University of Southern California Dissertations and Theses

Intravascular imaging on high-frequency ultrasound combined with optical modalities

(USC Thesis Other)

Intravascular imaging on high-frequency ultrasound combined with optical modalities

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

INTRAVASCULAR IMAGING ON HIGH-FREQUENCY ULTRASOUND
COMBINED WITH OPTICAL MODALITIES

by

Xiang Li

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOMEDICAL ENGINEERING)

August 2012

Copyright 2012 Xiang Li
ii

Dedication

To my beloved

Parents
Huafeng Hu & Shizhong Li

Wife
Hu Ye

iii

Acknowledgements
I am sincerely grateful to my advisors Professor K. Kirk Shung and Professor
Qifa Zhou, for their guidance, mentoring and support during my Ph.D. study at
Resource Center for Medical Ultrasonic Transducer Technology. They are truly great
mentors who pass to me not only their knowledge but also their life philosophy. I
would also like to express my sincere gratitude to my dissertation committees: K. Kirk
Shung, Qifa Zhou, Jesse Yen and Steven Nutt for their suggestion during my
preparation of this dissertation.
I am heartily thankful to Professor Zhongping Chen from UC Irvine, who is my
project supervisor in collaboration. Without Prof. Chen’s support, my research work
will never be possible. I would also like to thank my colleagues from UC Irvine:
Jiechen Yin, Joe Jing, Jiawen Li and Wei Wei. We have been working together for
three years and shared a lot pains and happiness.
I am greatly thankful to all my colleagues and friends at the transducer resource
center, especially Ruimin Chen and Dawei Wu, who taught me the knowledge of
transducer fabrication and ultrasound imaging system when I was firstly involved in the
center. Special thanks to Professor Tzung Hsiai and Fei Yu, who helped me a lot in
understanding atherosclerosis and reading intravascular ultrasound images.
Last but not the least, I would like to thank my parents who raised me up and
give me unconditional love; and my dear Hu Ye, who always stand on my side through
all the good times and bad.

iv

Table of Contents
Dedication ........................................................................................................................ ii
Acknowledgements ......................................................................................................... iii
List of Tables ................................................................................................................. vii
List of Figures ............................................................................................................... viii
Abstract ......................................................................................................................... xiv
Chapter 1 Introduction ......................................................................................................1
1.1 Atherosclerosis ................................................................................................1
1.2 Biomedical Imaging for Diagnosing Atherosclerosis .....................................2
1.2.1 Intravascular Ultrasound (IVUS) Imaging .......................................4
1.2.2 Intravascular Optical Coherence Tomography (OCT)
Imaging ...........................................................................................6
1.2.3 Intravascular Photoacoustic (IVPA) Imaging ..................................7
1.3 Plaque Characterization by Imaging Interpretation ........................................8
1.4 Scopes of the Dissertation .............................................................................11
Chapter 2 High Frequency Ultrasonic Transducer for IVUS Imaging ...........................13
2.1 B (Brightness)-Mode Ultrasound Imaging ...................................................13
2.2 Resolution, Attenuation and Backscattering in High Frequency
Ultrasound ...................................................................................................14
2.2.1 Resolution ......................................................................................14
2.2.2 Attenuation .....................................................................................15
2.2.3 Backscattering ................................................................................17
2.3 Single Element Ultrasonic Transducer .........................................................18
2.3.1 Piezoelectric Effects .......................................................................18
2.3.2 Piezoelectric Film Technology ......................................................19
2.3.3 Single Element Ultrasonic Transducer Design ..............................21
2.4 Prototype IVUS Transducer ..........................................................................22
2.4.1 IVUS Transducer Design and Fabrication .....................................22
2.4.2 Transducer Performance Characterization .....................................25
2.4.3 IVUS System Setup and Imaging Experiments .............................28
Chapter 3 Intravascular Ultrasound Combined with Optical Coherence
Tomography Imaging ..........................................................................................31
3.1 Advantages of Integrated IVUS-OCT Imaging Modality ............................31
3.2 Integrated IVUS-OCT System ......................................................................32
3.2.1 System Setup ..................................................................................32
3.2.2 Signal Acquisition and Processing .................................................35
v

3.3 First Generation Probe Design: Side-by-Side Arrangement .........................37
3.3.1 Probe Design ..................................................................................37
3.3.2 Imaging Results .............................................................................38
3.4 Second Generation Probe Design: Coaxial Arrangement .............................40
3.4.1 Probe Design ..................................................................................40
3.4.2 Ring Transducer Characterization and Mirror Effects ...................41
3.4.3 Imaging Results .............................................................................43
3.4.4 Downsized Version: Miniature Ring Transducer ..........................45
3.5 Third Generation Probe Design: Sequential Arrangement ...........................46
3.5.1 Probe Design ..................................................................................46
3.5.2 In-vitro Imaging on Rabbit Aorta Specimens ................................49
3.5.3 In-vitro Imaging on Human Coronary Artery
Specimens .....................................................................................55
3.5.4 In-vivo Imaging on Rabbit .............................................................56
3.5.5 In-vivo Imaging on Swine ..............................................................58
3.6 Discussion .....................................................................................................61
Chapter 4 Intravascular Ultrasound at 80 and 95 MHz ..................................................63
4.1 Introduction to Very High Frequency IVUS ................................................63
4.2 IVUS at 80 MHz ...........................................................................................64
4.2.1 PMN-PT Free-Standing Film Fabrication .....................................64
4.2.2 PMN-PT Free-Standing Film Characterization .............................67
4.2.3 Transducer Design and Fabrication ...............................................70
4.2.4 Free-Standing Film IVUS Transducer
Characterization ............................................................................71
4.2.5 80 MHz IVUS Imaging Results .....................................................73
4.3 IVUS at 95 MHz ...........................................................................................76
4.3.1 95 MHz IVUS Transducer Using 22 µ m PMN-PT
Free-Standing Film .......................................................................76
4.3.2 95 MHz IVUS Imaging Results .....................................................78
4.4 Discussion .....................................................................................................83
Chapter 5 Intravascular Ultrasound Combined with Photoacoustic Imaging .................86
5.1 Introduction to Intravascular Photoacoustic Imaging ...................................86
5.2 IVPA System Setup ......................................................................................87
5.3 Probe Design .................................................................................................88
5.3.1 Side-by-Side Arrangement.............................................................88
5.3.2 Coaxial Arrangement .....................................................................89
5.4 IVPA Imaging Results ..................................................................................91
5.4.1 Imaging on Tissue Mimicking Phantom by the
Coaxial Probe ...............................................................................91
5.4.2 In-vitro Imaging on Rabbit Aorta by the Coaxial
Probe .............................................................................................94
vi

5.4.3 Imaging on 6-µ m Wire Targets by the Side-by-Side
Probes at 35 and 80 MHz .............................................................96
5.4.4 In-vitro Imaging on Rabbit Aorta by the Side-by-Side
Probes at 35 and 80 MHz .............................................................98
5.5 Discussion ...................................................................................................102
Bibliography .................................................................................................................104

vii

List of Tables
Table 2-1: Design parameters of 40 MHz IVUS transducer ............................................. 23
Table 2-2: 40 MHz IVUS transducer testing results ......................................................... 27
Table 4-1: 80 MHz IVUS transducer testing results ......................................................... 73

viii

List of Figures
Figure 1-1: The formation of Atherosclerosis. The normal artery is
composed of three layers: Intima, Media and Adventitia (a);
As a fatty streak develops inside the intima (b), thin fibrous
cap and a fatty core will be formed (c); When a rupture of an
unstable plaque happens (d), it will cause blood clots and
block the artery entirely (e) (adapted from
www.medmovie.com) ..................................................................................2
Figure 1-2: An intravascular imaging catheter scanning inside a vessel
(adapted from www.trdesign.com/lightlab/oct.php) ....................................4
Figure 1-3: Two standard schematics of IVUS catheter: Single element
transducer (A); Array transducer (B) ...........................................................5
Figure 1-4: Schematics of an IVUS catheter (a) (adapted from
www.bostonscientific.com) and an OCT catheter (b) (adapted
from www.lightlab.com) ..............................................................................7
Figure 1-5: IVUS images and corresponding histology for fibrous (A, B),
calcified (C, D), and lipid-rich (E, F) plaque types. Fib,
fibrous plaque; Ca, calcified plaque; L, lipid-rich plaque
(adapted from (Potkin, Bartorelli et al. 1990)) ..........................................10
Figure 1-6: OCT images and corresponding histology for fibrous (A, B),
calcified (C, D), and lipid-rich (E, F) plaque types. Fib,
fibrous plaque; Ca, calcified plaque; L, lipid-rich plaque
(adapted from (Tearney, Jang et al. 2006)) ................................................10
Figure 2-1: The procedure of processing the IVUS RF data ..........................................14
Figure 2-2: Attenuation of variety tissues in the frequency range from 10
to 100 MHz (Foster, Pavlin et al. 2000).....................................................16
Figure 2-3: Plots of backscatter coefficients versus shear rates at 65 MHz
(Adapted from Foster, Obara et al. 1994) ..................................................18
Figure 2-4: Prototype IVUS transducers: needle type (a); and flexible type
(b) ...............................................................................................................24
Figure 2-5: Pulse-echo measurement of one 40 MHz IVUS transducer ........................26
Figure 2-6: Wire phantom image from one 40 MHz IVUS transducer ..........................27
ix

Figure 2-7: IVUS imaging system ..................................................................................29
Figure 2-8: Custom built rotational coupling joint .........................................................29
Figure 2-9: IVUS image of human coronary artery at 40 MHz. I, intima;
M, media; A, adventitia; Ca, calcified plaque; L, lipid-rich
plaque .........................................................................................................30
Figure 3-1: Schematic of IVUS-OCT imaging system. The blue blocks
represent signal flow; the orange blocks represent the
mechanical joints; the green blocks represent synchronizing
triggers .......................................................................................................33
Figure 3-2: A photograph of the integrated IVUS-OCT system .....................................33
Figure 3-3: A rotary joint device connects the rotational and pull-back
motor; and couples electrical and optical signals from the
rotational part to the stationary part ...........................................................35
Figure 3-4: Schematic of IVUS-OCT probe ...................................................................38
Figure 3-5: OCT (a) and IVUS (b) images of rabbit aorta acquired by the
first generation probe .................................................................................39
Figure 3-6: Schematic of coaxially arranged IVUS-OCT probe ....................................40
Figure 3-7: Ultrasound wire phantom images without a mirror(a) and with
a mirror(d), displayed with a dynamic range of 50 dB; axial
and lateral envelopes of the RF echo signals from the wire
located at transducer focal point without a mirror(b)(c) and
with a mirror(e)(f) ......................................................................................43
Figure 3-8: OCT (a), IVUS (b) and combined IVUS-OCT (c) images of a
rabbit aorta. I, Cork holder .........................................................................44
Figure 3-9: Downsized version of the coaxial arranged IVUS-OCT probe.
A miniature ring transducer with an outer diameter of 1 mm
and an inner diameter of 0.45 mm (a); the transducer is
assembled with an OCT probe in the center and a 45˚ prism
at the tip (b), the pink spot is the reflection from an visible
light source; and the whole probe (c) .........................................................45
Figure 3-10: Pulse-echo measurement of the miniature ring transducer,
which has a center frequency of 44.4 MHz with 51 %
bandwidth ...................................................................................................45
x

Figure 3-11: In vitro rabbit aorta images from the downsized version of
coaxially arranged OCT and IVUS probe ..................................................46
Figure 3-12: Schematic (a) and photo (b) of miniature IVUS-OCT probe .....................48
Figure 3-13: A miniature IVUS-OCT probe ...................................................................48
Figure 3-14: A finished integrated IVUS-OCT catheter and the imaging
probe is inside a 3.6 Fr sheath ....................................................................49
Figure 3-15: OCT (a), IVUS (b) and corresponding H&E histology (c)
images of an atherosclerotic rabbit aorta with eccentric
plaque. I, intima; M, media; Fib, fibrous plaque. The yellow
and black arrows point to a plaque cap in OCT and histology
images, respectively ...................................................................................51
Figure 3-16: OCT (a), IVUS (b) and corresponding H&E histology (c)
images of an atherosclerotic rabbit aorta with concentric
plaque. The plaque on left panel is dramatically thicker than
that on the right panel. I, intima; M, media; Fib, fibrous
plaque. The yellow and black arrows point to a plaque cap in
OCT and histology images, respectively ...................................................52
Figure 3-17: OCT (a) and IVUS (b) images of a thick thrombus. The
thrombus (T1, T2) rapidly attenuates OCT signals and blocks
the view of vessel (V) in OCT image. Ultrasound could
penetrate the thrombus to image the vessel (V), but the
boarder of the vessel is unclear ..................................................................54
Figure 3-18: OCT (a) and IVUS (b) images of a thin thrombus. The
thrombus (T) signal appears weaker than that from vessel (V)
in OCT image, thus the boarder of vessel is clearly identified.
Ultrasound could penetrate the thrombus to image the vessel
(V), but the boarder of the vessel is unclear ..............................................54
Figure 3-19: OCT (a), ultrasound (b), and combined IVUS-OCT (c)
images of a human coronary artery specimen ............................................56
Figure 3-20: A photograph of in vivo rabbit experiment ................................................57
Figure 3-21: OCT (a) and IVUS (b) images of a rabbit abdominal aorta
without flushing. OCT (c), IVUS (d) and combined (e)
IVUS-OCT images with flushing. L, lipid-rich content; FC,
fibrous cap ..................................................................................................58
xi

Figure 3-22: in vivo experiment on swine .......................................................................60
Figure 3-23: OCT (a) and IVUS (b) images of a swine coronary artery
after flushing. G, guide wire; V, vessel; T, tissue surrounding
outside the vessel .......................................................................................61
Figure 4-1: Illustration of resolution versus penetration depth of
conventional IVUS (40 MHz), OCT, and very high frequency
IVUS (80~100 MHz) .................................................................................64
Figure 4-2: An SEM image of the top surface of a PMN-PT green tape (a);
An optical image of a PMN-PT green tape (tan color) on a
black bench top (b).....................................................................................66
Figure 4-3: X-ray diffraction (XRD) pattern of the PMN-PT free-standing
film .............................................................................................................67
Figure 4-4: SEM micrographs of the 30 µ m PMN-PT free-standing film: a
cross-section view (a) and an enlarged view (b). .......................................68
Figure 4-5: Frequency dependence of dielectric constant and loss of PMN-
PT free-standing film. ................................................................................69
Figure 4-6: Polarization-electric field hysteresis loop of the annealed
PMN-PT free-standing film .......................................................................69
Figure 4-7: Frequency dependence of electrical impedance of the 30 µ m
PMN-PT free-standing film .......................................................................70
Figure 4-8: SEM photograph of an acoustic stack that is composed of
acoustic matching layer, PMN-PT film layer, and acoustic
backing layer ..............................................................................................71
Figure 4-9: Pulse-echo measurement of one representative 80 MHz PMN-
PT free-standing film transducer ...............................................................72
Figure 4-10: Ultrasound wire phantom image (a), displayed with a
dynamic range of 45 dB; axial (b) and lateral (c) envelopes of
echo signals from the wire located at 1.2 mm away from the
transducer surface ......................................................................................73
Figure 4-11: Images of human cadaver coronary artery specimen from 80
MHz PMN-PT freestanding-film transducer (a); and 40 MHz
PMN-PT single-crystal transducer (b). Fib, fibrous plaque;
Ca, calcified plaque; L, lipid-rich plaque; .................................................74
xii

Figure 4-12: Images of human cadaver coronary artery specimen from 80
MHz PMN-PT freestanding-film transducer (a); and 40 MHz
PMN-PT single-crystal transducer (b). Three layer structures
are displayed with higher contrast in the 80 MHz image than
the 40 MHz one. I, intima; M, media; A, Adventitia .................................75
Figure 4-13: SEM micrographs of a 22 µ m PMN-PT free-standing film ......................77
Figure 4-14: SEM photograph of an acoustic stack that is composed of 22
µ m PMN-PT free-standing film and conductive backing layer .................77
Figure 4-15: Frequency dependence of electrical impedance of the 22 µ m
PMN-PT free-standing film .......................................................................77
Figure 4-16: Pulse-echo measurement of one representative 95 MHz
PMN-PT free-standing film transducer .....................................................78
Figure 4-17: Images of a human cadaver coronary artery specimen from a
95 MHz PMN-PT free-standing-film transducer (a) and a 40
MHz PMN-PT single-crystal transducer (b). And the
corresponding histology by Von Kossa staining. ca, calcified
plaque; A fibrous plaque cap was pointed out by an arrow.
The calcium appears black in Von Kossa staining. The blank
area at 9 o’clock in (c) is due to the dropping of calcium in
staining process. ca, calcified plaque; cap, fibrous cap .............................79
Figure 4-18: Images of human cadaver coronary artery specimen from a
95 MHz PMN-PT free-standing-film transducer (a) and a 80
MHz PMN-PT free-standing-film transducer (b) ......................................80
Figure 4-19: Images of human cadaver coronary artery specimen from a
95 MHz PMN-PT free-standing-film transducer (a) and an
OCT probe (b). ca, calcified plaque; L, lipid-rich plaque; cap,
fibrous cap ..................................................................................................81
Figure 4-20: Images of a human cadaver coronary artery specimen from a
40 MHz PMN-PT single-crystal transducer (a); a 95 MHz
PMN-PT free-standing-film transducer (b); an OCT probe (c);
and H&E histology (d). Fib, fibrous plaque; I, intima; M,
media; A, Adventitia ..................................................................................82
Figure 5-1: Schematic of the hybrid IVUS and IVPA imaging system. .........................88
xiii

