultrahigh-frequency
single-beam acoustic tweezer, IEEE transactions on Ultrasonics, Ferroelectrics, and
Frequency control, 63, 1988-1995.
Liu, Y., Sonek, G. J., Berns, M. W., and B. J. Tromberg. (1996). Physiological
monitoring of optically trapped cells: assessing the effects of confinement by 1064-nm
laser tweezers using microfluorometry, Biophysical Journal, 71, 2158-2167.
Lockwood, G. R., D. H. Turnball, D. A. Christopher, and F. S. Foster. (1996). Beyond 30
MHz: Applications of high-frequency ultrasound imaging, IEEE Engineering in Medicine
and Biology Magazine, 12.
Lodish, H., Berk, A., Zipursky, S. L., Matsudaira, P., Baltimore D., and J. Darnell. (2000).
Molecular Cell Biology 4th edition, New York: W. H. Freeman.
Loupas, T., J. T. Powers, and R. W. Gill, 1995, An axial velocity estimator for ultrasound
blood flow imaging, based on a full evaluation of the Doppler equation by means of a
two-dimensional autocorrelation approach: Ultrasonics, Ferroelectrics, and Frequency
Control, IEEE Transactions on, v. 42, p. 672-688.
Lukacs, M., J. Yin, G. Pang, R. C. Garcia, E. Cherin, R. Williams, J. Mehi, and F. S.
Foster. (2006). Performance and characterization of new micromachined high-frequency
linear arrays, IEEE Trans Ultrason Ferroelectr Freq Control, 53, 1719-1729.
Marigo, F. A., K. Esaki, P. T. Finger, H. Ishikawa, D. S. Greenfield, J. M. Liebmann, and
R. Ritch. (1999). Differential diagnosis of anterior segment cysts by ultrasound
biomicroscopy, Ophthalmology, 106, 2131-2135.
McCloy, R. A., Rogers, S., Caldon, C. E., Lorca, T., Castro, A., and A. Burgess. (2014).
Partial inhibition of Cdk1 in G 2 phase overrides the SAC and decouples mitotic events,
Cell Cycle, 13, 1400-1412.
Mehier-Humbert, S., and R. H. Guy. (2005). Physical methods for gene transfer:
improving the kinetics of gene delivery into cells, Adv Drug Deliv Rev, 57, 733-753.
101
Merks, E. J. W., N. Bom, A. Bouakaz, N. de Jong, and A.F.W. van der Steen, 2004,
Design of a multi-layer transducer for acoustic bladder volume assessment, Ultrasonics
Symposium (IUS), 2004 IEEE International, 145-148.
Milan, D. J., I. L. Jones, P. T. Ellinor, and C. A. MacRae, 2006, In vivo recording of
adult zebrafish electrocardiogram and assessment of drug-induced QT prolongation: Am
J Physiol Heart Circ Physiol, v. 291, p. H269-73.
Miltenyi, S., Müller, W., Weichel, W., and A. Radbruch. (1990). High gradient magnetic
cell separation with MACS. Cytometry, 11, 231-238.
Mthunzi, P., Dholakia, K., and F. Gunn-Moore. (2010). Phototransfection of mammalian
cells using femtosecond laser pulses: optimization and applicability to stem cell
differentiation, J Biomed Opt, 15, 041507.
Neuman, K. C., Chadd, E. H., Liou, G. F., Bergman, K., and S. M. Block. (1999).
Characterization of photodamage to Escherichia coli in optical traps, Biophysical Journa ,
77, 2856-2863.
Nieminen, T., Knoner, G., Heckenberg, N., and H. Rubinsztein-Dunlop. (2007). Physics
of optical tweezers. Laser Manipulation Cells Tissues, 82, 207-236.
Oldham, K. B. (2004). The RC time ‘‘constant’’ at a disk electrode, Electrochem.
Commun. 6, 210-214.
Ophir, M.J., Liu, B.C., and S. C. Bunnell. (2013). The N terminus of SKAP55 enables T
cell adhesion to TCR and integrin ligands via distinct mechanisms, J Cell Biol, 203,
1021-1041.
Paeng, D. G., M. J. Choi, S.W. Yoon, and K. K. Shung. (2004). Echogenicity from
disturbed blood flow by an eccentric stenosis under pulsatile flow, Journal of the
Acoustical Society of America, 116, 2560.
Palomba, M. L., Piersanti, K., Ziegler, C. G., Decker, H., Cotari, J.W., Bantilan, K., Rijo,
I., Gardner, J. R., Heaney, M., Bemis, D., Balderas, R., Malek, S. N., Seymour, E.,
Zelenetz, A. D., van den Brink, M. R., and G. Altan-Bonnet. (2014). Multidimensional
single-cell analysis of BCR signaling reveals proximal activation defect as a hallmark of
chronic lymphocytic leukemia B cells, PloS One, 9, e79987.
Pfeifer, A., and I. M. Verma. (2001). Gene therapy: promises and problems, Annu. Rev.
Genomics Hum. Genet, 2, 177-211.
Piret, J., Goyette, N., and G. Boivin. Novel method based on real-time cell analysis for
drug susceptibility testing of herpes simplex virus and human cytomegalovirus, J. Clin
Microbiol, 54, 2120-2127.
102
Piwowarska, E., and A. Sidlarewicz. (2006). Analysis of spiral inductor model in CMOS
circuit, The International Conference of Mixed Design of Integrated Circuits and System,
454 -459.
Poss, K. D., L. G. Wilson, and M. T. Keating. (2002). Heart Regeneration in Zebrafish,
Science, 298, 2188-2190.
Pozar, D.M. (1998). Microwave Engineering 2nd edition, New York: Wiley.
Ramser, K. and D. Hanstorp. (2010). Optical manipulation for single-cell studies, J.
Biophotonics, 3, 187-206.
Reinstein, D. Z., R. H. Silverman, and D. J. Coleman. (1993). High-frequency ultrasound
measurement of the thickness of the corneal epithelium, Refract Corneal Surg, 9, 385-
387.
Rhea, R. (2006). The Yin-Yang of Matching: Part 2 – Practical Matching Techniques,
High Frequency Electronics.
Ritch, R., and J. M. Liebmann, 1998, Role of ultrasound biomicroscopy in the
differentiation of block glaucomas: Curr Opin Ophthalmol, v. 9, p. 39-45.
Ritter, T. A., T. R. Shrout, R. Tutwiler, and K. K. Shung, 2002, A 30-MHz piezo-
composite ultrasound array for medical imaging applications: IEEE Trans Ultrason
Ferroelectr Freq Control, v. 49, p. 217-30.
Rogers, J., and C. Plett. (2003). Radio Frequency Integrated Circuit Design, Artech
House.
Rué, P. and A. A. Martinez. (2015). Cell dynamics and gene expression control in tissue
homeostasis and development, Mol Syst Biol, 11, 792.
Said, T. M., Agarwal, A., Zborowski, M., Grunewald, S., Glander, H. J., and U. Paasch.
(2008). Utility of magnetic cell separation as a molecular sperm preparation technique. J
Androl, 29, 134-142.
Sanchez-Freire, V., Ebert, A. D., Kalisky, T., Quake, S. R. and J. C. Wu. (2012).
Microfluidic single-cell real-time PCR for comparative analysis of gene expression
patterns, Nat. Protoc. 7, 829-838.
Schoell, W. M., Klintschar, M., Mirhashemi, R. and B. Pertl. (1999). Separation of sperm
and vaginal cells with flow cytometry for DNA typing after sexual assault, Obstet
Gynecol, 94, 623-627.
Sedmera, D., M. Reckova, A. deAlmeida, M. Sedmerova, M. Biermann, J. Volejnik, A.