Figure 5-2: Schematic of the integrated IVUS/IVPA probe: top view (a)
and front view (b). Light beam is in green and acoustic beam
in gray ........................................................................................................89
Figure 5-3: Schematic of the combined IVPA probe (a); Photo of the ring-
shaped ultrasonic transducer (b); Photo of the combined
IVPA probe after packaging (c) .................................................................91
Figure 5-4: Schematic of the phantom in which the 0.5 mm diameter
graphite rod and 0.5 mm diameter air lumen are embedded.
The graphite rod and air lumen work as two contrast
inclusions which have different absorption coefficient within
a tissue-mimicking phantom. .....................................................................92
Figure 5-5: Cross-sectional IVUS (a), IVPA (b), and combined image of
the phantom (c). The field of view is 6.5 mm in radius in all
images. IVUS and IVPA images are displayed in dynamic
ranges of 50 dB and 40 dB. ........................................................................94
Figure 5-6: Cross-sectional IVUS (a), IVPA (b), Hematoxylin-Eosin
(H&E) stained histological image (c), and combined image of
a normal rabbit aorta (d). The IVPA image was obtained
using 532 nm optical excitation wavelength and 39 MHz
ring-shaped transducer. IVUS and IVPA images are
displayed in dynamic ranges of 50 dB and 40 dB. ....................................95
Figure 5-7: 35 MHz photoacoustic image of 6-µ m tungsten wires (a); axial
and lateral envelopes (b) of photoacoustic signal from the
wire located at 2.5 mm away from the transducer surface ........................97
Figure 5-8: 80 MHz photoacoustic image of 6-µ m tungsten wires (a); axial
and lateral envelopes (b) of photoacoustic signal from the
wire located at 2.5 mm away from the transducer surface ........................97
Figure 5-9: RF signals and spectrums of photoacoustic pulses generated by
the wire targets at 35 MHz (a) and 80 MHz (b) .........................................98
Figure 5-10: Cross-sectional IVUS (a), IVPA (b) and fused (c) images of a
healthy rabbit aorta at 35 MHz; and Hematoxylin-Eosin
(H&E) stained histology image (d) ..........................................................100
Figure 5-11: Cross-sectional IVUS (a), IVPA (b) and fused (c) images of a
healthy rabbit aorta at 80 MHz; and Hematoxylin-Eosin
(H&E) stained histology image (d) ..........................................................101
xiv

Abstract
Atherosclerosis is a complex syndrome characterized by plaques build up on the
inner lining of arteries, which is the leading cause of morbidity in developed countries.
Conventional gold standard for diagnosing the vulnerability of plaques is intravascular
ultrasound (IVUS) imaging working in the 20~40 MHz range. At this range, the
resolution (~100 µ m) is insufficient for detecting a number of critical microstructures
that are in the scope of less than 65 µm. Driving the working frequency above 80 MHz
could be a feasible solution for improving IVUS resolution. Meanwhile, intravascular
Optical Coherence Tomography (OCT) is a new imaging modality recently approved
for clinical use. OCT features high resolution (10~30 µm) but limited penetration depth
(~1 mm). The other emerging optical modality: Intravascular Photoacoustic (IVPA) is
based on the tissue optical absorption properties and has been proved to offer better
contrast of tissue composition.
In this dissertation, three intravascular imaging modalities have been
investigated: integrated IVUS-OCT imaging, very high frequency IVUS, and IVPA
imaging, which aim to overcome the limitations in resolution, penetration and imaging
contrast of current technologies. Imaging systems and different types of ultrasonic or
optical-ultrasonic hybrid probes have been designed and fabricated for using in each
modality. In vitro and in vivo studies have been carried out to validate these new
imaging concepts. The imaging results successfully support the enhancements brought
by the new modalities. The complementary synergy of ultrasonic and optical modalities
xv

provides extra diagnosing information. This research suggests a bright future for multi-
modality imaging that combines the conventional IVUS and new optical technologies.

1

Chapter 1 Introduction
1.1 Atherosclerosis
The targeted clinical disease in this research is atherosclerosis, which is one of
the major causes of morbidity and mortality in developed countries. Atherosclerosis is
characterized by the thickening of the arterial vessel wall due to the building up of
atheromatous plaque on the inner lining of arteries. The formation of atherosclerosis is
illustrated in Figure 1-1. The normal artery is composed of three layers: intima, media
and adventitia. As a fatty streak develops inside the intima, thin fibrous plaque with a
fatty core will be formed gradually. If the fibrous plaque is composed of thin cap, it can
be vulnerable to rupture, which may result in narrowing or even occlusion of entire
arterial lumen (Ross 1999). As a result, many related diseases, such as coronary heart
disease, carotid artery disease, and peripheral arterial disease will be developed.
Vulnerable plaques are the major reasons of myocardial infarction and stroke (Lusis
2000). Although the understanding of vulnerable plaques is still at an early stage,
previous research has demonstrated that the thin-cap fibroatheroma (TCFA) accounts
for 80% of sudden cardiac death (Kolodgie, Burke et al. 2001). TCFA has been
histologically identified with the following pathological features: thin fibrous cap
(<65μm), large lipid pool, and macrophages near or within the fibrous cap. TCFA is the
most prevalent precursor of vulnerable plaque and responsible for acute coronary
syndromes (ACS) (Rathore, Terashima et al. 2011). Therefore it is important to detect
2

the vulnerable plaques or TCFA in early time and manage them before they become
painful or life threatening.

Figure 1-1: The formation of Atherosclerosis. The normal artery is composed of three
layers: Intima, Media and Adventitia (a); As a fatty streak develops inside the intima
(b), thin fibrous cap and a fatty core will be formed (c); When a rupture of an unstable
plaque happens (d), it will cause blood clots and block the artery entirely (e) (adapted
from www.medmovie.com)
1.2 Biomedical Imaging for Diagnosing Atherosclerosis
Biomedical imaging techniques aiming at imaging and assessing vulnerable
plaques have been applied in clinical interventions or reported in literatures, such as
angiography, magnetic resonance imaging (MRI), intravascular ultrasound (IVUS),
optical coherence tomography (OCT) and intravascular photoacoustic (IVPA) (Foster,
Pavlin et al. 2000; Pasterkamp, Falk et al. 2000; Tearney, Jang et al. 2006; Wang, Su et
3

al. 2010). Angiography provides a projection view of the vessels with the flushing of
radio-opaque contrast agent. Angiography is sensitive to the vascular stricture or blood
occlusion, however, it is hard to evaluate the vulnerability of plaques since no detailed
characterization of the plaque structure or composition can be extracted from the
projection imaging. In clinic, angiography is usually used to give a bird view of the
vascular network and locate the potential lesions. MRI is performed to study the
progression and regression of plaque over time, however, its insufficient resolution
cannot render accurate measurements of the micro-structures of vessel wall or plaque
cap (Pasterkamp, Falk et al. 2000).
IVUS and OCT are currently the two most commonly used intravascular
imaging modalities in clinic for diagnosing cardiovascular diseases, which allow the
direct visualization of vessel wall from inside the lumen (Landini and Verrazzani 1990;
Potkin, Bartorelli et al. 1990; Huang, Swanson et al. 1991; Tearney, Jang et al. 2006).
Both modalities are based on imaging catheters. In the imaging process, a long catheter
is advanced into the coronary artery and the imaging probe is electrically scanned or
mechanically rotated 360˚ to get a cross sectional view of vessel wall. The process is
illustrated in Figure 1-2. By performing imaging pullbacks, 3D volumetric data sets
could be acquired. IVPA is a newly developed catheter based optical-acoustic hybrid
modality, which is capable of providing optical absorption mapping of the arterial wall.
IVPA is still under bench-top research.
4

Figure 1-2: An intravascular imaging catheter scanning inside a vessel (adapted from
www.trdesign.com/lightlab/oct.php)
1.2.1 Intravascular Ultrasound (IVUS) Imaging
Intravascular ultrasound (IVUS) imaging is currently the gold standard for
diagnosing coronary artery diseases, which has been researched and applied in clinic
for over 20 years. The common working frequencies are ranging from 20 to 60 MHz. In
clinical interventions, two ways of scanning schematic are widely utilized, shown in
Figure 1-3. The first type (iCross, Boston Scientific, Co. Natick, MA; Revolution,
Volcano, Co. San Diego, CA) involves a single element high frequency (40~45 MHz)
ultrasonic transducer, as shown in Figure 1-3 (a), which is mechanically rotated 360˚ by
a motor through a long drive cable. The mechanical rotation method could achieve a
frame rate of 30 fps. The fabrication of single element transducer and imaging system
are relatively less sophisticated. The second type (Eagle Eye, Volcano, Co. San Diego,
5

CA) applies a ring array transducer (20 MHz, 64 elements), as shown in Figure 1-3 (b),
which electrically scans the ultrasound beams inside the vessel lumen without rotating.
The electrical scanning methods could achieve a frame rate of 100 fps. The improved
scanning speed not only avoids motion distortion caused by transducer rotation, but
also reduces the time required for imaging pull-back, which increases surgery safety.
However, such scanning method relies on advanced ultrasonic array transducer and
sophisticated transmitting/receiving imaging electronics. Current technology has
limitation in fabricating ring array over 20 MHz, which is insufficient in resolution for
detecting certain critical vessel and plaque structures. Moreover, due to the relatively
few array elements, side lobes become a big concern and reduce the lateral resolution.
On the other hand, when goes back to the mechanical rotation method, with the
advancement of rotational shaft and drive cable, the obstacle of high speed frame rate at
100 fps could be overcome in the near future. IVUS at 40 MHz range could achieve
axial resolution of 60-100 µ m, and penetration depth more than 5 mm.

Figure 1-3: Two standard schematics of IVUS catheter: Single element transducer (A);
Array transducer (B)
6

1.2.2 Intravascular Optical Coherence Tomography (OCT) Imaging
Optical coherence tomography (OCT) is an optical modality for high-resolution
(10~30 μm), cross-sectional and real-time imaging (Huang, Swanson et al. 1991;
Patwari, Weissman et al. 2000). OCT is analogous to B-mode ultrasound except that
infrared light is used instead of ultrasonic waves. In OCT, a technique known as low-
coherence interferometry is applied to measure the time delay of back-scattered infrared
light instead of electronically measured as in ultrasound.
OCT has been performed for intravascular imaging and evaluation of vulnerable
plaques for over 10 years, and the use of OCT in coronary artery diseases detection is
intensively increasing (Pasterkamp, Falk et al. 2000; Tearney, Jang et al. 2006; Farooq,
Khasnis et al. 2009). The outstanding resolution advantage of OCT over other imaging
modalities makes it “the only method demonstrated to be capable of measuring all of
the microscopic features associated with TCFAs to date.”(Tearney, Jang et al. 2006).
However, the major limitation of OCT is the shallow penetration depth, which is less
than 1.25 mm (Jang, Bouma et al. 2002).
The first generation time-domain OCT product for intravascular imaging was
released in 2004 in Europe by LightLab Imaging, now acquired by St Jude Medical, Inc.
The second generation frequency-domain OCT has been approved by US FDA until
2011. The second generation OCT achieves a frame rate of 100 fps, which significantly
reduces the time and amount of saline for blood clearance.
7

The intravascular OCT catheter is similar as the single-element-transducer type
of IVUS catheter, which also requires mechanical rotation. The schematics of the IVUS
and OCT are shown in Figure 1-4. The usable length of the catheter is normally more
than 135 cm, and the imaging pull-back length is around 15-20 cm. The pull-back
speed is around 0.5-1 mm/s for IVUS, while the speed is three times faster in the
second generation OCT.

Figure 1-4: Schematics of an IVUS catheter (a) (adapted from
www.bostonscientific.com) and an OCT catheter (b) (adapted from www.lightlab.com)
1.2.3 Intravascular Photoacoustic (IVPA) Imaging
Intravascular photoacoustic (IVPA) imaging is a newly developed catheter
based optical-acoustic hybrid modality, which is capable of providing optical
absorption mapping of the arterial wall, offering not only morphological but also
functional information (Sethuraman, Aglyamov et al. 2007; Karpiouk, Wang et al. 2010;
Wang, Su et al. 2010; Jansen, van der Steen et al. 2011; Wei, Li et al. 2011). From a
short laser irradiation on biological tissues, electromagnetic waves are absorbed and
8

transient thermal expansion is followed. As a result, broad band ultrasonic waves are
generated and received by an ultrasonic transducer. In photoacoustic imaging, not only
back-reflected photons are used for generating signals as in OCT, the absorbed photons
are also applied to generate ultrasonic signals due to the thermal expansion effects,
hence the penetration depth goes beyond the light diffusion limit (Wang and Hu 2012).
From the synergy of acoustic and optical mechanisms, complementary structure and
composition information could be extracted from the IVUS and IVPA images.
The IVPA catheter is still under bench-top research. Early researchers use free
space light shinning from outside of the vessel for in vitro experiments (Sethuraman,
Aglyamov et al. 2007; Sethuraman, Amirian et al. 2007). More recently, integrated
IVUS-IVPA catheters that can be used inside the vessel lumen has been extensively
investigated. The head of the IVPA catheter normally incorporates an optical fiber for
light delivering and an ultrasonic transducer for acoustic wave detection. Cone-shaped
scan head, sequentially and con-focally arranged fiber and transducer have been
reported (Hsieh, Chen et al. 2010; Wang, Su et al. 2010; Jansen, van der Steen et al.
2011; Wei, Li et al. 2011).
1.3 Plaque Characterization by Imaging Interpretation
The atherosclerotic plaques are generally divided into three types: fibrous
plaque, which contains majorly fibrous constituents, such as smooth muscle cells with
subsequent synthesis of collagen, elastin, and proteoglycans; calcified plaque, which
9

contains calcific deposits; lipid-rich plaque, which contains fat-containing macrophages,
or foam cells. Some plaques are classified as hybrid types due to the complicated
constituents, such as fibroclacific and fibrofatty. Both IVUS and OCT images could be
used to differentiate the three types of plaques. A summary of the specific features in
charactering the three types of plaques in IVUS and OCT images are shown in Figure
1-5 and Figure 1-6 (Potkin, Bartorelli et al. 1990; Tearney, Jang et al. 2006).
The fibrous plaque, as shown in Figure 1-5 (B) and Figure 1-6 (B), displays as
concentric or eccentric thickening of intima in histology images. The corresponding
region in IVUS image, as shown in Figure 1-5 (A), appears bright and homogeneous,
while the region in the corresponding OCT image, as shown in Figure 1-6 (A), also
appears bright and homogeneous.
The calcified plaque is shown in Figure 1-5 (D) and Figure 1-6 (D). In IVUS
image, as shown in Figure 1-5 (C), the calcified plaque demonstrates very bright
surface with acoustic shadow behind it because acoustic waves could hardly penetrate it.
The calcified region in OCT image, as shown in Figure 1-6 (C), appears signal poor
with sharply delineated boarders.
The lipid-rich plaque, as shown in Figure 1-5 (F) and Figure 1-6 (F), appears
dark and hypoechoic in IVUS image, as shown in Figure 1-5 (E); while the region in
the corresponding OCT image, as shown in Figure 1-6 (E), appears signal poor but with
diffused boarders. During the tissue preparation for histology stain, the lipid
composition could be partially dissolved, as shown in Figure 1-5 (F).
10

Figure 1-5: IVUS images and corresponding histology for fibrous (A, B), calcified (C,
D), and lipid-rich (E, F) plaque types. Fib, fibrous plaque; Ca, calcified plaque; L, lipid-
rich plaque (adapted from (Potkin, Bartorelli et al. 1990))

Figure 1-6: OCT images and corresponding histology for fibrous (A, B), calcified (C,
D), and lipid-rich (E, F) plaque types. Fib, fibrous plaque; Ca, calcified plaque; L, lipid-
rich plaque (adapted from (Tearney, Jang et al. 2006))
11

1.4 Scopes of the Dissertation
In this dissertation, new intravascular imaging techniques have been
investigated either at higher working frequencies for IVUS or with integration of
conventional IVUS and optical modalities for diagnosing atherosclerosis. The goal of
this research is to improve intravascular imaging technology to provide higher
resolution, deeper penetration, and better contrast. The work could be divided into three
major parts: integration of IVUS and OCT for intravascular imaging; development of
very high frequency IVUS imaging; and IVPA imaging.
In chapter 2, the basic knowledge of B-mode ultrasound imaging, especially
IVUS imaging is introduced. The most important properties of ultrasound in high
frequency pulse-echo imaging are discussed. Starting from single element ultrasonic
transducer design and fabrication, the prototype IVUS transducer and imaging system
setup is introduced. The prototype IVUS transducer and system represents the current
IVUS technology that available in clinical use, and it is also the basis from which our
advancement progresses.
Chapter 3 works on the development of integrated IVUS-OCT imaging
catheters and system. The significance of developing the dual-function hybrid modality
is discussed. The imaging system setup is presented. After that, three generations of the
integrated IVUS-OCT probes are described in detail. Based on the current generation of
probe, in vitro and in vivo imaging results are presented, which demonstrate the
feasibility and superiority of the combination of the two modalities.
12

Chapter 4 focuses on the advancement of very high frequency IVUS technology.
The fabrication and characterization of 80 and 95 MHz IVUS transducers is introduced.
The in vitro imaging results are presented, discussed, and compared with 40 MHz
IVUS and OCT imaging results. The comparison reveals the pros and cons for IVUS
working at such high frequencies.
Chapter 5 presents our preliminary study on IVPA imaging. The system setup
and two types of probe designs are described first. A tissue mimicking phantom and in
vitro rabbit aorta imaging results are then presented and discussed.