Sarre, E. Raddatz, R. A. McCarthy, R. G. Gourdie, and R. P. Thompson. (2003).
103
Functional and morphological evidence for a ventricular conduction system in zebrafish
and Xenopus hearts, Am J Physiol Heart Circ Physiol, 284, 1152-60.
Sharei, A., Zoldan, J., Adamo, A., Sim, W. Y, Cho, N., Jackson, E., Mao, S., Schneider,
S., Han, M., Lytton-Jean, A., Basto, P. A., Jhunjhunwala, S., Lee, J., Heller, D.A, Kang,
J.W., Hartoularos, G. C., Kim, K., Anderson, D. G., Langer, R., and K. F. Jensen. (2013).
A vector-free microfluidic platform for intracellular delivery. Proc. Natl. Acad. Sci. U. S.
A., 110, 2082-2087.
Shi, C., Guo, D., Xiao, K.,Wang, X., Wang, L., and J. Luo. (2015). A drug-specific
nanocarrier design for efficient anticancer therapy, Nat Commun, 6, 7449.
Shung, K. K. (2005). Diagnostic Ultrasound: Imaging and Blood Flow Measurements,
CRC Press.
Silverman, R. H., F. L. Lizzi, B. G. Ursea, L. Cozzarelli, J. A. Ketterling, C. X. Deng, R.
Folberg, and D. J. Colemen. (2001). Safety levels for exposure of cornea and lens to very
high-frequency ultrasound, Journal of Ultrasound in Medicine, 20, 979-986.
Silverman, R. H. (2009). High-resolution ultrasound imaging of the eye - a review, Clin
Experiment Ophthalmol, 37, 54-67.
Snook, K. A., Zhao, J. Z., Alves, C. H., Cannata, J. M., Chen, W. H., Meyer, R. J. Jr.,
Ritter, T. A., and K. K. Shung. (2002). Design, fabrication, and evaluation of high
frequency, single-element transducers incorporating different materials, IEEE
Transactions on Ultrasonics, Ferroelectrics and Frequency Control, 49, 169-176.
Stephen, N. G. (2006). On the Maximum Power Transfer Theorem Within
Electromechanical Systems, Journal of Mechanical Engineering Science, 220, 1261-
1267.
Sung, K., and T. Won. (2011). Design of a 100 V high-side n-channel LDMOS transistor
for breakdown voltage enhancement, Journal of the Korean Physical Society, 59, 412-
415.
Tunis, A. S., Czarnota, G. J., Giles, A., Sherar, M. D., Hunt, J. W., and M. C. Kolios.
(2005). Monitoring structural changes in cells with highfrequency ultrasound signal
statistics, Ultrasound Med. Biol, 31, 1041-1049.
VanDelinder, V. and A. Groisman. (2006). Separation of plasma from whole human
blood in a continuous cross-flow in a molded microfluidic device, Analytical Chemistry,
78, 3765-3771.
van Unen, J., Rashidfarrokhi, A., Hoogendoorn, E., Postma, M., Gadella, T. W. J., and J.
Goedhart. (2016). Quantitative single cell analysis of signaling pathways activated
immediately downstream of histamine receptor subtypes, Mol. Pharmacol, 90, 162-176.
104
Van Wamel, A., Kooiman, K., Harteveld, M., Emmer, M., Ten Cate, F. J., Versluis, M.,
and N. De Jong. (2006). Vibrating microbubbles poking individual cells: drug transfer
into cells via sonoporation, J Control Release, 112, 149-155.
Xu, Y., Xie, J., Chen, R., Cao, Y., Ping, Y., Xu, Q., Hu, W., Wu, D., Gu, L., Zhou, H.,
Chen, X., Zhao, Z., Zhong, J., and R. Li. (2016). Fluorescence- and magnetic-activated
cell sorting strategies to separate spermatozoa involving plural contributors from
biological mixtures for human identification, Scientific Reports, 6, 36515.