13

Chapter 2 High Frequency Ultrasonic Transducer for IVUS
Imaging
2.1 B (Brightness)-Mode Ultrasound Imaging
B-mode ultrasound imaging is also called brightness-mode or gray scale
ultrasound imaging. It usually operates in pulse-echo mode, where the same ultrasonic
transducer transmits and receives ultrasonic waves.
Ultrasound is a compressional wave that can only propagate inside a medium.
Once the transmitting ultrasonic waves hit an interface of two media, where acoustic
impedances mismatch in the two media, parts of the ultrasonic waves will be bounced
back and received by the same transducer. The rest of ultrasonic waves continue
propagate until hit other interfaces. The position of each interface can be timely
resolved according to equation 2-1, where is the distance between transducer surface
and interface of acoustic impedance discontinuity, is sound speed in the medium, and
is the time delay of echo signal.
2-1
The echo signals received along one transmitting/receiving route is named one
A-line. If the ultrasonic transducer is physically scanned linearly or rotationally, a 2-D
image can be formed by incorporating multiple A-lines. The echo signal magnitude is
mapped to the brightness in the gray scale image. In IVUS imaging, a single element
14

transducer is rotated 360˚ to scan a cross section of vessel wall and form a radial format
image.
The IVUS imaging processing procedure is described in Figure 2-1. The raw RF
data is first digitally filtered within the effective bandwidth, in order to remove the
unwanted noises that are out of the bandwidth. After filtering, the RF data goes through
envelop detection and log compression. The compressed data is then digitally converted
into radial format for displaying. Multiple averaging methods could be applied to the
RF data or compressed data to reduce the background noise level. However, in the case
that high frame rate is desirable, less A-lines are acquired, thus averaging may not be
feasible.

Figure 2-1: The procedure of processing the IVUS RF data
2.2 Resolution, Attenuation and Backscattering in High Frequency Ultrasound
2.2.1 Resolution
Ultrasound refers to the sound with frequency above the human hearing range
(20~20,000 Hz). High frequency ultrasound generally refers to ultrasound with
frequency above 15 MHz. The working frequency directly determines the spatial
resolutions of an image, described in equations 2-2 and 2-3.
15

2-2

2-3
where is the speed of sound,

is the center frequency of a transducer,

is defined
as the ratio of focal distance to the aperture diameter of a transducer, and

represents the -6 dB bandwidth of a transducer. It is obvious that increasing the center
frequency will improve both axial and lateral resolutions. Equation 2-2 is valid only for
a focused transducer in the focal zone.
For an unfocused piston single element transducer, the focal distance can be
considered as the near field size, or the nature focus, which is described as below,

2-4
where

is the nature focus, is the radius of the transducer aperture, is the
wavelength at the transducer’s center frequency. For a single element transducer
working at 40 MHz with an aperture radius of 0.25 mm, the nature focus is around 1.7
mm in water. For an 80 MHz transducer at the same size, the nature focus increases to
3.4 mm. Within the nature focusing zone, the beam width is around to 2 . Beyond
the nature focusing zone, acoustic beams start to diverge.
2.2.2 Attenuation
Although high frequency ultrasound is able to provide excellent spatial
resolutions, a major concern is the strong attenuation effect, which will drastically
16

limits the penetration depth of ultrasonic waves. The attenuation of acoustic energy is
caused by absorption and scattering in soft tissues, which could be described
mathematically as below,
0
z
P P e
 
 2-5
Where P
0
is the ultrasonic pressure at transducer surface, α is attenuation coefficient
and z is the sound traveling distance. Attenuation coefficient α is frequency dependent
and also has strong dependence on tissue types. It is defined as below,


  f
0
 2-6
where
0
 is the attenuation coefficient at 1 MHz and γ is the frequency dependence
parameter. Attenuation coefficient from 10 to 100 MHz is summarized and plotted in
Figure 2-2.

Figure 2-2: Attenuation of variety tissues in the frequency range from 10 to 100 MHz
(Foster, Pavlin et al. 2000)

17

At 80 MHz, an attenuation coefficient of 10 dB/mm is expected for coronary
artery wall, which means that a penetration depth of 3 mm can be achieved for a system
with a dynamic range of 60 dB (Foster, Pavlin et al. 2000). Yet such a system requires
highly sensitive miniature IVUS transducers, which is a great challenge.
2.2.3 Backscattering
Another important issue in very high frequency IVUS application is the blood
backscattering effect. Early investigator has reported that while the blood backscatter at
35 MHz is not significant, however, at 65 MHz, the value is closed to or exceeds that
from muscular media and greatly exceeds the values from soft plaque, as shown in
Figure 2-3. The blood backscatter becomes even stronger when goes beyond 65 MHz
(Foster, Obara et al. 1994). The finding also implies that backscatter is stronger at body
temperature than that at room temperature. The backscattering effect may dramatically
degrade IVUS imaging contrast between blood and vessel at frequencies beyond 65
MHz; limit its ability to see through blood and even further reduce the penetration
depth in vessel wall. In this case, saline flushing for blood clearance could be used to
overcome this problem.
18

Figure 2-3: Plots of backscatter coefficients versus shear rates at 65 MHz (Adapted
from Foster, Obara et al. 1994)
2.3 Single Element Ultrasonic Transducer
2.3.1 Piezoelectric Effects
The finding of piezoelectric effect could be dated back to 1880s, when Curie
brothers discovered that mechanical pressure could be generated when a rapid changing
electrical potential was applied on quartz crystal, and vice versa.
Not only quartz, many polymers, ceramics and crystals also display superior
piezoelectric properties, such as polyvinylidene fluoride (PVDF), lead zirconate titanate
(PZT), and lead magnesium niobate-lead titanate (PMN-PT). Piezoelectric material
19

plays a crucial role in fabricating ultrasonic transducers, since it is the only active
material inside the transducers.
The merits of the piezoelectric materials are mostly determined by
electromechanical coupling coefficient (k
t
), piezoelectric strain constant (d
33
) and
dielectric permittivity ( ε
r
/ε
0
). Piezoelectric materials with high k
t
and d
33
values are
efficient in energy conversion, suggesting improved sensitivity of the transducers.
Meanwhile, ε
r
/ε
0
is a critical issue considering the electrical impedance matching of the
transducers to the 50 Ω imaging electronics, which would affect the sensitivity in both
transmitting and receiving and the ring-down effect in imaging. The electrical
impedance of a transducer is inversely proportional to the piezoelectric material’s
surface area and ε
r
/ε
0
value (Cannata, Ritter et al. 2003). For a miniature IVUS
transducer, materials with high dielectric permittivity value are desirable.
Among all piezoelectric materials, [Pb(Mg
1/3
Nb
2/3
)O
3
]
0.63
[PbTiO
3
]
0.37
(PMN-PT)
single crystal is a promising candidate with high k
t
(0.58, HC materials, Bolingbrook,
IL), ε
r
/ε
0
(5,229) (Kosec, Holc et al. 2007) and d
33
(2000 pCN
-1
) (Calzada, L. et al. 2009)
values. This material is the major type of piezo-material used throughout my IVUS
researches.
2.3.2 Piezoelectric Film Technology
When researching on very high frequency (>80 MHz) transducers, the difficulty
comes from the preparation of a very thin piezoelectric layer with properties similar to
20

bulk materials. Traditional lapping of crystal or ceramic bulk down to the thickness of
20-30 μm is extremely difficult and time consuming. Relatively large thickness
variation (~3 μm) and crack nature make the quality of transducers hard to predict.
Additionally, the degradation of dielectric and electromechanical properties of single
crystals with decreasing thickness may downgrade the sensitivity of a high frequency
transducer (Lee, Zhang et al. 2010). A feasible solution is to pursue piezoelectric thin
film technology, which is a promising solution for building miniature very high
frequency transducers.
In previous work, sol-gel lead zirconate titanate (PZT) and sputtered zinc oxide
(ZnO) thin films have been investigated for fabricating very high frequency (>80 MHz)
ultrasonic transducers (Cannata, Williams et al. 2008; Zhu, Wu et al. 2008; Lee, Zhang
et al. 2010; Zhou, Lau et al. 2011). However, the sensitivity of sol-gel PZT transducers
is relatively low. The drawbacks of ZnO as active transducer elements are the relative
low k
t,
(0.28) and ε
r
/ε
0
(8) values, which make them only suitable for fabricating large
aperture transducers.
In order to investigate IVUS imaging at 80 MHz or higher, we choose to use
PMN-PT free-standing film technology for building the miniature very high frequency
transducers. In contrast to normal thin films that require substrates which degrade the
piezoelectric properties of films, free-standing films were produced without substrate.
PMN-PT free-standing film was synthesized with a modified precursor coating method
(Luo, Shih et al. 2007).
21

2.3.3 Single Element Ultrasonic Transducer Design
Single element ultrasonic transducer and imaging system are relatively less
sophisticated to build, but require mechanical scanning, compared to the array
transducers and system. However, single element transducer could be easily
miniaturized to the size less than 1 mm, which is critical for IVUS applications. The
criteria for a high quality ultrasonic transducer involve wide frequency response or
broad bandwidth; good acoustic impedance matching to biological tissues; and high
efficiency in both transmitting and receiving.
To achieve these criteria, the selection of piezoelectric material is very
important. Moreover, two layers of acoustic matching scheme is suggested to improve
the performance of transducers (Cannata, Ritter et al. 2003). The acoustic impedances
of the two matching layers should be respectively equal to: (Desilets, Fraser et al. 1978)
7 / 1 3 4
1
) (
l p m
Z Z Z  2-7
6 1/7
2
()
m p l
Z Z Z  2-8
where Z
l
and Z
p
are acoustic impedances of the loading medium (water, 1.5 MRayls)
and piezoelectric material, respectively. Popular acoustic matching material includes
silver epoxy (7.3 MRayls) and Parylene (2.3 MRayls), which are close to the
requirements above. A conductive material, E-solder 3022 (VonRoll Isola, New Haven,
CT) is applied to the bottom as the acoustic backing. After centrifuging, the backing
material has an acoustic impedance of 5.92 MRayls and loss of 110 dB/mm (Cannata,
Ritter et al. 2003).
22

The thicknesses and dimension of each matching layer and piezoelectric layer
could be optimized on a commercial software based on Krimhotz, Leedom, and
Matthaei (KLM) equivalent circuit model (PiezoCad, Sonic Concepts, Woodinville,
WA), in order to achieve the best performance of a transducer.
2.4 Prototype IVUS Transducer
2.4.1 IVUS Transducer Design and Fabrication
For bench top experiments, we designed and fabricated 40 MHz side-viewing
miniature IVUS transducers as a prototype. The prototype transducer was used as a
representative of current 40 MHz single element IVUS transducer technology.
A piece of PMN-PT single crystal sheet (HC materials, Bolingbrook, IL, USA)
was used as the active piezoelectric layer to fabricate the side-viewing miniature
transducers. According to PiezoCad simulations, the design parameters were
summarized in Table 2-1.

23

Table 2-1: Design parameters of 40 MHz IVUS transducer
Specifications Values
Parameters used in transducer design
Center frequency 40 MHz
Bandwidth 50~60%
Piezoelectric material (PM) PMN-PT
Thickness of PM 48 µ m
1
st
matching layer (ML) 2-3 µ m Silver epoxy
Thickness of 1
st
ML 18 µ m
2
nd
ML Parylene
Thickness of 2
nd
ML 13 µ m
Backplate E-Solder 3022
Thickness of Backplate 0.4 mm
Aperture Dimension 0.4 mm× 0.4 mm

The PMN-PT sheet was first lapped to the desired thickness then sputtered with
Cr/Au (500Å/1000Å) layers as electrodes at top and bottom. A silver epoxy matching
layer made from Insulcast 501, Sulcure 9 (American Safety Technologies, Roseland,
NJ) and 2-3 µ m silver particles (Sigma-Aldrich Inc., St. Louis, MO) was then cured
over the top of the PMN-PT sheet and lapped to designed 18 µ m. A conductive backing
material, E-solder 3022 (VonRoll Isola, New Haven, CT) was then applied to the
bottom of the film and lapped to 0.4 mm. The active stack was diced along the
thickness direction into small posts with the aperture of 0.4 mm × 0.4 mm. The post
was housed within a 0.57-mm-ID polyimide tube (Small Parts, Inc., Miramar, FL), on
the side of which a window was opened to allow the transducer to stand toward side
face. A 0.25-mm-OD 50 Ω electrical wire was connected to the conductive backing
using E-solder 3022 inside the polyimide tube. The polyimide tube provided the
electrical isolation from the outer stainless steel needle housing. The outer needle
housing with an ID of 0.66 mm and OD of 0.92 mm was opened a window on the side
24

for acoustic wave to go through. 5-min epoxy (Henkel Corporation, Irvine, CA) was
filled into the gap between piezoelectric post and needle housing to insulate the inner
electrode. Another Cr/Au electrode was sputtered over the silver epoxy matching layer
and stainless steel needle housing to form the ground connection. A 13-µm-thick
parylene layer (± 0.5 µ m) was vapor-deposited onto the transducer and needle housing
to serve as second matching and protecting layer. The transducer was finally connected
to a brass holder and SMA connector for mechanical holding and electrical connection.
To enhance the piezoelectric activity of the PMN-PT single crystal, the finished
transducer was poled in a DC electric field of 20 KV/cm for 5 minutes at room
temperature. The prototype needle IVUS transducer is shown in Figure 2-4 (a). Another
flexible version is fabricated with 1.5 m long double wound flexible torque coil with an
outer diameter of 0.65 mm, as shown in Figure 2-4 (b). The torque coil allows for
smooth torque translation to the distal end throughout the whole catheter.

Figure 2-4: Prototype IVUS transducers: needle type (a); and flexible type (b)
25

2.4.2 Transducer Performance Characterization
The side-viewing miniature transducer’s performance was characterized in a de-
ionized water bath at room temperature. Pulse-echo test (Cannata, Ritter et al. 2003)
was conducted with an X-cut quartz as the signal reflecting target. A broadband
negative pulse with approximately 100 Vpp and 200 Hz repetition rate emitted from a
pulser/receiver unit (PR5900 Olympus NDT, Inc., Kennewick, WA) was used to excite
the transducer. Echo signals were received and digitized by a 1 GHz oscilloscope
(LC534, LeCroy Corp., Chestnut Ridge, NY). The frequency response of the transducer
was analyzed from the echo waveform, shown in Figure 2-5. The Centre frequency (
c
f )
and -6 dB fractional bandwidth (BW) could be determined by the following Equations:
2
u l
c
f f
f

 2-9
% 100 


c
l u
f
f f
BW 2-10
where
l
f and
u
f were defined as lower and upper -6dB frequencies, at which the
magnitude of the amplitude in the spectrum is 50% (-6dB) of the maximum. The
measured center frequency was 42 MHz and -6dB fractional bandwidth was 53%.
26

Figure 2-5: Pulse-echo measurement of one 40 MHz IVUS transducer

Two-way insertion loss (IL) was calculated using the ratio of the frequency
spectrum of the transmitted and received responses (1MΩ coupling). The following
equation was used for calculation,
2 4
2 10 2 . 2 9 . 1 log 20
c
T
R
f d
V
V
IL      

2-11
where
T
V and
R
V were transmitting and receiving amplitudes (V); d was the distance
(mm) between target and transducer surface. The imperfect reflection from quartz
crystal was compensated by 1.9 dB. The signal loss due to attenuation in water was
compensated by 2.2· 10
-4
dB/mm· MHz
2
(Cannata, Ritter et al. 2003). Two way insertion
loss value was measured to be 14 dB at 40 MHz. Five 6-μm-OD tungsten wire targets
were imaged to determine axial and lateral resolutions of the transducer, as shown in
2.3982 2.5237 2.6492 2.7747 2.9002
-500
-250
0
250
500
Time ( s)
Amplitude (mV)

10 30 50 70 90
-24
-18
-12
-6
0
Frequency (MHz)
Magnitude (dB)
Pulse-echo Responsee
Spectrum
27

Figure 2-6. The axial and lateral resolutions were determined from the -6 dB envelope
width for the wire located at around 1 mm, which were 57 μm and 425 μm, respectively.

Figure 2-6: Wire phantom image from one 40 MHz IVUS transducer

The performance results were summarized in Table 2-2.
Table 2-2: 40 MHz IVUS transducer testing results
Specifications Values
Measurements of prototype IVUS transducer
Center frequency 42 MHz
Bandwidth 53%
Insertion loss 14 dB
Axial resolution (-6dB) 57 μm
Lateral resolution (-6dB) 425 μm
Lateral distance [mm]
Axial distance (Depth) [mm]
PMN-PT
1 2 3 4 5 6 7 8 9 10
1
2
3
4
28

2.4.3 IVUS System Setup and Imaging Experiments
The imaging system is illustrated in Figure 2-7. A Panametrics PR5900
pulser/receiver (Olympus NDT, Inc., Kennewick, WA) is used to excite the transducer
and also receive the echo signals. RF data is digitized by a 12-bit data acquisition board
(Gage Applied Technologies, Lockport, IL) with a sampling rate of 400 MHz. The
IVUS transducer is mounted to a custom built rotational joint for electrical coupling
from the rotational part to the stationary part, as shown in Figure 2-8. An electrical slip
ring is used to connect the rotational motor (Animatics, Santa Clara, CA) and IVUS
transducer. The electrical wire on the rotational part of the slip ring is connected to
transducer and the wire on the stationary part is connected to the pulser/receiver. The
stepper motor provides trigger signals at every step to trigger a function generator
which then synchronizes the pulser/receiver and data acquisition board. 1000 A-lines
are acquired during each revolution and the scanning procedure is controlled by a
custom built LabVIEW (National Instruments, Austin, TX) program. RF data are saved
and post processed for image display. Images are displayed with 50 dB dynamic range.
29

Figure 2-7: IVUS imaging system
In vitro imaging of post mortem human coronary artery specimens was
performed. During experiment, the tip of a transducer was positioned inside lumen area
of a specimen, which was immersed in water and supported by a sponge to stand in a
water tank. Only the part of sample above the sponge was imaged. The circumferential
scanning was achieved by rotating the transducer while the specimen was kept
immobile.

Figure 2-8: Custom built rotational coupling joint
30

An in vitro IVUS image from a 40 MHz transducer was shown in Figure 2-9.
The image was able to differentiate the intima, media and adventitia layers, and also
able to see a calcified plaque and lipid content behind it. However, it was not accurate
to provide detailed information such as the thickness of plaque cap, due to the
insufficient resolution at 40 MHz.

Figure 2-9: IVUS image of human coronary artery at 40 MHz. I, intima; M, media; A,
adventitia; Ca, calcified plaque; L, lipid-rich plaque

31

Chapter 3 Intravascular Ultrasound Combined with Optical
Coherence Tomography Imaging
3.1 Advantages of Integrated IVUS-OCT Imaging Modality
The advantages of ultrasound as a popular means for intravascular imaging stem
from its large penetration depth and moderate resolution in resolving plaque, lipid pool
and vessel structures (Foster, Pavlin et al. 2000; Sawada, Shite et al. 2008). In addition,
blood serves as a nature coupling medium for ultrasound, rather than an obstruction for
optical modalities. However, the insufficient resolution of ultrasound imaging becomes
a major limitation in characterizing certain critical microstructures of atheromatous
plaque, for example, an accurate estimation of cap thickness of fibrous plaque in thin-
cap fibroatheroma (TCFA). On the other hand, OCT has been used for atherosclerosis
evaluation and shown to provide valuable information about microscopic features
(Patwari, Weissman et al. 2000; Kawasaki, Bouma et al. 2006; Tearney, Jang et al.
2006). However, the limited penetration depth (1.25 mm) prevents OCT from imaging
the whole depth of a large lipid pool of a plaque. Meanwhile, flushing or occlusion is
needed for blood clearance in the clinical settings.
Since the two modalities possess complementary properties to each other, where
OCT’s high resolution could resolve superficial microstructures and ultrasound could
penetrate the whole depth of vessel wall, an integrated imaging system combining the
two would be more beneficial than either alone. Moreover, a minimal amount of
32

flushing agent would be needed for OCT since ultrasound can serve as guidance while
searching for targets in the blood vessel. The feasibility of combining information from
intravascular OCT and IVUS for detecting TCFA has been studied by several
investigators (Sawada, Shite et al. 2008),

who acquired ultrasound and OCT images
from separate systems. The results indicate that neither modality alone is sufficient for
detecting TCFA, while the combined use of OCT and IVUS provides much better
sensitivity and specificity for evaluating TCFA. However, using two separate catheters
and systems is time consuming, which requires patients endure extra suffer and may
increase safety risks. Moreover, separately acquired IVUS and OCT images may not be
fully coregistered, which could result in inaccurate diagnosis. An integrated IVUS-OCT
catheter and system is a solution to overcome all the problematic issues.
3.2 Integrated IVUS-OCT System
3.2.1 System Setup
The schematic of the integrated IVUS-OCT system is shown in Figure 3-1 and a
photograph of the system is shown in Figure 3-2. The system can be roughly divided
into four parts: IVUS subsystem, OCT subsystem, timing and motor controller, data
acquisition and processing components.
In the IVUS subsystem, a Panametrics PR5900 pulser/receiver (Olympus NDT,
Inc., Kennewick, WA) is used to excite the transducer and also receive the echo signals.
33

2 µ J pulse energy is used to drive the transducer, and 26 dB gain plus 10~100 MHz
band-pass filter is applied to the received RF signal.

Figure 3-1: Schematic of IVUS-OCT imaging system. The blue blocks represent signal
flow; the orange blocks represent the mechanical joints; the green blocks represent
synchronizing triggers

Figure 3-2: A photograph of the integrated IVUS-OCT system

34

For the swept-source OCT (SS-OCT) subsystem (Su, Zhang et al. 2008), light
from a SS (center wavelength, 1310 nm; FWHM bandwidth, 100 nm; output power, 2.7
mW; scanning rate, 20 KHz; Santec Corp., Komaki, Aichi, Japan) is split by an 80/20
1× 2 coupler, with 80% of the power directed to the IVUS-OCT probe and the
remaining 20% to the reference arm. Two circulators are used in both arms to redirect
back-scattered and back-reflected light to the two input ports of a 50/50 2× 2 coupler for
balanced detection.
The light source generates 20 KHz trigger signals which drive a function
generator (Agilent Technologies, Inc., Santa Clara, CA) serving as a frequency divider
to provide up to 10 KHz triggers to synchronize the data acquisition board and the
pulser/receiver. Both OCT and IVUS signals are digitized by a two-channel, 12 bit data
acquisition board (Alazar Technologies Inc., Pointe-Claire, QC, Canada) working at a
sampling rate of 200 MHz. The 200 MHz clock is provided by an external voltage
controlled oscillator. The acquired IVUS and OCT signals are real time processed,
saved and displayed by a custom built program.
A custom built rotary joint device is used for motion control and signal coupling
from the rotational part to the stationary part. The device is shown in Figure 3-3. The
rotational motor (Animatics, Santa Clara, CA, USA) is mounted to a gear fixture with
gear ratio of 2:1. A fiber optic rotary joint (Princetel, Inc., Pennington, New Jersey) and
an electrical slip ring (Prosperous, Co., Hangzhou, China) are used for signal coupling.
All these components are fixed to a translational stepper motor which functions for
imaging pull-back.
35

For the OCT subsystem, the axial and lateral resolutions are optimized and
measured to be 8 μm and 30 μm and the working distance is 3 mm. For the IVUS
subsystem, the axial and lateral resolutions are measured to be 57 μm and 425 μm for
using 40 MH transducers and working distance is over 7 mm. The frame rate of the
system is up to 20 fps with 500 A-lines per revolution and 8192 sampling points in each
A-line for real time displaying and raw data saving. Pull-back speed is adjustable,
normally 0.5~1 mm/s.

Figure 3-3: A rotary joint device connects the rotational and pull-back motor; and
couples electrical and optical signals from the rotational part to the stationary part
3.2.2 Signal Acquisition and Processing
The image processing for the combined IVUS-OCT system works across 4
threads which are capable of working concurrently with each other. The first major
thread is the data acquisition thread, which handles interfacing with the Alazar 9350
digitizer (ATS 9350 Alazar Technologies Inc, QC, Canada) and acquiring the OCT and
IVUS data from two channels and storing them into buffers for processing. For each
A-line scan, the digitizer samples 8192 12-bit values at 200 MHz using its internal
36

clock, ultimately scanning 500 or 1000 A-lines in buffers for 1 frame of data. When a
buffer has been filled, the data acquisition thread signals the image processing thread
that it may begin to process the captured data and also signals the save thread to save
the raw data if that is desired. Meanwhile, the data acquisition thread begins the process
for capturing the next buffer of data. The image process begins by separating the
interleaved OCT and IVUS data. Using the OpenMP API, the two data sets are
processed concurrently using all 16 CPU threads. The OCT data is split into 6 portions
and each portion is computed on its own processor thread. Each thread is performing a
linear K-domain transform and FFT on each A-line within the data finally followed by
a logarithm and contrast scaling. At the same time, the remaining 10 threads are
processing the IVUS set of data, performing a logarithm and scaling of the data out of
the transducer. When all the parallel threads finish their processing, a signal is issued to
the display thread to display and optionally save BMP images of the data.
A newly created GPU version of the code works similarly to the CPU version
with the exception of having many more threads available to process the data. The
DAQ, save and display thread all work the same. The image process thread now
transfers the entire buffer of data straight to the GPU’s on board memory. The data is
again split into an OCT array and an IVUS array, but afterwards the GPU’s parallel
processing takes over. For the OCT portion, instead of having to repeatedly perform a
FFT for each individual A-line, the GPU is able to perform the multiple FFTs at the
same time. Likewise, the processing for the linear K transform and logarithms also
occur in parallel instead of sequentially like the CPU’s version does (nested for loops).
37

The IVUS data naturally also reaps the benefits of parallel processing. The overall
speed up is fairly significant; the CPU version of the system was able to process 4000
lines of data in about 0.9 seconds. The new GPU version is able to process 20K lines of
data in the same amount of time and there is still some room for improvement in the
code. Another means of improvement could be using two GPU’s, one for the OCT data
and one for the IVUS data to further decrease processing time.
3.3 First Generation Probe Design: Side-by-Side Arrangement
3.3.1 Probe Design
The schematic of the first generation dual-modality probe that combines OCT
optical components with an IVUS transducer is shown in Figure 3-4. A 0.5-mm-diam
gradient index lens (NSG America, Inc., Somerset, NJ) is used to focus light from a
single-mode fiber tip, followed by a microprism (Tower Optical Corp., Boynton Beach,
FL) reflecting the focused light beam into tissue. These optical components are fixed
and sealed in a transparent thin-wall polyertrafluoroethylene tubing (Zeus, Inc.,
Orangeburg, SC) with an inner diameter of 0.6 mm and a wall thickness of 50 µ m. The
axial and lateral resolutions of the OCT system with probe are 8 and 20 µ m,
respectively. The rest of the bare optical fiber is protected by a 0.51-mm-diam flexible
stainless steel tubing, which ensures efficient transmission of rotational torque from the
rotary joint to the probe distal end. As for the IVUS part, a single element, unfocused
38

PZT transducer (PZT-5H, 40 MHz, aperture area 0.16 mm
2
) is fabricated with a
fractional bandwidth of 50% at -6 dB. The measured two-way insertion loss for this
miniature transducer is 26 dB at 40 MHz in water, and its axial and lateral resolutions
are approximately 38 and 400 µ m, respectively. The OCT probe and IVUS transducer
are fixed side by side in a stainless steel tube housing on which a window is made to
allow the light beam and sound wave to exit. The integrated OCT-IVUS probe has a
maximum outer diameter of 2.4 mm. To ensure coregistered images, the transducer
element is placed next to the exit light beam so that both OCT and IVUS viewed at
approximately the same cross section during imaging.

Figure 3-4: Schematic of IVUS-OCT probe
3.3.2 Imaging Results
In vitro imaging of normal rabbit aorta was performed using the integrated
IVUS-OCT system to demonstrate its feasibility in intravascular imaging, as shown in
Figure 3-5. It can be clearly indentified that the OCT image, as shown in Figure 3-5 (a),
offers higher resolution than the IVUS image, as shown in Figure 3-5 (b), and provides
39

more detailed information on tissue structure, which is essential for thickness detection
of the thin fibrous cap on a TCFA. However, only about a 500-µ m-thick maximum
depth can be visualized in Figure 3-5 (a); the penetration depth is even shallower
(region I) when the tissue is far away from the probe due to sensitivity roll-off. On the
other hand, although its resolution is inferior to that of OCT, the IVUS image provides
a full-depth cross-sectional image of the aorta, even the cork that was used to position
the tissue can be seen (region II). The deep imaging depth is critical for visualizing the
large lipid pool within a TCFA, which is one of the important features for plaque
characterization. Therefore, the integration of the two imaging modalities can offer
complementary information for atherosclerosis diagnosis that cannot be obtained by
either modality alone.

Figure 3-5: OCT (a) and IVUS (b) images of rabbit aorta acquired by the first
generation probe

40

3.4 Second Generation Probe Design: Coaxial Arrangement
3.4.1 Probe Design
The schematic of the second generation IVUS-OCT probe is shown in Figure
3-6 (Li, Yin et al. 2010). The home-made 50MHz focused ring transducer has an
effective aperture of 2 mm OD with a 0.8 mm hole at the center to make room for the
OCT probe which has an OD of 0.7 mm. The coaxial ultrasound and light beams have a
common focal length of 4 mm, and both are steered into tissue by a 45˚ mirror along
their pathway. The glass mirror, coated with Aluminum, is fixed close to the anterior
surface of the hybrid probe to ensure both beams are focused at tissue target. The
mirror and hybrid probe are properly aligned and packaged in a brass tube housing on
which a window is made to allow ultrasound and light beams to exit. After packaging,
the monolithic IVUS-OCT probe has a maximum OD of 2.5 mm. The probe is then
connected to a stepper motor and rotary joint device to enable rotational scanning inside
vessel lumen.

Figure 3-6: Schematic of coaxially arranged IVUS-OCT probe
41

3.4.2 Ring Transducer Characterization and Mirror Effects
In order to determine the performance of ultrasound subsystem, especially the
mirror reflecting effects on the propagation of ultrasonic beams, a series of
measurements were performed and showed that there was slightly loss in energy or
change in beam profile. As an ultrasonic wave encounters an interface between two
media, the directions of reflected and refracted waves are governed by Snell’s law
(Shung 2006; Cobbold 2007). The incident critical angle (
ic
 ) of total internal
reflection are defined as ) c / c ( sin
2 1
1
ic

  , where
1
c and
2
c are sound speeds in each
medium.
ic


at water and glass mirror interface can be calculated to be 15
˚
. In our
device, the incident angles of the focused sound beam with

of 2 encountering a 45
˚

mirror are calculated to be between 31˚ to 59˚, which are much greater than
ic
 . Thus,
‘total internal reflection’ occurs at the interface of water and mirror, or there is no
propagation loss due to refraction or shear wave mode conversion (Cobbold 2007).
However, in B-mode ultrasound, images are acquired in pulse-echo mode. Ideally,
ultrasound echo would return along the same pathway as incident wave, or ‘total
internal reflection’ also occurs for echo wave, which was further investigated.
To evaluate the mirror effects on the two-way propagation of ultrasonic wave,
several tests without/with a mirror were carried out. First, pulse-echo and two-way
insertion loss measurements were conducted. Comparing results without/with a mirror,
no significant difference in center frequency (51/50 MHz), fractional bandwidth
42

(58/58 %) or sensitivity (-24/-25 dB) are observed. Next, to determine the ultrasound
subsystem’s spatial resolution, the linearly B-scan images of five 6-μm-diam tungsten
wires were examined, which could be treated as ideal line targets since they were much
thinner than the average ultrasound wavelength (30 μm). IVUS images of wire
phantoms without/with a mirror are shown in Figure 3-7(a)(d) The two images clearly
demonstrate that the ultrasound beams are still focused at the designed focal point (4
mm) after reflected by the glass mirror. Figure 3-7 (b)(c)(e)(f) display the envelopes of
the RF signals acquired from the wires located at transducer’s focal point. Axial (R
axial
)
and lateral (R
lateral
) resolutions of ultrasound are determined from the -6 dB envelope
width. Without a mirror reflection, R
axial
and R
lateral
are 25 μm and 48 μm, respectively;
while with a mirror reflection, those are 22 μm and 44 μm, remain almost unchanged.
Both situations are consistent with the theoretical values of 25 μm and 45 μm for a ring
transducer. (Theoretically: R
axial
=c/(2BW); R
lateral
≈0.75λf/d (Cobbold 2007); c, sound
velocity in water; BW, -6 dB bandwidth; λ, average wavelength; f, focal length; and d,
outer diameter of transducer aperture.)
43

Figure 3-7: Ultrasound wire phantom images without a mirror(a) and with a mirror(d),
displayed with a dynamic range of 50 dB; axial and lateral envelopes of the RF echo
signals from the wire located at transducer focal point without a mirror(b)(c) and with a
mirror(e)(f)
3.4.3 Imaging Results
In order to demonstrate the feasibility of this coaxial IVUS-OCT probe in
intravascular imaging, an in vitro study of a healthy rabbit aorta was performed. The
aorta was fixed in 10% formalin for 24 h and preserved in phosphate buffer. During the
experiment, the aorta was pinned to a piece of cork and immersed into saline. The
results are shown in Figure 3-8. It is not surprising that OCT image [Figure 3-8 (a)]
44

resolves the inner profile of aorta more precisely and offers much higher resolution
(smaller speckles) than ultrasound image [Figure 3-8 (b)]. However, only about 500-μm
depth could be visualized in OCT image. On the other hand, ultrasound image provides
a full-depth cross-section image of the aorta. A combined IVUS-OCT image is
displayed in Figure 3-8 (c), which clearly demonstrated the advantage in high
resolution of OCT and deep penetration of IVUS, which can offer complementary
information for atherosclerosis diagnosis that cannot be obtained by either modality
alone. Meanwhile, the similar shapes of OCT and IVUS images illustrate the
superiority in depicting coregistered intravascular images of the coaxial hybrid probe
design and integrated imaging system.

Figure 3-8: OCT (a), IVUS (b) and combined IVUS-OCT (c) images of a rabbit aorta. I,
Cork holder
45

3.4.4 Downsized Version: Miniature Ring Transducer
We have further downsized this coaxial design to 1.2 mm of OD in order for
future animal and clinic studies, as shown in Figure 3-9. The miniature ring transducer
has an outer diameter of 1 mm and an inner diameter of 0.45 mm. PMN-PT single
crystal is used for fabricating this miniature transducer, as it has high k
t
and ε
r
/ε
0
values,
which could improve transducer’s sensitivity. Pulse echo test result is shown in Figure
3-10. The transducer has a center frequency of 44.4 MHz with 51 % bandwidth. Two-
way insertion loss test shows the sensitivity is 23 dB.

Figure 3-9: Downsized version of the coaxial arranged IVUS-OCT probe. A miniature
ring transducer with an outer diameter of 1 mm and an inner diameter of 0.45 mm (a);
the transducer is assembled with an OCT probe in the center and a 45˚ prism at the tip
(b), the pink spot is the reflection from an visible light source; and the whole probe (c)

Figure 3-10: Pulse-echo measurement of the miniature ring transducer, which has a
center frequency of 44.4 MHz with 51 % bandwidth
5.1046 5.2298 5.3551 5.4803 5.6056
-150
-75
0
75
150
Time ( s)
Amplitude (mV)

15 28.75 42.5 56.25 70
-24
-18
-12
-6
0
Frequency (MHz)
Magnitude (dB)
Pulse-echo Responsee
Spectrum
46

By using this downsized version, we imaged a segment of rabbit aorta, as shown
in Figure 3-11. The results demonstrated the feasibility of this smaller version of
coaxial design.