Wang, D. and S. Bodovitz. (2010). Single cell analysis the new frontier in ‘omics',
Trends Biotechnol., 28, 281-290.
Watanabe, W., Arakawa, N., Matsunaga, S., Higashi, S., Fukui, K., Isobe, K., and K. Itoh.
(2004). Femtosecond laser disruption of subcellular organelles in a living cell, Opt.
Express, 12, 4203-4213.
World Health Organization. (2014). Health Statistics and Information Systems: WHO
Mortality Database, World Cancer Report, 2014.
Yamada, M., Nakashima, M. and M. Seki. (2004). Pinched flow fractionation: continuous
size separation of particles utilizing a laminar flow profile in a pinched microchannel,
Anal Chem, 76, 5465-5471.
Yin, D., and D. Marshall. (2012). Microfluidics for single cell analysis, Curr. Opin.
Biotechnol., 23, 110-119.
Yoon, C., Kang, B. J., Lee, C., Kim, H. H. and, K. K. Shung. (2014). Multi-particle
trapping and manipulation by a high-frequency array transducer, Applied Physics Letters,
105, 214103.
Yoon, S., M. G. Kim, J. A. Williams, C. Yoon, B. J. Kang, N. Cabrera-Munoz, K. K.
Shung, and H. H. Kim. (2015). Dual element needle transducer for intravascular
ultrasound imaging, J Med Imaging, 2, 027001.
Yoon, S., Kim, M. G., Wang, Y., and K. K. Shung., (2015). Programmable delivery of
macromolecules using high frequency ultrasound, Ultrasonics Symposium (IUS) IEEE
International, 1-3.
Yoon, S., J. A. Williams, B. J. Kang, C. Yoon, N. Cabrera-Munoz, J. S. Jeong, S. G. Lee,
K. K. Shung, and H. H. Kim. (2015). Angled- focused 45 MHz PMN-PT single element
transducer for intravascular ultrasound imaging, Sens Actuators A Phys, 228, 16-22.
Yoon, S., Kim, M. G., Chiu, C. T., Hwang, J. Y., Kim, H. H., Wang, Y., and K. K. Shung.
(2016). Direct and sustained intracellular delivery of exogenous molecules using
acoustic-transfection with high frequency ultrasound, Sci Rep, 6, 20477.
105
Youla, D.C. (1964). A New Theory of Broad-band Matching, IEEE Transactions on
Circuit Theory, 11, 30-50.
Zeira, E., Manevitch, A., Khatchatouriants, A., Pappo, O., Hyam, E., Darash-Yahana, M.,
Tavor, E., Honigman, A., Lewis, A., and E. Galun. (2003). Femtosecond infrared laser-an
efficient and safe in vivo gene delivery system for prolonged expression, Mol Ther, 8,
342-350.
Zeira, E., Manevitch, A. Manevitch, Z., Kedar, E., Gropp, M., Daudi, N., Barsuk, R.,
Harati, M., Yotvat, H., Troilo, P.J., 2nd Griffiths, T.G., Pacchione, S.J., Roden, D.F., Niu,
Z., Nussbaum, O., Zamir, G., Papo, O., Hemo, I., Lewis, A., and E. Galun. (2007).
Femtosecond laser: a new intradermal DNA delivery method for efficient, long-term gene
expression and genetic immunization, FASEB J, 21, 3522-3533.