Figure 3-11: In vitro rabbit aorta images from the downsized version of coaxially
arranged OCT and IVUS probe
3.5 Third Generation Probe Design: Sequential Arrangement
3.5.1 Probe Design
In order to transfer this technology to animal study and clinical use, we further
developed a miniature version of probe. The schematic of the miniature IVUS-OCT
probe is shown in Figure 3-12 (a). Figure 3-12 (b) shows the tip of the probe. Unlike
previously two designs which had a parallel or coaxial arrangement, this new version of
the probe design features a sequential arrangement of the IVUS transducer and OCT
probe so that over-all probe size is significantly decreased by roughly 4-fold. Within
47

the OCT probe, a 0.35 mm-diameter gradient index (GRIN) lens (NSG America, Inc.,
Somerset, NJ) was used for light focusing, followed by a 0.3 mm-diameter micro-prism
(Tower Optical Corp., Boynton Beach, FL) for reflecting the focused light beam into
tissue. All the optical components were fixed in a polyimide tube with an outer
diameter of 0.41 mm and wall thickness of 0.025 mm (Small Parts, Inc., Miramar, FL).
The working distance of the OCT probe was about 3 mm. The ultrasonic transducer
with an aperture size of 0.4 mm × 0.4 mm was built using a PMN-PT single crystal
(H.C. Materials, Bolingbrook, IL) which has superior piezoelectric properties for
building high sensitivity ultrasound transducers in a small size. The center frequency of
the ultrasound transducer was 40 MHz with a fractional bandwidth of 52%. The two
way insertion loss was measured to be 15 dB at the center frequency. The transducer
was fixed in the distal end of a thin-wall stainless tube (OD 0.64 mm, wall thickness
0.051 mm, Cadence, Inc., Staunton, VA) within which the OCT probe was also fixed.
A window was made on the tube to let both the light beam and acoustic wave exit.
Finally, the transducer wire and optical fiber were confined in a double wound flexible
torque coil with an outer diameter of 0.65 mm. It was known beforehand that the
transducer and prism were 2 mm apart, thus coregistered OCT and IVUS images could
easily be matched from the 3D pull-back data sets by offsetting OCT and IVUS images
at this distance. A finished miniature IVUS-OCT probe is shown in Figure 3-13. The
probe is more than 1.6 m long, with the maximum outer diameter of 0.65 mm, which
can be fit in a 3.6 Fr catheter sheath for in vivo animal study and clinical use.
48

Figure 3-12: Schematic (a) and photo (b) of miniature IVUS-OCT probe

Figure 3-13: A miniature IVUS-OCT probe

49

By making this probe into a catheter, we achieved many features that are
required for clinical use. A finished integrated IVUS-OCT catheter is shown in Figure
3-14. The catheter is 3.6 Fr in diameter, and effective length is more than 1.6 m. It has
an imaging window at the distal end of the catheter which is transparent to both
ultrasound and OCT. The imaging window is around 15~20 cm long. An X-ray opaque
marker for angiography and a slot for guide wire were located at the tip of catheter. The
catheter is able for flushing and pulling back. The flexible torque coil shaft ensured the
rotational torque can be smoothly translated through the whole catheter.

Figure 3-14: A finished integrated IVUS-OCT catheter and the imaging probe is inside
a 3.6 Fr sheath
3.5.2 In-vitro Imaging on Rabbit Aorta Specimens
To test the miniature catheter’s functionality, we conducted in vitro imaging on
animal specimens. The atherosclerotic plaque model was built from New Zealand white
rabbits. The rabbits were first undergone a balloon surgery, and then placed on a high-
50

cholesterol diet for over 4 months. Segments of the rabbit abdominal aortas were
harvested freshly for imaging. The imaged segments were pinpointed and fixed in
formalin after imaging for Hematoxylin-Eosin (H&E) stained histology examination.
One pair of IVUS and OCT images acquired from the integrated IVUS-OCT
catheter on a segment of rabbit aorta is shown in Figure 3-15. The histology result
[Figure 3-15 (c)] shows eccentric fibrous plaque on the left panel, while the intima on
the right panel is not significantly thickened. The IVUS image [Figure 3-15 (b)] clearly
distinguishes the dramatically thickened intima from the media layer on the left panel.
The entire plaque is identified in the IVUS image and the thickness of the plaque is
measurable. In OCT image [Figure 3-15 (a)], the intima on the right panel appears thin
and relatively moderate in brightness. On the left panel, OCT signal is strong and
becomes diffused inside the plaque, which demonstrates a typical characteristic of
fibrous plaque. Due to the superior resolution, the fibrous plaque in OCT image
displays two layers. The bright boarder (yellow arrow) of the fibrous plaque in OCT
image represents a fibrous plaque cap, which is also seen in histology [Figure 3-15 (c)]
as a very thin but darker boarder (black arrow) on the surface of plaque. However, this
feature is not detected in IVUS image due to the inferior resolution.
51

Figure 3-15: OCT (a), IVUS (b) and corresponding H&E histology (c) images of an
atherosclerotic rabbit aorta with eccentric plaque. I, intima; M, media; Fib, fibrous
plaque. The yellow and black arrows point to a plaque cap in OCT and histology
images, respectively

On a second pair of IVUS and OCT images, shown in Figure 3-16. Concentric
fibrous plaque is observed around the lumen in the histology [Figure 3-16 (c)]. The
fibrous plaque on the left panel is thicker than that on the right panel. In the OCT image
[Figure 3-16 (a)], the fibrous plaque appears thick on the left panel, and the plaque on
the right panel also displays strong and rich signals. A plaque cap is observed in OCT
52

image (yellow arrow) as bright and sharp boarder and so is in the histology image
(black arrow). The cap is not well identified in IVUS image [Figure 3-16 (b)]. However,
the fibrous plaque is imaged in full depth, and well distinguished from the media layer
in IVUS image.

Figure 3-16: OCT (a), IVUS (b) and corresponding H&E histology (c) images of an
atherosclerotic rabbit aorta with concentric plaque. The plaque on left panel is
dramatically thicker than that on the right panel. I, intima; M, media; Fib, fibrous
plaque. The yellow and black arrows point to a plaque cap in OCT and histology
images, respectively

53

Detection and differentiation of thrombus from plaque is critical for accurate
plaque characterization (Tearney, Jang et al. 2006). Two types of thrombus are
classified: platelet-rich (“white” thrombus) versus red-blood-cell-rich (“red” thrombus).
In this in vitro imaging experiment, rabbit aorta is excised with blood residue inside the
lumen on purpose. The blood residue clots attaching to the inner surface of vessel wall
is used to mimic thrombus. The thrombus is primarily composed of red blood cells,
which could be regarded as “red” thrombus.
Figure 3-17 shows images of a thick thrombus detected in IVUS and OCT
images. In OCT image [Figure 3-17 (a)], the “red” thrombus (T1, T2) has strong
reflection of light, and rapidly attenuates the light inside the thrombus, which is similar
as the situation of whole blood. The vessel (V) behind the thrombus is not clearly
identified. In most cases, only the surface of “red” thrombus could be imaged (T2) and
the vessel wall behind it is not detected.
In other cases, as shown in Figure 3-18 (a), where the thrombus is thin and
covering on the vessel surface (T1). The OCT signal is not totally attenuated, therefore
the vessel boarder (V) could be sharply distinguished from the thrombus. In IVUS
image [Figure 3-18 (b)], ultrasound could penetrate the thrombus to image the entire
vessel wall without significant attenuation, however, the thrombus appears bright and
rich even the thrombus is thin, which lacks of adequate contrast to be differentiated
from the vessel wall. If no prior knowledge of the thrombus or comparison with OCT
was known, the thrombus might be misread as fibrous plaque.
54

Figure 3-17: OCT (a) and IVUS (b) images of a thick thrombus. The thrombus (T1, T2)
rapidly attenuates OCT signals and blocks the view of vessel (V) in OCT image.
Ultrasound could penetrate the thrombus to image the vessel (V), but the boarder of the
vessel is unclear

Figure 3-18: OCT (a) and IVUS (b) images of a thin thrombus. The thrombus (T) signal
appears weaker than that from vessel (V) in OCT image, thus the boarder of vessel is
clearly identified. Ultrasound could penetrate the thrombus to image the vessel (V), but
the boarder of the vessel is unclear
55

3.5.3 In-vitro Imaging on Human Coronary Artery Specimens
Post mortem human coronary artery specimens were fixed in formalin and then
preserved in phosphate buffer for imaging. During the in vitro experiment, the
integrated IVUS-OCT catheter was inserted into the lumen of specimen which was
immersed in saline. Imaging pull-back was conducted.
OCT and IVUS images of a human coronary artery specimen with calcified
plaque is shown in Figure 3-19. The plaque can be identified in both images. In the
OCT image, the plaque cap, which is bright and with sharp boarder, can be clearly
identified due to the high axial resolution and high contrast of OCT. However, the OCT
image could not visualize the entire depth of the vessel wall, and its maximum
penetration depth was about 1 mm. On the other hand, in the IVUS image, the acoustic
shadow as pointed by the arrow indicates the location of the plaque. However, one can
hardly distinguish the border between calcium and the surrounding tissue due to the
inferior resolution of the ultrasound image, but its penetration depth was much deeper
than that of OCT (the radius of the image is 4.5 mm). The contour of the vessel in the
OCT image matches very well with that in the IVUS image, as shown in Figure 3-19 (c)
which indicates that the two images were taken at approximately the same cross section.
56

Figure 3-19: OCT (a), ultrasound (b), and combined IVUS-OCT (c) images of a human
coronary artery specimen
3.5.4 In-vivo Imaging on Rabbit
For the in vivo experiment, atherosclerotic plaque model was built from the
New Zealand white rabbits. The rabbits were undergone a balloon surgery, then placed
on a high-cholesterol diet for over 4 months.
57

In the in vivo experiment, the rabbit was anesthetized and incubated during the
surgical and imaging procedures. The rabbit was cut open chest, then the imaging
catheter was inserted from thoracic aorta down to the abdominal aorta. Imaging
pullback and saline flushing for blood clearance was performed. The integrated IVUS-
OCT catheter was spinning within a 3.6 Fr catheter sheath (Boston Scientific Corp.,
Natick, MA) to avoid contamination and causing trauma to the vessel wall.

Figure 3-20: A photograph of in vivo rabbit experiment
The blood is highly scattering for light and rapidly attenuate the OCT signals,
thus no vessel image is obtained by OCT at the presence of blood [Figure 3-21 (a)]. On
the other hand, since blood serves as a nature coupling medium for ultrasound, the
vessel structure and tissue surrounding outside the vessel is identified clearly in IVUS
image [Figure 3-21 (b)]. After flushing 10cc saline to clear blood, OCT displayed a
clear view of the aorta [Figure 3-21 (b)]. At the same time, since blood was cleared, the
blood backscattered signals in IVUS image was significantly reduced and a clear view
of aorta lumen was obtained. At 5 o’clock, a fibrous plaque with lipid content (L) was
detected in both OCT and IVUS images. Due to the higher resolution, the plaque cap
(FC) was defined more explicitly.
58

Figure 3-21: OCT (a) and IVUS (b) images of a rabbit abdominal aorta without
flushing. OCT (c), IVUS (d) and combined (e) IVUS-OCT images with flushing. L,
lipid-rich content; FC, fibrous cap
3.5.5 In-vivo Imaging on Swine
For testing the feasibility of the integrated IVUS-OCT probe and system, in vivo
rabbit experiments have been successfully conducted. In rabbit models, only the
thoracic or abdominal aortas are achievable for imaging. The rabbit aortas are similar in
size of human coronary artery and can develop multiple degrees of plaques. However,
59

the access to the straight rabbit abdominal aorta for a catheter is much easier than the
access to a human coronary artery; hence the rabbit experiment cannot fully simulate
the real situation where coronary artery imaging is conducted. In clinic, when the IVUS
or intravascular OCT imaging is performed, the catheter is inserted from right or left
femoral artery and all the way up through abdominal aorta to the aortic arc, where the
catheter is turned 180˚. After passing the aortic arc, the catheter is guided into coronary
arteries and further into any branch of interest. The long passage requires the catheter to
be at least 135 cm in length and flexible for turning 180˚ and rotating, while also
maintaining certain stiffness to avoid kinking.
A swine is a better model for testing our probes, for the reasons that porcine
body size is close to human; heart and coronary artery structure is similar to human;
and physiological functions are suitable for monitoring the surgery procedures. During
an in vivo swine experiment, a 50 kg healthy pig was used for imaging. During the
experiment, the pig was anesthetized and incubated. Right femoral artery was opened
for inserting a 7 Fr guide catheter. The guide catheter was advanced through the
femoral artery, abdominal aorta, and aortic arc into the opening of one of coronary
arteries. After placing the guide catheter in position, a guide wire was inserted through
the 7 Fr guide catheter into the coronary or its branches. The imaging catheter was then
slid through the rail of guide wire into the coronary artery. After using the X-ray
angiography to confirm the imaging catheter was in proper position, the imaging
process began. During the imaging, saline flushing and imaging pull-back was
performed. The imaging was conducted at 20 fps and pull-back speed was 1 mm/s. 10
60

cc saline was used at each flushing. A photograph of the in vivo experiment on swine is
shown in Figure 3-22. The in vivo images acquired by the integrated IVUS-OCT
catheter are shown in Figure 3-23. The guide wire is seen in both images as a bright
spot with shadow in OCT image and ring-down in IVUS image. The coronary artery is
relatively thin in the healthy pig. The OCT image delineated the vessel with very high
resolution and is able to distinguish the thin vessel and other tissues. In IVUS image,
though the imaging depth is deeper than that in OCT image, the vessel is hardly
differentiated from the tissue surrounding outside. The overall experiment was
successful and demonstrated the feasibility and capability of the integrated IVUS-OCT
imaging modality.

Figure 3-22: in vivo experiment on swine
61

Figure 3-23: OCT (a) and IVUS (b) images of a swine coronary artery after flushing. G,
guide wire; V, vessel; T, tissue surrounding outside the vessel
3.6 Discussion
Since 2009, we have been researching on the integrated IVUS-OCT catheters
and systems for intravascular imaging. Three generation of probes have been designed,
fabricated and tested in vitro and in vivo. The trend of catheter evolution is getting
smaller but more durable. Many engineering problems have been solved to achieve the
performance of real time imaging of dual modalities in the highly interventional
environment. Through all the in vitro and in vivo tests, we successfully demonstrated
the advantages of the combination of two most commonly used imaging modalities.
With the complementary information from the two modalities, extra knowledge of the
plaques could be distracted for the evaluation of plaque vulnerability. Yet the extra
knowledge could not be obtained from either modality alone. The remarkable
62

achievement of in vivo experiment in swine has pushed the technology one big step
toward the translation of clinical use.
However, there is still a long way to go before this technology can be applied in
human use. More animal tests are needed for evaluating the safety of the catheter using
in clinical environments. The image quality of each single modality needs to be
enhanced to achieve the current state of the art. The imaging speed still has big room to
improve and 100 fps will be the ideal speed. Flushing techniques need to be optimized
since the two modalities are now working simultaneously and both are under the
influence of flushing. Image fusing and analysis methods need to be developed, so as to
maximally utilize the complementary information from the two modalities.
After solving these and other issues that may pop up, the dual-function modality
could make a great change to the diagnosis of atherosclerosis.

63

Chapter 4 Intravascular Ultrasound at 80 and 95 MHz
4.1 Introduction to Very High Frequency IVUS
Currently, IVUS and OCT are the most commonly used coronary artery
imaging modalities. IVUS at 40 and 45 MHz has resolution on the order of 100 µ m and
penetration around 5~8 mm. OCT, on the other hand, can provide much better
resolution of 10~20 µm, but could only achieve the penetration depth around 1.5 mm.
In the previous chapter, an integrated IVUS-OCT probe was introduced, which could
provide high resolution for superficial micro-structures and moderate resolution for
deep tissues. However, there is still a gap between conventional IVUS and OCT in
resolution and penetration for tissue beyond OCT’s field of view. Figure 4-1 illustrates
the comparison of resolution versus penetration of conventional IVUS (40 MHz), OCT,
and very high frequency IVUS (above 80 MHz). Increasing the center frequency to 80
MHz or higher can improve imaging resolution and fill the gap between conventional
IVUS and OCT, but at the cost of losing penetration depth. At 80 MHz, an attenuation
coefficient of 10 dB/mm is expected for coronary artery, which means that a
penetration depth of 3 mm can be achieved for a system with a dynamic range of 60 dB
(Foster, Pavlin et al. 2000).
Building such high frequency transducers is challenging because the preparation
of very thin piezoelectric layers with properties similar to the bulk material is difficult.
In Chapter 2.3.2, the benefits of using piezoelectric film technology for fabricating very
64

high frequency ultrasonic transducers was discussed. In this chapter, we introduce the
utilization of [Pb(Mg
1/3
Nb
2/3
)O
3
]
0.63
[PbTiO
3
]
0.37
(PMN-PT) free-standing film for very
high frequency IVUS application.

Figure 4-1: Illustration of resolution versus penetration depth of conventional IVUS (40
MHz), OCT, and very high frequency IVUS (80~100 MHz)
4.2 IVUS at 80 MHz
4.2.1 PMN-PT Free-Standing Film Fabrication
The PMN-PT free-standing film was synthesized using a modified precursor
coating approach (Luo, Shih et al. 2007). Nb
2
O
5
(99.99%, Aldrich Chemical Co.,
Milwaukee, WI), titanium isopropoxide (Ti(OCH(CH
3
)
2
)
4
, 99.9% Alfa Aesar, Ward
Hill, MA), lead acetate anhydrous (Pb(CH
3
COO)
2
.2Pb(OH)
2
, Fluka, St. Louis, MO),
65

Mg(Ac)
2
.6H
2
O (99.9%, Alfa Aesar, Ward Hill, MA), and NH
4
OH (5M, Aldrich
Chemical Co., Milwaukee, WI) were used in this fabrication.
0.1 mol of Nb
2
O
5
powder was first suspended in 500 ml distilled water then
ultrasonicated (Ultrasonic Homogenizer 4710 series, Cole-Parmer Instrument Co.,
Vernon Hills, IL) for 10 min to break up the Nb
2
O
5
agglomerates. To precipitate
Mg(OH)
2
on to the Nb
2
O
5
surface, Mg(CH
3
COO)
2
.6H
2
O (0.105 mol) was continuously
dropped into the suspension during this coating process. NH
4
OH (5M) was used to
keep pH level above 10.5 and the suspension was kept stirring for 30 min after mixing.
The suspension was dried at 150˚C by a hotplate. After drying, a precursor slurry was
made from a lead acetate anhydrous (Pb(CH
3
COO)
2
.2Pb(OH)
2
) solution in ethylene
glycol (EG) (HOCH
2
CH
2
OH, Alfa Aesar, Ward Hill, MA) with 15% excess lead. The
suspension was then dried at 230˚C on a hot plate. Pyrochlore-free perovskite PMN
powder was obtained by the dried PMN precursor powder, which was first heated at a
rate of 1˚C/min to 360˚C for 2 hrs, followed by 5˚C/min heating to 950˚C for 2 hrs.
A PT precursor solution in EG with a stoichiometric quantity of
Pb(CH
3
COO)
2
.2Pb(OH)
2
and Ti(OCH(CH
3
)
2
)
4
was prepared by dissolving in EG
before mixing. The perovskite PMN powder was then suspended in a PT precursor
solution containing lead acetate and titanium isopropoxide (Ti(OCH(CH
3
)
2
)
4
, 99.9%
Alfa Aesar, Ward Hill, MA) in EG and ball-milled for 48 hrs. PMN-PT green powder
was obtained by drying at 230˚C (with constant stirring) followed by heat treatment at a
rate of 1˚C/min to 360˚C for 2 hrs. After drying, the resultant powder was ball-milled
again for 24 hrs to ensure intimate mixing of the PMN powder with the PT precursor,
66

which was ready for tape casting. The sizes of the PMN and PT precursor particles
were 430 and 44 nm, respectively, by BET (Luo, Shih et al. 2007). A scanning electron
microscopy (SEM) micrograph of the green tape is shown in Figure 4-2 (a), where the
submicron-sized PMN is intimately mixed with the nano-sized PT precursor. The tape
casting process was carried out by a tape casting company (Maryland Tape Casting Co.,
Bel Air, MD). The solids loading and viscosity of the slurry had been optimized by
adding binders and dispersants. The slurry was ball-milled for 48 hrs and vacuumed to
remove bubbles before casting. The solids loading of the slurry was 60.74% by weight.
The thickness of the tape is controlled by adjusting the gap of the doctor’s blade. The
resultant tape was smooth without warping or cracking, shown in Figure 4-2 (b). The
PMN-PT green tape was sintered at 1200˚C for 2 hrs and under saturated PbO vapor
atmosphere at normal pressure to prevent the evaporation of Pb element from PMN-PT
tape.

Figure 4-2: An SEM image of the top surface of a PMN-PT green tape (a); An optical
image of a PMN-PT green tape (tan color) on a black bench top (b)
67

4.2.2 PMN-PT Free-Standing Film Characterization
The crystalline phases of the sintered PMN-PT free-standing film were
characterized by X-ray diffractometry (XRD), shown in Figure 4-3. The XRD pattern
displayed a pure perovskite phase, indicating the film was well crystallized. The
crystalline microstructures of the film were observed by scanning electron microscopy
(SEM), shown in Figure 4-4. The film was fully dense and without any crack or pore.
Density was measured to be 7.76 g/cm
3
. d
33
was measured to be 478 pC/N. Grain size
was around 1-3 µ m and film thickness was 30 µ m. According to (Lee, Zhang et al.
2010), fine grained PMN-PT films would offer greater mechanical strength and
improved property stability, compared to coarse-grained (7-10 µ m) PMN-PT ceramics.

Figure 4-3: X-ray diffraction (XRD) pattern of the PMN-PT free-standing film
68

Figure 4-4: SEM micrographs of the 30 µ m PMN-PT free-standing film: a cross-section
view (a) and an enlarged view (b).

For the measurement of electrical properties, chrome/gold (Cr/Au) electrodes
with dimension of 0.4× 0.4 mm
2
were sputtered onto the film as top electrodes. Another
layer of Cr/Au was sputtered onto the bottom to serve as ground electrodes. Film
samples were poled in DC electric field of 30 KV/cm for 5 minutes at room
temperature before measurements. The dielectric and ferroelectric properties were
measured with an Agilent 4292A impedance analyzer (Agilent Technologies, Santa
Clara, CA) and an RT6000 ferroelectric test system (Radiant Technology, Albuquerque,
NM), respectively. The frequency dependence of the free relative dielectric permittivity
( ε
r
/ε
0
) and the loss (tan ) of the PMN-PT film were measured from 1 KHz to 1 MHz,
shown in Figure 4-5. The ε
r
/ε
0
and tan  at 1 KHz were found to be 4,364 (± 221) and
0.033 (± 0.004), respectively. The high relative permittivity and low loss of the PMN-
PT free-standing film implied a better electrical impedance match and improved
sensitivity for miniature high frequency transducers. The polarization-electric field
hysteresis loop was shown in Figure 4-6. The remnant polarization (P
r
) and coercive
69

field (E
c
) were 28 μC/cm
2
(bulk, 12.3-33.1 μC/cm
2
)

and 18.43 KV/cm (bulk, 3.3-4.3
KV/cm) (Kosec, Holc et al. 2007), respectively. The saturation polarization (P
s
) was
around 46 μC/cm
2
.

Figure 4-5: Frequency dependence of dielectric constant and loss of PMN-PT free-
standing film.

Figure 4-6: Polarization-electric field hysteresis loop of the annealed PMN-PT free-
standing film
10
3
10
4
10
5
10
6
0
1250
2500
3750
5000
Frequency (Hz)
Relative Dielectric Constant

10
3
10
4
10
5
10
6
0
0.125
0.25
0.375
0.5
Loss
Relative Dielectric Constant
Loss
-150 -100 -50 0 50 100 150
-50
-25
0
25
50
Electric Field (KV/cm)
Polarization ( C/cm
2
)
70

The frequency dependence of the electrical impedance and phase are displayed
in Figure 4-7, which showed that the electrical impedance at resonant peak is 29 Ω at
75 MHz, the series and parallel resonant frequencies are 69 MHz and 80 MHz. k
t
was
found to be 0.55 which was comparable to the bulk PMN-PT single crystal (0.58, HC
materials, Bolingbrook, IL, USA).

Figure 4-7: Frequency dependence of electrical impedance of the 30 µ m PMN-PT free-
standing film
4.2.3 Transducer Design and Fabrication
After characterization of the PMN-PT free-standing film, a piece of 7 mm × 10
mm film was used as the active piezoelectric layer to fabricate side-viewing miniature
transducers. Two layers of acoustic matching scheme was applied to improve the
performance of transducer. Simulations on a Krimhotz, Leedom, and Matthaei (KLM)
equivalent circuit model (PiezoCad, Sonic Concepts, Woodinville, WA) predicted that
60% bandwidth at center frequency of 81 MHz could be achieved by incorporating
40 50 60 70 80 90 100
20
25
30
35
40
Frequency (MHz)
Z (Ohms)

40 50 60 70 80 90 100
-90
-72
-54
-36
-18
Phase (Degrees)
Z
Phase
71

4.4 µ m silver epoxy (7.3 MRayl) and 1.8 µ m parylene (2.3 MRayl) as first and second
matching layers. However, even a 1 µ m variance of the two layers would decrease the
bandwidth by 20% or more. This made quality control in the fabrication process a
great challenge. Figure 4-8 shows an acoustic stack where the PMN-PT free-standing
film was deposited with silver epoxy matching layer and conductive backing layer. The
IVUS transducer fabrication process is similar to that described in Chapter 2.4.1.

Figure 4-8: SEM photograph of an acoustic stack that is composed of acoustic matching
layer, PMN-PT film layer, and acoustic backing layer
4.2.4 Free-Standing Film IVUS Transducer Characterization
Pulse-echo result of a representative transducer, whose matching layers well
matched the designed parameters, was shown in Figure 4-9. Measured center frequency
was 82 MHz and -6dB fractional bandwidth was 65%. However, due to the difficulty in
controlling the thickness of silver epoxy matching layer, transducers with thinner layers
(2~3 µ m) possessed inferior bandwidth (35%) but slightly higher center frequencies
(83~84 MHz), whereas transducers with thicker layers (5~7 µ m) functioned at lower
72

center frequencies (70~75 MHz) and had fair bandwidths (35~45%). The decrease in
center frequency could be explained by the clamping and attenuating effects of the
matching layer.

Figure 4-9: Pulse-echo measurement of one representative 80 MHz PMN-PT free-
standing film transducer

Two-way insertion loss (IL) was measured to be 23 dB at 80 MHz, which
indicated the transducer’s sensitivity is comparable with that of an 80 MHz large
aperture LiNbO
3
single crystal transducer (10-25 dB) (Cannata, Ritter et al. 2003). 6-
μm-diameter tungsten wire imaging is shown in Figure 4-10 (a). The envelopes of echo
signals from the wire located at 1.2 mm away from the transducer surface, are
displayed in Figure 4-10(b) and (c). The axial and lateral resolutions were determined
from the -6 dB envelope widths, which were 35 μm and 176 μm, respectively.
1.2645 1.3898 1.515 1.6403 1.7655
-400
-200
0
200
400
Time ( s)
Amplitude (mV)

40 60 80 100 120
-24
-18
-12
-6
0
Frequency (MHz)
Magnitude (dB)
Pulse-echo Responsee
Spectrum
73

Figure 4-10: Ultrasound wire phantom image (a), displayed with a dynamic range of 45
dB; axial (b) and lateral (c) envelopes of echo signals from the wire located at 1.2 mm
away from the transducer surface

The performance results were summarized in Table 4-1.
Table 4-1: 80 MHz IVUS transducer testing results
Specifications Values
Measurements of 80 MHz transducer
Center frequency 82 MHz
Bandwidth 65%
Insertion loss 23 dB
Axial resolution (-6dB) 35μm
Lateral resolution (-6dB) 176 μm
4.2.5 80 MHz IVUS Imaging Results
In vitro imaging of a human cadaver coronary artery specimen was performed
by using the 80 MHz PMN-PT free-standing film transducer, as shown in Figure 4-11
(a). For comparison, the same cross section of the specimen was imaged with a 40 MHz
Lateral distance [mm]
Axial distance [mm]
6 um Wire Phantom
1 2 3 4 5
1.2
1.4
1.6
1.8
1 1.2 1.4
-50
0
Axial distance (Depth) [mm]
0 0.5 1 1.5 2 2.5
-50
0
Relative Magnitude [dB]
Lateral distance [mm]
(b)
(c)
(a)
74

PMN-PT single crystal transducer, as shown in Figure 4-11 (b). All in vitro
experiments were conducted in de-gassed water.

Figure 4-11: Images of human cadaver coronary artery specimen from 80 MHz PMN-
PT freestanding-film transducer (a); and 40 MHz PMN-PT single-crystal transducer (b).
Fib, fibrous plaque; Ca, calcified plaque; L, lipid-rich plaque;

The images at both frequencies are able to identify the three-layer structures,
fibrous (Fib), lipid-rich (L), and calcified (Ca) plaques. Compared to the 40 MHz
images, the 80 MHz image displays superior resolution with denser speckles. A thin
fibrous cap covering on a lipid-rich area is detected at 11 o’clock, as marked by an
arrow. Detailed plaque structure, especially the plaque cap thickness, which is around
84 µ m, is measureable in 80 MHz image, but not in 40 MHz image. The hypoechoic
region behind the cap, which represents lipid-rich area, is not identified in the 40 MHz
image due to the inferior resolution and relatively higher backscattering property than
that at 80 MHz. At 40 MHz, the ring down of echo signal from the cap may extend to
the lipid region, thus making the originally signal-poor region appear as brighter.
75

Figure 4-12: Images of human cadaver coronary artery specimen from 80 MHz PMN-
PT freestanding-film transducer (a); and 40 MHz PMN-PT single-crystal transducer (b).
Three layer structures are displayed with higher contrast in the 80 MHz image than the
40 MHz one. I, intima; M, media; A, Adventitia

In a second pair of images of coronary artery, as shown in Figure 4-12, the
layered structures are distinguished in the 80 MHz image with higher contrast than that
in the 40 MHz image. The improvement of contrast is resulting from the shorter pulse
76

width at higher frequency. The penetration depth observed in the 80 MHz images is
around 2 mm. At this frequency, ultrasound is still able to see through the entire arterial
wall into the tissue surrounding outside the vessel.
4.3 IVUS at 95 MHz
4.3.1 95 MHz IVUS Transducer Using 22 µ m PMN-PT Free-Standing Film
Making a step forward, we continue researched on 95 MHz IVUS transducer
with a 22 µm PMN-PT free-standing film, as shown in Figure 4-13. At this time, the
silver epoxy matching layer was not applied because the thickness was required to be
on the level of 1 to 2 µ m, which was impossible to achieve by the deposit-and-lapping
method. In order to match the transducer’s electrical impedance to 50 Ω, the acoustic
stack was designed to be 0.3 × 0.3 mm
2
. The acoustic stack after depositing conductive
backing layer is shown in Figure 4-14. The frequency dependence of the electrical
impedance and phase are displayed in Figure 4-15, which illustrates that the electrical
impedance at resonant peak is 34 Ω at 100 MHz, and the series and parallel resonant
frequencies are 94 MHz and 104 MHz. k
t
was found to be 0.46.

77

Figure 4-13: SEM micrographs of a 22 µ m PMN-PT free-standing film

Figure 4-14: SEM photograph of an acoustic stack that is composed of 22 µ m PMN-PT
free-standing film and conductive backing layer

Figure 4-15: Frequency dependence of electrical impedance of the 22 µ m PMN-PT
free-standing film
80 85 90 95 100 105 110
27
29.5
32
34.5
37
Frequency (MHz)
Z ( )

80 85 90 95 100 105 110
-80
-72.5
-65
-57.5
-50
Theta ( )
Z
Theta
78

A 4 µ m parylene layer was deposited after finishing the transducer as the only
matching layer. Figure 4-16 shows the pulse echo testing result of a 22 µ m PMN-PT
free-standing film transducer. The center frequency is 95 MHz, and the bandwidth is
only 35%, which is due to lacking of the silver epoxy matching layer. This transducer
can achieve an axial resolution of 22 µm.

Figure 4-16: Pulse-echo measurement of one representative 95 MHz PMN-PT free-
standing film transducer
4.3.2 95 MHz IVUS Imaging Results
The 95 MHz PMN-PT free-standing film transducer has been applied for in
vitro IVUS imaging to evaluate its performance. Human cadaver specimens were
imaged in de-gassed water. The 95 MHz IVUS images were compared with images of
lower frequency IVUS and OCT, which were acquired at the same cross section of
1.2645 1.3898 1.515 1.6403 1.7655
-1000
-500
0
500
1000
Time ( s)
Amplitude (mV)

60 75 90 105 120
-24
-18
-12
-6
0
Frequency (MHz)
Magnitude (dB)
Pulse-echo Response
Spectrum
79

specimens for each comparison. The co-registration of each comparison was achieved
by positioning the probes at the same depth into a segment of aorta, referencing from
the top of the segment of aorta. The position of transducer was precisely controlled by a
positioning stage.

Figure 4-17: Images of a human cadaver coronary artery specimen from a 95 MHz
PMN-PT free-standing-film transducer (a) and a 40 MHz PMN-PT single-crystal
transducer (b). And the corresponding histology by Von Kossa staining. ca, calcified
plaque; A fibrous plaque cap was pointed out by an arrow. The calcium appears black
in Von Kossa staining. The blank area at 9 o’clock in (c) is due to the dropping of
calcium in staining process. ca, calcified plaque; cap, fibrous cap

80

Figure 4-17 shows images of a cross section of human cadaver coronary artery
specimen from a 95 MHz PMN-PT free-standing-film transducer Figure 4-17 (a) and a
40 MHz PMN-PT single-crystal transducer Figure 4-17 (b). The corresponding segment
histology processed by Von Kossa staining is shown in Figure 4-17 (c). This staining
method is for displaying the presence of calcium content, which appears as black. In the
histology image, the calcium contents are identified in two locations: 5 o’clock and 9
o’clock. In both 95 and 40 MHz IVUS images, the two calcium contents are identified
as a bright boarder followed by acoustic shadow region. Compared to the 40 MHz
IVUS image, the boarder of fibrous cap covering the calcium at 9 o’clock in 95 MHz
IVUS image, as marked by an arrow, is delineated sharper due to the higher resolution.

Figure 4-18: Images of human cadaver coronary artery specimen from a 95 MHz PMN-
PT free-standing-film transducer (a) and a 80 MHz PMN-PT free-standing-film
transducer (b)
The comparison of 95 MHz IVUS image with 80 MHz IVUS image is shown in
Figure 4-18. The speckle size in 95 MHz image is slightly denser than that in 80 MHz
one. Both images demonstrate clear views of the thickened intima, media and adventitia
81

layers. The penetration depth at 95 MHz is similar to that at 80 MHz, but slightly
shallower, which can be demonstrated at 9 o’clock. A signal poor region, which is
corresponding to lipid-rich area, is identified at 9 o’clock in both images. The 80 MHz
image is able to see through the lipid pool to the rear boarder, but the rear boarder is not
detected in the 95 MHz image.
The comparison of 95 MHz IVUS with OCT image is shown in Figure 4-19. On
the right hand panel of Figure 4-19 (a), the IVUS image at 95 MHz could still see
through all the three layers even the intima layer is significant thickened. A calcified
plaque is seen on the left hand panel in both 95 MHz IVUS and OCT images, so does a
lipid-rich area At 7 o’clock. The OCT image delineates the lipid plaque cap with higher
contrast and resolution than that appears in IVUS image.