Abstract (if available)
Abstract
Single-cell analysis is an essential step to elucidate and understand cellular behaviors and functional responses in a large cell population at the individual cell level. The application of a high frequency ultrasound is a promising approach for single-cell analysis, providing unique characteristics such as being easy-to-use and cost-effective as well as having minimal effects on normal cell physiology and a label-free investigation under high specificity and spatial resolution. This dissertation presents single-cell analysis with high frequency ultrasound and covers single-cell analysis and custom-built high frequency ultrasound electronics for single-cell analysis. For single-cell analysis, a new transfection technique using a highly focused high frequency ultrasound was proposed and demonstrated by live-cell fluorescence imaging of time-based intensity changes for a fluorescent propidium iodide (PI) molecule and the cytoplasmic delivery of 3 kDa dextran. This transfection technique was developed further as a versatile and adaptable transfection tool for the intensive investigation of treatment conditions between different human cancer cell lines. This approach was verified by the intracellular delivery of 70 kDa dextran and the simultaneous intracellular delivery of two molecules including 70 kDa dextran and PI into four human cancer cell lines using optimal treatment conditions with high delivery efficiency with minimal cell membrane disruption. Also, a label-free single-cell separation technique using the highly focused high frequency ultrasound was presented and demonstrated by individually identifying the size of the trapped single object based on the ultrasound backscattering coefficient. The proposed strategy provides beneficial characteristics, including it being an easy-to-use, cost-effective, label-free investigation of physical properties with minimal effects on normal cell physiology with a much more accurate analysis at the single-cell level. Moreover, for custom-built high frequency ultrasound electronics for single-cell analysis, an impedance matching network was developed to maximize energy transmission from the excitation source to high frequency ultrasonic transducers for cell manipulation as well as to achieve low-input parameters for the safe operation of ultrasonic transducers. Also, a high frequency ultrasound all-in-one front-end system with a novel driving condition to overcome the limitations of commercially available systems was designed and developed. This system could be used for a label-free acoustic sensing method and for multifunctional high frequency ultrasound imaging.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Microfluidic cell sorting with a high frequency ultrasound beam
PDF
Quantification of cellular properties using ultra-high frequency single-beam acoustic tweezer
PDF
Development of high frequency focused transducers for single beam acoustic tweezers
PDF
Multi-modality intravascular imaging by combined use of ultrasonic and opticial techniques
PDF
High-frequency ultrasonic transducers for photoacoustic applications
PDF
High frequency ultrasonic imaging for the development of breast biopsy needle with miniature ultrasonic transducer array
PDF
High frequency ultrasonic phased array system and its applications
PDF
High frequency ultrasound array for ultrasound-guided breast biopsy
PDF
Development of front-end circuits for high frequency ultrasound system
PDF
High-frequency ultrasound array-based imaging system for biomedical applications
PDF
High-frequency ultrasound imaging system with Doppler features for biomedical applications using 30~35 mHz linear arrays and an analog beamformer
PDF
A high frequency array- based photoacoustic microscopy imaging system
PDF
Development of high-frequency (~100mhz) PZT thick-film ultrasound transducers and arrays
PDF
Ultrasound-mediated mechanotransduction and its applications
PDF
Intravascular imaging on high-frequency ultrasound combined with optical modalities
PDF
High frequency ultrasound elastography and its biomedical applications
PDF
Array transducers for high frequency ultrasound imaging
PDF
Membrane-bound regulation of hematopoietic stem cells
PDF
Development of back-end processing system for high frequency ultrasound b-mode imaging
PDF
Improved contrast in ultrasound imaging using dual apodization with cross-correlation
Asset Metadata
Creator
Kim, Min Gon
(author)
Core Title
Single-cell analysis with high frequency ultrasound
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Biomedical Engineering
Publication Date
06/05/2017
Defense Date
03/23/2017
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
high frequency ultrasound,OAI-PMH Harvest,single-cell analysis
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Shung, K. Kirk (
committee chair
), Chow, Robert H. (
committee member
), McCain, Megan L. (
committee member
), Shen, Keyue (
committee member
), Yen, Jesse T. (
committee member
)
Creator Email
mingonki@gmail.com,mingonki@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-379617
Unique identifier
UC11258033
Identifier
etd-KimMinGon-5379.pdf (filename),usctheses-c40-379617 (legacy record id)
Legacy Identifier
etd-KimMinGon-5379.pdf
Dmrecord
379617
Document Type
Dissertation
Rights
Kim, Min Gon
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
high frequency ultrasound
single-cell analysis