Figure 4-19: Images of human cadaver coronary artery specimen from a 95 MHz PMN-
PT free-standing-film transducer (a) and an OCT probe (b). ca, calcified plaque; L,
lipid-rich plaque; cap, fibrous cap
The comparison of 40 MHz IVUS, 95 MHz IVUS, OCT and corresponding
H&E histology images on a human coronary artery with fibrous plaque are shown in
82

Figure 4-20. At 12 o’clock, a fibrous plaque (Fib) is detected in 40 MHz IVUS, 95
MHz IVUS, and OCT images. The media and adventitia layers behind the plaque are
detected in 40 MHz IVUS and 95 MHz IVUS images but not in OCT image. The tissue
surrounding outside adventitia is detected in 40 MHz IVUS image, but not other two
modalities. At 3 o’clock, all the three layer structures are identified in 40 and 95 MHz
IVUS images, while only the inner two layers are detected in OCT image.

Figure 4-20: Images of a human cadaver coronary artery specimen from a 40 MHz
PMN-PT single-crystal transducer (a); a 95 MHz PMN-PT free-standing-film
transducer (b); an OCT probe (c); and H&E histology (d). Fib, fibrous plaque; I, intima;
M, media; A, Adventitia
83

In this comparison, where only fibrous plaque is present in the specimen, 40
MHz IVUS is able to detect the important features, such as the fibrous plaque and
three-layer structures, however, 95 MHz IVUS and OCT does make it easier to identify
those features at higher resolution.
4.4 Discussion
Driving the IVUS transducer’s working frequency to 80 MHz and 95 MHz is
breathtaking but also challenging. The remarkable resolution of IVUS improved at 80
and 95 MHz achieves 20 to 30 µ m in axial direction. By comparing to the current 40
MHz IVUS technology, we demonstrated that IVUS above 80 MHz were able to detect
some features that is beyond the scope of 40 MHz, such as thin fibrous cap, small
region of lipid content, and boarder of calcium. However, compared with OCT, the
resolution of IVUS at 95 MHz is still inferior to that of OCT. Yet 95 MHz IVUS is
sufficient in detect certain thin caps in TCFA in some cases, and its penetration is still
deeper than OCT.
The penetration depth at such high frequencies is another issue needs carefully
examination. The deepest penetration at 80 MHz ever observed is 2 mm into vessel
wall. IVUS at 95 MHz is able to see through all the three layers for healthy aorta or
vessels whose intima is not significantly thickened. However, all those conclusions are
derived in the testing condition without the presence of blood. In the in vivo situations,
red blood cells are very strong ultrasound scatters, which would scatter portion of
84

ultrasound energy and cause strong backscattered signals. The scattering of ultrasound
would lead to significant attenuation of ultrasound signal strength before ultrasound
waves are able to arrive at the vessel wall. At the meantime, backscattered signals
would lead to strong blood signals that appear very bright in the image, making the
boarder of vessel wall hard to identify. Preliminarily studies on a human cadaver
specimen in blood mimicking solution showed it was impossible for 80 MHz
ultrasound waves to penetrate the solution without significant reduce in signal
amplitude. And with the blood, the penetration of 80 MHz IVUS image is less than 1
mm into vessel wall. This finding is consistent with (Foster, Obara et al. 1994), which
shows the backscatter coefficient of blood is higher than soft plaque at 65 MHz and
higher. The blood backscatter coefficient is even higher at body temperature than room
temperature, and it does not drop when shear rate increases. Considering all those
limitations, the only solution for IVUS at 80 MHz or above might be saline flushing at
in vivo or clinical environments.
Looking back at the 40 MHz IVUS, though the resolution is inferior to 80 MHz
IVUS or OCT, the decent penetration depth makes it irreplaceable. And in certain cases,
where only fibrous plaque is present, the 40 MHz IVUS is still workable. However, for
some critical features that determine the vulnerability of a plaque, such as TCFA, 80 or
95 MHz IVUS and OCT are the only modalities that have the potential to detect.
Considering the pros and cons of IVUS at lower and higher frequencies, an optimal
method might be integrating the conventional 40 MHz IVUS with a higher frequency
transducer to provide IVUS images on different resolution and penetration depth at the
85

same time. From a technical point of view, this integration is achievable at a moderate
cost.
In view of future improvement on the very high frequency IVUS, there is still
room to improve the performance of transducer, especially for the bandwidth and
sensitivity. The potential methods include optimizing the PMN-PT free-standing film
synthesis; improving matching layer depositing techniques; and optimizing the
transducer ground shielding. On the other hand, better electrical system is needed for
enhance the performance. The system used for very high frequency IVUS in this
research is from commercial pulser and receiver, which are not optimized for very high
frequency imaging application. The system noise level could be reduced by custom
designing an integrated transmitting and receiving system. And the system should have
the flexibility in generating arbitrary waveform pulses in order to maximizing the
transmitting energy, such as coded excitation.
The attenuation and reflection of IVUS catheter sheath needs to be evaluated
and optimized. In previous chapters, the 95 MHz IVUS images were acquired while the
transducer was inside a polyolefin tube, whose acoustic impedance is closed to that of
commercially used low density polyethylene tube. Strong reflection was observed when
the transducer is normal to the tube. By tilting the transducer to a certain angle, the
reflection could be reduced a bit amount. However, the tilting angle was hardly
controlled in current transducer fabricating procedures. Optimizing the sheath material
and the tilting angle could reduce the loss of acoustic energy, hence improve the
penetration depth.
86

Chapter 5 Intravascular Ultrasound Combined with
Photoacoustic Imaging
5.1 Introduction to Intravascular Photoacoustic Imaging
For Intravascular Photoacoustic (IVPA) application, usually a short duration
unfocused light pulse is used. The image axial resolution is determined by the
ultrasonic transducer’s working frequency. Higher frequency transducer can provide
better image resolution. Acoustic signal strength or image contrast is determined by the
light absorption coefficient (μ
a
) of tissue (Ku, Wang et al. 2004).
Human and animal arterial walls normally display three-layer structures: intima,
media and adventitia (Fitzgerald, St Goar et al. 1992; Sarkola, Redington et al. 2010).
Intimal thickening is considered to be an early and continuous manifestation of
atherosclerosis (Kume, Akasaka et al. 2005; Puri, Worthley et al. 2011). Examining the
optical absorption spectra of the arterial wall at 532 nm shows that the intima and
adventitia have similar μ
a
values (10.04 and 13.29), but almost twice higher than that of
media (5.323) (Keijzer, Richards-Kortum et al. 1989). On this basis, we hypothesize
that the three-layer structures could be better resolved by IVPA imaging, with optical
contrast and ultrasonic resolution. Intima and adventitia would appear brighter than
media in IVPA image. The optical contrast between layered structures is believed to be
more valid than that induced by acoustic impedance, which is only a few percent for
uncalcified tissues (Beard and Mills 1997).
87

Previous investigators demonstrated the capability of spectroscopic IVPA
imaging in detecting lipid (Wang, Su et al. 2009; Jansen, Steen et al. 2011). However,
the application of IVPA to the assessment of arterial wall layered structures has not
been investigated. Moreover, while earlier studies were mostly conducted under 40
MHz range, IVPA at 80 MHz with higher resolution could be a more effective
approach in detecting the layered structures.
5.2 IVPA System Setup
The ultrasound based integrated IVUS/IVPA imaging system is shown in Figure
5-1. A pulsed Q-switched Nd:YAG laser (532 nm, 3-5 ns pulse width, 10 Hz repetition
rate, Continuum, Inc., Santa Clare, CA) is used as IVPA excitation source. The free
space laser output is coupled by a 4 × objective lens into an optical fiber then delivered
to the prism at distal end of the probe. The integrated probe is inserted into the
specimen’s lumen and internally illuminates the specimen with surface fluence energy
around 10 mJ/cm
2
. The circumferential scanning is achieved by rotating a water tank
with specimen inside using a stepper motor, while the probe is kept immobile. A
PR5900 pulser/receiver (Olympus NDT, Inc., Kennewick, WA) is used to generate
ultrasound pulses and receiving both ultrasound and IVPA waves. Received signals are
digitized and processed in computer. For each IVPA scan, 1000 A-lines are acquired
and averaged by every two A-lines to reduce the background noise. IVUS scan is then
followed with the same scanning pattern at the same cross section. Co-registration
88

between IVUS and IVPA is ensured by the precision of motor stepping. The scanning
procedure is controlled by a custom built LabVIEW program (National Instruments,
Austin, TX) and synchronized by the laser trigger signals.

Figure 5-1: Schematic of the hybrid IVUS and IVPA imaging system.
5.3 Probe Design
5.3.1 Side-by-Side Arrangement
The integrated IVUS/IVPA probe is composed of parallel arranged side-firing
optical probe and side-viewing ultrasonic transducer, shown in Figure 5-2. For the
optical part, a 200-µm-core multimode fiber is used to deliver 532 nm pulsed laser
beams. At the distal end, a 45˚ polished microprism (0.25× 0.25× 2 mm
3
Bern Optics
89

Inc., Westfield, MA) is connected to the fiber tip and sealed inside a glass capillary
tube (0.4 mm ID; 0.55 mm OD). Air is trapped inside the tube to form air/glass
interface at the prism polished surface to redirect laser beams by 90˚, following “total
internal reflection” effects. For the acoustic part, 35 and 80 MHz ultrasonic transducers
are fabricated in our laboratory (0.4× 0.4 mm
2
aperture) (Li, Wu et al. 2011) and
assembled into IVUS/IVPA probes, respectively. The bandwidth of the 35 MHz
transducer is 50% and that of the 80 MHz transducer is 45%, which are determined
from ultrasound pulse-echo tests. The optical prism and ultrasonic transducer are
arranged side by side with an angle of approximately 160˚ and packaged in a Polyimide
Tubing (0.0453” ID, 0.0473” OD, Small Parts, Inc., Miramar, FL) with outer diameter
of 1.2 mm.

Figure 5-2: Schematic of the integrated IVUS/IVPA probe: top view (a) and front view
(b). Light beam is in green and acoustic beam in gray
5.3.2 Coaxial Arrangement
Normally, the probe is designed by mounting the polished fiber tip and
ultrasound transducer either parallel or sequentially (Karpiouk, Wang et al. 2010;
90

Jansen, van der Steen et al. 2011). Based on this geometry, the overlap portion of laser
beam and ultrasound beam is determined by the face angle of the polished fiber and the
distance between the fiber tip and the transducer. In this case, the working distance of
the probe is finite (e.g., 0.5 to 4.5 mm) (Jansen, van der Steen et al. 2011). Additionally,
in order to get highly efficient acoustic signal, the multiple components need to be
precisely aligned. In this chapter, a novel coaxial integrated probe design is presented.
Figure 5-3 (a) shows the schematic of the coaxially arranged integrated
IVUS/IVPA probe (Wei, Li et al. 2011). The 532 nm pulse laser beams are delivered by
a 200-µm-core multimode optical fiber and emitted through the central hollow of the
ring-shaped ultrasonic transducer. This design steers both the coaxial laser beams and
the ultrasound beams into the sample by a customized micro rod mirror (platinum
coating, 2.0 mm OD, 4 mm length, with the reflective surface angled at 45˚ to the
probe’s axis). Also, the ultrasound echoes and the excited PA waves from the sample
are deflected by the mirror and detected by the ultrasound transducer. The micro rod
mirror shared by the coaxial laser and acoustic beams assures the total overlap of the
laser and acoustic beams along the transmitting path. Snell’s law governs the path of
reflected and refracted waves when the acoustic wave encounters an interface of two
media (Cobbold 2007). The ratio of sound-propagation speeds (1.5/5.1, longitudinal
wave; 1.5/3.3 shear wave) in water and glass is large enough so that the total internal
reflection occurs at the interface of water and glass (Yang, Maslov et al. 2009). In other
words, there is no additional propagation loss on the transmitting path of the ultrasonic
wave. The mirror, optical fiber and ultrasonic transducer are fixed and packaged in a
91

polyimide tube in which a window is made to allow the laser beams and ultrasound
beams to pass through. Figure 5-3 (b) shows the ring-shaped ultrasonic transducer with
2.2 mm outer diameter and 0.6 mm inner diameter. Figure 5-3 (c) is the whole
integrated probe that is packaged in a polyimide tube with a final packaged diameter of
2.3 mm.

Figure 5-3: Schematic of the combined IVPA probe (a); Photo of the ring-shaped
ultrasonic transducer (b); Photo of the combined IVPA probe after packaging (c)
5.4 IVPA Imaging Results
5.4.1 Imaging on Tissue Mimicking Phantom by the Coaxial Probe
In order to test the performance of the coaxial probe, a tissue-mimicking
phantom study was conducted to obtain IVUS and IVPA images. The cylindrical
92

phantom was made out of 10 wt% gelatin. Then, 2 wt% 0.5 to 15 µm silica dioxide
powder and 10 vol% low-fat milk were added as ultrasonic scatters and optical scatters,
respectively. As shown in Figure 5-4, a 5 mm diameter hollow was drilled in the center
of the phantom and acted as the lumen of tissue. A 0.5 mm diameter graphite rod,
serving as the optical absorption contrast, was embedded in the phantom at a distance
of 3.5 mm from the center. By drilling the phantom, a 0.5 mm diameter air lumen was
created, the same dimension as the graphite rod, which appeared as an air bubble in the
cross-section scanned. The tiny air lumen was 5 mm away from the center and acted as
the acoustic contrast.

Figure 5-4: Schematic of the phantom in which the 0.5 mm diameter graphite rod and
0.5 mm diameter air lumen are embedded. The graphite rod and air lumen work as two
contrast inclusions which have different absorption coefficient within a tissue-
mimicking phantom.

The images of a tissue-mimicking phantom from the same cross-section were
obtained by the combined IVUS and IVPA system. All images in Figure 5-5 cover a 13
mm diameter field of view. Figure 5-5 (a) is the IVUS image showing the general
93

structure of the phantom, lumen, and two inclusions. The 8 o’clock inclusion has a
strong signal at a depth of ~2.5 mm and that of the 1 o’clock has a strong signal at a
depth of ~1.0 mm. As prepared previously, it is known that one is the graphite rod
while the other one is the air lumen. These two inclusions have relatively distinct
acoustic impedance compared to the surrounding tissues so both of them appear as
hyperechoic regions. Besides located at different depths, it is difficult to distinguish the
two inclusions from each other by looking at the pattern of the IVUS image. In this case,
IVPA, which has the ability to show the optical specificity and the contrast within a
phantom tissue, is required. The IVPA image shown in Figure 5-5 (b) demonstrates an
optical absorption contrast within the phantom. According to the PA imaging
mechanism, the bright area in Figure 5-5 (b) implies an optical inclusion which has
high absorption at the 532 nm wavelength while the other tissue predominantly scatters
the light. Figure 5-5 (c) is the combined image of the phantom. Compared to the
surrounding tissue and the air lumen, the absorption of the graphite rod is higher at
wavelength 532 nm. Hence, the reflective area located at the 1 o’clock position in
Figure 5-5 should be the graphite rod, which matches the depth illustration in Figure
5-4. On the other hand, it can be deduced that the 8 o’clock position in the image is the
air lumen by comparing the IVUS and IVPA images. Therefore, by combining these
two imaging systems together, the whole tissue structure can be displayed while the
relative area of the inclusion can be recognized. The complementary information from
the images of the IVUS and IVPA above illustrate the synergy of the combined system.
94

Figure 5-5: Cross-sectional IVUS (a), IVPA (b), and combined image of the phantom
(c). The field of view is 6.5 mm in radius in all images. IVUS and IVPA images are
displayed in dynamic ranges of 50 dB and 40 dB.
5.4.2 In-vitro Imaging on Rabbit Aorta by the Coaxial Probe
By using the coaxially arranged probe, in vitro images of a normal rabbit aorta
were obtained to demonstrate imaging ability. The imaging result is shown in Figure
5-6.
95

Figure 5-6: Cross-sectional IVUS (a), IVPA (b), Hematoxylin-Eosin (H&E) stained
histological image (c), and combined image of a normal rabbit aorta (d). The IVPA
image was obtained using 532 nm optical excitation wavelength and 39 MHz ring-
shaped transducer. IVUS and IVPA images are displayed in dynamic ranges of 50 dB
and 40 dB.

It is clear that the IVUS image, shown in Figure 5-6 (a), provides a full-depth
cross-sectional image of the aorta. The ultrasound pulse echo signals are relatively
homogeneous from the vessel wall. Figure 5-6 (b) is the IVPA image from the same
96

cross-section of the aorta. The moderate response from the aorta implies the tissue
absorption at 532 nm is sufficient to get an image. The PA signal intensity is also
relatively homogeneous, which corresponds well to the IVUS image. At the 6 o’clock
position in Figure 5-6 (b), the IVPA image shows a maximum imaging depth of
approximately 1 mm. Figure 5-6 (c) is the hematoxylin and eosin (H&E) stained
normal rabbit aorta histological image. Figure 5-6 (d) shows the combined IVUS-IVPA
image. The similar shapes of ultrasound and PA images indicate the superiority of the
coaxial probe design in obtaining co-registered images.
5.4.3 Imaging on 6-µ m Wire Targets by the Side-by-Side Probes at 35 and 80 MHz
The photoacoustic images of the 6-µ m wire targets at 35 and 80 MHz by using
the side-by-side arrangement are shown in Figure 5-7 (a) and Figure 5-8 (a),
respectively. Figure 5-7 (b) and Figure 5-8 (b) show the envelopes of the photoacoustic
signals from the wire located around 2.5 mm away from the transducer surface. The
axial and lateral resolutions are determined from the -6 dB envelope width, which are
59 and 232 µ m at 35 MHz; 35 and 181 µ m at 80 MHz. The RF signals and spectrums
of photoacoustic pulses generated by the wire targets at 35 MHz and 80 MHz are
shown in Figure 5-9 (a) and (b), respectively. The pulse width at 35 MHz is roughly
twice as long as that at 80 MHz. The center frequency of the 35 MHz transducer down
shifted to 32 MHz and -6dB bandwidth is increased to 83%.
97

Figure 5-7: 35 MHz photoacoustic image of 6-µ m tungsten wires (a); axial and lateral
envelopes (b) of photoacoustic signal from the wire located at 2.5 mm away from the
transducer surface

Figure 5-8: 80 MHz photoacoustic image of 6-µ m tungsten wires (a); axial and lateral
envelopes (b) of photoacoustic signal from the wire located at 2.5 mm away from the
transducer surface
98

Figure 5-9: RF signals and spectrums of photoacoustic pulses generated by the wire
targets at 35 MHz (a) and 80 MHz (b)
5.4.4 In-vitro Imaging on Rabbit Aorta by the Side-by-Side Probes at 35 and 80
MHz
Healthy rabbit abdominal aortas were harvested and preserved in phosphate
buffer. The rabbit aortas were imaged by the Side-by-Side probes at 35 and 80 MHz,
respectively. During the experiment, the specimen was immersed and supported by a
sponge to stand in a water tank. Only the part of sample above the sponge was imaged.
99

Imaged sections were pinpointed for Hematoxylin-Eosin (H&E) stained histology
examination.
IVUS and IVPA images of a rabbit aorta at 35 MHz are shown in Figure 5-10.
Both images could see through the vessel wall. However, the layered structures in the
IVUS image looks more ambiguous than that in IVPA image. The IVPA imaging depth
could be demonstrated up to 4 mm at 12~2 o’clock in Figure 5-10(b). A rough
brightness comparison of adventitia and media areas in IVPA image reveals that
adventitia is 10~15 dB higher than media, while the number is 5~10 dB in IVUS, which
indicates that the contrast in IVPA is around 5~10 dB higher than that in IVUS. The
optical contrast enables IVPA to distinguish the boundaries of the three layers more
effectively than IVUS.
100

Figure 5-10: Cross-sectional IVUS (a), IVPA (b) and fused (c) images of a healthy
rabbit aorta at 35 MHz; and Hematoxylin-Eosin (H&E) stained histology image (d)
Higher working frequency of ultrasonic transducer implies better image
resolution. Further investigation of 80MHz IVPA is shown in Figure 5-11. The three
layers in 80 MHz IVPA image [Figure 5-11(b)] are differentiated much clearer around
the whole circumference of vessel wall than that in 35 MHz IVPA image. With
improved resolution and contrast, the profile of intima layer was depicted without any
ambiguity. However, the 80 MHz IVUS in this case didn’t show clear three-layer
structures, owing to the low acoustic contrast. High attenuation of 80 MHz ultrasonic
101

waves is responsible for the weak IVPA signals at certain parts of adventitia layer, e.g.
at the 5 o’clock in Figure 5-11(b).

Figure 5-11: Cross-sectional IVUS (a), IVPA (b) and fused (c) images of a healthy
rabbit aorta at 80 MHz; and Hematoxylin-Eosin (H&E) stained histology image (d)
The improved contrast of IVPA and higher resolution at 80 MHz in
differentiating layered structures could be more beneficial for diagnosing
atherosclerosis at an early stage, where intima is not significantly thickened and under
the detection threshold of IVUS. However, IVUS is still valuable to provide anatomical
information of thickened intima and plaque. Moreover, since blood has strong
102

absorption at 532 nm, which requires saline flushing, IVUS could serve as guidance
before flushing.
5.5 Discussion
In this IVPA research, two types of integrated IVPA-IVUS probes have been
introduced. By introducing the fiber delivery system to the ultrasound based catheter,
the sample could be illuminated internally which showed its clinical potential. In the
phantom study, the performance of the probe was evaluated and the complementary
IVUS/IVPA images emphasized the synergy of the combined ultrasound and PA
system. The coaxially designed probe provided co-registered IVUS and IVPA images
of a rabbit aorta which also demonstrated the feasibility of this dual-modality system.
Compared to IVUS, IVPA differentiated the aorta layered structures with higher clarity.
At 80 MHz, IVPA displayed extraordinary resolution, although the penetration depth
has not been fully investigated.
The preliminary results on IVPA imaging at 40 MHz and 80 MHz suggest the
potential of this new modality which can provide better contrasts of the vascular
layered structures. More experiments on rabbit and human specimens are needed to
verify the results. The current setup of the IVPA system was based on sample rotating,
which is not realizable for in vivo or clinical settings. To achieve the rotation of probe,
an electrical slip ring and an optical rotary joint, or an integrated rotary joint, is required
for electrical and optical signal coupling. Our current researches were focused on the
103

different probe design, and more researches on the various transducers should be
examined extensively, including the different transducer material, geometries, and
working frequencies. Optimizing the transducer’s receiving performance is a way to
improve the imaging resolution and contrast.
104

Bibliography
Beard, P. C. and Mills, T. N. (1997). "Characterization of post mortem arterial tissue
using time-resolved photoacoustic spectroscopy at 436, 461 and 532 nm."
Physics in medicine and biology 42(1): 177-198.
Calzada, M. L., Alguero, M., et al. (2009). "Piezoelectric, ferroelectric
Pb(Mg[1/3]Nb[2/3])O[3]-PbTiO[3] thin films with compositions around the
morphotropic phase boundary prepared by a sol-gel process of reduced thermal
budget." Journal of materials research 24(2): 526-533.
Cannata, J. M., Ritter, T. A., et al. (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(11):
1548-1557.
Cannata, J. M., Willianms, J. A., et al. (2008). "Self-focused ZnO transducers for
ultrasonic biomicroscopy." Journal of applied physics 103(8): 084109.
Cobbold, R. S. C. (2007). Foundations of biomedical ultrasound. Oxford ; New York,
Oxford University Press.
Desilets, C. S., Fraser, J. D., et al. (1978). "The design of efficient broad-band
piezoelectric transducers." IEEE transactions on sonics and ultrasonics 25(3):
115-125.
Farooq, M. U., Khasnis, A., et al. (2009). "The role of optical coherence tomography in
vascular medicine." Vascular medicine 14(1): 63-71.
Fitzgerald, P. J., St Goar F. G., et al. (1992). "Intravascular ultrasound imaging of
coronary arteries. Is three layers the norm?" Circulation 86(1): 154-158.
Foster, F. S., Obara, H., et al. (1994). "Ultrasound backscatter from blood in the 30 to
70 MHz frequency range." IEEE ultrasonics symposium proceedings 3: 1599-
1602.
Foster, F. S., Pavlin, C. J., et al. (2000). "Advances in ultrasound biomicroscopy."
Ultrasound in medicine & biology 26(1): 1-27.
Hsieh, B. Y., Chen, S. L., et al. (2010). "Integrated intravascular ultrasound and
photoacoustic imaging scan head." Optics letters 35(17): 2892-2894.
Huang, D., Swanson, E. A., et al. (1991). "Optical coherence tomography." Science
254(5035): 1178-1181.
105

Jang, I.-K., Bouma, B. E., et al. (2002). "Visualization of coronary atherosclerotic
plaques in patients using optical coherence tomography: comparison with
intravascular ultrasound." Journal of the american college of cardiology 39(4):
604-609.
Jansen, K., Van Der Steen, A. F., et al. (2011). "Intravascular photoacoustic imaging of
human coronary atherosclerosis." Optics letters 36(5): 597-599.
Karpiouk, A. B., Wang, B., et al. (2010). "Development of a catheter for combined
intravascular ultrasound and photoacoustic imaging." The review of scientific
instruments 81(1): 014901.
Karpiouk, A. B., Wang, B., et al. (2010). "Integrated catheter for intravascular
ultrasound and photoacoustic imaging." SPIE photonics west proceedings 7564:
756408.
Kawasaki, M., Bouma, B. E., et al. (2006). "Diagnostic accuracy of optical coherence
tomography and integrated backscatter intravascular ultrasound images for
tissue characterization of human coronary plaques." Journal of the american
college of cardiology 48(1): 81-88.
Keijzer, M., Richards-Kortum, R. R., et al. (1989). "Fluorescence spectroscopy of
turbid media: Autofluorescence of the human aorta." Applied optics 28(20):
4286-4292.
Kolodgie, F. D., Burke, A. P., et al. (2001). "The thin-cap fibroatheroma: a type of
vulnerable plaque: the major precursor lesion to acute coronary syndromes."
Current opinion in cardiology 16(5): 285-292.
Kosec, M., Holc, J., et al. (2007). "Pb(Mg1/3Nb2/3)O3-PbTiO3 thick films from
mechanochemically synthesized powder." Journal of the european ceramic
society 27(13-15): 3775-3778.
Ku, G., Wang, X., et al. (2004). "Multiple-bandwidth photoacoustic tomography."
Physics in medicine and biology 49(7): 1329-1338.
Kume, T., Akasaka, T., et al. (2005). "Assessment of coronary intima-media thickness
by optical coherence tomography: comparison with intravascular ultrasound."
Circulation journal : official journal of the japanese circulation society 69(8):
903-907.
Landini, L. and Verrazzani, L. (1990). "Spectral characterization of tissues
microstructure by ultrasounds: a stochastic approach." IEEE transactions on
ultrasonics, ferroelectrics, and frequency control 37(5): 448-456.
106

Lee, H. J., Zhang, S., et al. (2010). "Thickness-dependent properties of relaxor-PbTiO3
ferroelectrics for ultrasonic transducers." Advanced functional materials 20(18):
3154-3162.
Li, X., Wu, W., et al. (2011). "80-MHz intravascular ultrasound transducer using PMN-
PT free-standing film." IEEE transactions on ultrasonics, ferroelectrics and
frequency control 58(11): 2281-2288.
Li, X., Yin, J., et al. (2010). "High-resolution coregistered intravascular imaging with
integrated ultrasound and optical coherence tomography probe." Applied
physics letters 97(13): 133702.
Luo, H., Shih, W. Y., et al. (2007). "Double precursor solution coating approach for
low-temperature sintering of [Pb(Mg1/3Nb2/3)O3]0.63[PbTiO3]0.37 solids."
Journal of the american ceramic society 90(12): 3825-3829.
Lusis, A. J. (2000). "Atherosclerosis." Nature 407(6801): 233-241.
Pasterkamp, G., Falk, E., et al. (2000). "Techniques characterizing the coronary
atherosclerotic plaque: influence on clinical decision making?" Journal of the
american college of cardiology 36(1): 13-21.
Patwari, P., Weissman, N. J., et al. (2000). "Assessment of coronary plaque with optical
coherence tomography and high-frequency ultrasound." The american journal of
cardiology 85(5): 641-644.
Potkin, B. N., Bartorelli, A. L., et al. (1990). "Coronary artery imaging with
intravascular high-frequency ultrasound." Circulation 81(5): 1575-1585.
Puri, R., Worthley, M. I., et al. (2011). "Intravascular imaging of vulnerable coronary
plaque: current and future concepts." Nature reviews. Cardiology 8(3): 131-139.
Rathore, S., Terashima, M., et al. (2011). "In-vivo detection of the frequency and
distribution of thin-cap fibroatheroma and ruptured plaques in patients with
coronary artery disease: an optical coherence tomographic study." Coronary
artery disease 22(1): 64-72.
Ross, R. (1999). "Atherosclerosis-an inflammatory disease." The new england journal
of medicine 340(2): 115-126.
Sarkola, T., Redington, A., et al. (2010). "Transcutaneous very-high-resolution
ultrasound to quantify arterial wall layers of muscular and elastic arteries:
validation of a method." Atherosclerosis 212(2): 516-523.
107

Sawada, T., Shite, J., et al. (2008). "Feasibility of combined use of intravascular
ultrasound radiofrequency data analysis and optical coherence tomography for
detecting thin-cap fibroatheroma." European heart journal 29(9): 1136-1146.
Sethuraman, S., Aglyamov, S. R., et al. (2007). "Intravascular photoacoustic imaging
using an IVUS imaging catheter." IEEE transactions on ultrasonics,
ferroelectrics, and frequency control 54(5): 978-986.
Sethuraman, S., Amirian, J. H., et al. (2007). "Ex vivo characterization of
atherosclerosis using intravascular photoacoustic imaging." Optics express
15(25): 16657-16666.
Shung, K. K. (2006). Diagnostic ultrasound : imaging and blood flow measurements.
Boca Raton, FL, Taylor&Francis.
Su, J., Zhang, J., et al. (2008). "Real-time swept source optical coherence tomography
imaging of the human airway using a microelectromechanical system endoscope
and digital signal processor." Journal of biomedical optics 13(3): 030506.
Tearney, G. J., Jang, I. K., et al. (2006). "Optical coherence tomography for imaging
the vulnerable plaque." Journal of biomedical optics 11(2): 021002.
Wang, B., Su, J., et al. (2009). "On the possibility to detect lipid in atherosclerotic
plaques using intravascular photoacoustic imaging." Annual international
conference of the IEEE engineering in medicine and biology society. 2009:
4767-4770.
Wang, B., Su, J. L., et al. (2010). "Intravascular photoacoustic imaging." IEEE journal
of quantum electronics 16(3): 588-599.
Wang, L. V. and Hu, S. (2012). "Photoacoustic tomography: in vivo imaging from
organelles to organs." Science 335(6075): 1458-1462.
Wei, W., Li, X., et al. (2011). "Integrated ultrasound and photoacoustic probe for co-
registered intravascular imaging." Journal of biomedical optics 16(10): 106001.
Yang, J. M., Maslov, K., et al. (2009). "Photoacoustic endoscopy." Optics letters 34(10):
1591-1593.
Yin, J., Li, X., et al. (2011). "Novel combined miniature optical coherence tomography
ultrasound probe for in vivo intravascular imaging." Journal of biomedical
optics 16(6): 060505.
108

Yin, J., Yang, H.-C., et al. (2010). "Integrated intravascular optical coherence
tomography ultrasound imaging system." Journal of biomedical optics 15(1):
010512.
Zhou, Q., Lau, S., et al. (2011). "Piezoelectric films for high frequency ultrasonic
transducers in biomedical applications." Progress in materials science 56(2):
139-174.
Zhu, B. P., Wu, D. W., et al. (2008). "Lead zirconate titanate thick film with enhanced
electrical properties for high frequency transducer applications." Applied
physics letters 93(1): 012905.

Abstract (if available)

Abstract Atherosclerosis is a complex syndrome characterized by plaques build up on the inner lining of arteries, which is the leading cause of morbidity in developed countries. Conventional gold standard for diagnosing the vulnerability of plaques is intravascular ultrasound (IVUS) imaging working in the 20~40 MHz range. At this range, the resolution (~100 µm) is insufficient for detecting a number of critical microstructures that are in the scope of less than 65 µm. Driving the working frequency above 80 MHz could be a feasible solution for improving IVUS resolution. Meanwhile, intravascular Optical Coherence Tomography (OCT) is a new imaging modality recently approved for clinical use. OCT features high resolution (10~30 µm) but limited penetration depth (~1 mm). The other emerging optical modality: Intravascular Photoacoustic (IVPA) is based on the tissue optical absorption properties and has been proved to offer better contrast of tissue composition. ❧ In this dissertation, three intravascular imaging modalities have been investigated: integrated IVUS-OCT imaging, very high frequency IVUS, and IVPA imaging, which aim to overcome the limitations in resolution, penetration and imaging contrast of current technologies. Imaging systems and different types of ultrasonic or optical-ultrasonic hybrid probes have been designed and fabricated for using in each modality. In vitro and in vivo studies have been carried out to validate these new imaging concepts. The imaging results successfully support the enhancements brought by the new modalities. The complementary synergy of ultrasonic and optical modalities provides extra diagnosing information. This research suggests a bright future for multi-modality imaging that combines the conventional IVUS and new optical technologies.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Multi-modality intravascular imaging by combined use of ultrasonic and opticial techniques

PDF

Development of high frequency focused transducers for single beam acoustic tweezers

PDF

High-frequency ultrasonic transducers for photoacoustic applications

PDF

High-frequency ultrasound array-based imaging system for biomedical applications

PDF

Advamces in intravascular ultrasound (IVUS)-based technology

PDF

A high frequency array- based photoacoustic microscopy imaging system

PDF

High frequency ultrasonic phased array system and its applications

PDF

High frequency ultrasound elastography and its biomedical applications

PDF

Development of novel 1-3 piezocomposites for low-crosstalk high frequency ultrasound array transducers

PDF

High frequency ultrasonic imaging for the development of breast biopsy needle with miniature ultrasonic transducer array

PDF

Development of high-frequency (~100mhz) PZT thick-film ultrasound transducers and arrays

PDF

Array transducers for high frequency ultrasound imaging

PDF

Single-cell analysis with high frequency ultrasound

PDF

Configurable imaging platform for super-harmonic contrast-enhanced ultrasound imaging

PDF

Microfluidic cell sorting with a high frequency ultrasound beam

PDF

High resolution elastography in ophthalmology

PDF

Quantification of cellular properties using ultra-high frequency single-beam acoustic tweezer

PDF

Development of front-end circuits for high frequency ultrasound system

PDF

Novel beamforming with dual-layer array transducers for 3-D ultrasound imaging

PDF

Development of back-end processing system for high frequency ultrasound b-mode imaging

Asset Metadata

Creator Li, Xiang (author)

Core Title Intravascular imaging on high-frequency ultrasound combined with optical modalities

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Biomedical Engineering

Publication Date 12/13/2013

Defense Date 05/08/2012

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag high frequency,intravascular,IVPA,IVUS,OAI-PMH Harvest,OCT,photoacoustic,ultrasound

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Shung, Kirk Koping (committee chair), Zhou, Qifa (committee chair), Nutt, Steven R. (committee member), Yen, Jesse T. (committee member)

Creator Email xiangli@usc.edu,xiangli1223@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-46527

Unique identifier UC11290497

Identifier usctheses-c3-46527 (legacy record id)

Legacy Identifier etd-LiXiang-888.pdf

Dmrecord 46527

Document Type Dissertation

Rights Li, Xiang

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