A 1.5D Array for Acoustic Radiation Force (ARF)-Induced Peak Displacement Based Tissue Anisotropy Assessment with a Row-Column Excitation Method

Abstract

Many biological tissues including muscle or kidney are mechanically anisotropic and degree of anisotropy (DoA) in mechanical properties is diagnostically relevant. DoA can be assessed either using the ratio of shear wave velocities (SWV) or acoustic radio forced impulse (ARFI)-induced peak displacements (PD) measured longitudinal over transverse orientations. Whether using SWV or PD as a basis, DoA expressed as the ratio of values requires 90° transducer rotation when a linear array is employed. This large rotation angle is prone to misalignment errors. One solution is the use of a fully sampled matrix array for electronic rotation of point spread function (PSF). However, the challenges of matrix array are its high fabrication cost and complicated fabrication procedures. The cheaper and simpler alternative of matrix array is the use of a row-column array. A 3×64 elements 1.5D array with a row-column excitation mode is proposed to assess DoA in mechanical properties using the PD ratio. Different numbers of elements in elevational and lateral-directions were selected to have orthogonal ARFI excitation beams without rotating the transducer. A custom-designed flex circuit was used to fabricate the array with a simpler electrode connection than a fully sampled matrix array. The performance of the array was evaluated in Field II simulation and experiment. The output pressure was 0.57 MPa output under a 40 Vpp excitation with a −6 dB point spread dimension of 14×4 mm² in orthogonal directions. The peak displacement was measured to be 1.4 μm in an isotropic elastic phantom with Young’s modulus of 5.4 kPa. These results suggest that the array is capable of assessing DoA using PD ratio without physical rotation of the transducer. The array has the potential to reduce the misalignment errors for DoA assessment.

I. Introduction

Mechanical properties of many biological tissues including muscle [1], tendon [2], and kidney [3], are directionally dependent. These directional differences or degree of anisotropy (DoA) has shown to be diagnostically relevant. For example, muscle fiber fragmentation and disordered fatty/fibrous deposition with dystrophic degeneration degrade the DoA in affected muscles [4-5]. In kidney, the DoA of renal parenchyma changed with perfusion and inflammation [6-8]. For breast cancer diagnosis, malignant cancer has shown higher anisotropy compared with benign cancer l [9].

Conventionally, the DoA of mechanical properties has been assessed as the ratio of shear wave velocities (SWV) measured longitudinal over transverse orientation [10-11]. The longitudinal and transverse direction is relative to the direction of muscle fibers or nephron direction in the kidney. More recently, Hossain et. al developed a novel method to assess DoA of mechanical properties as the ratio of PDs achieved when the long axis of an asymmetric ARFI excitation point spread function (PSF) is aligned along versus across the material’s axis of symmetry (AoS) in transverse isotropic (TI) materials [12]. The advantage of this method is that the DoA is assessed in the ARFI region of excitation (ROE) without observing shear wave propagation, so measurements are less susceptible to error due to tissue inhomogeneities and impaired shear wave propagation. Similarly, this DoA assessment can be extended for Viscoelastic response (VisR) derived relative elasticity and relative viscosity [13-15]. VisR is a novel ARFI based method to interrogate tissue viscoelastic properties of tissue. Previously, VisR derived DoA has been used to detect renal inflammation [7-8] and dystrophic muscle degeneration [16].

Whether using SWV or ARFI PD or VisR derived relative elasticity and viscosity as a basis, DoA expressed as the ratio of values measured along versus across the fiber requires 90° transducer rotation when a linear array is employed. This large rotation angle is prone to misalignment error and could lead to intra- and inter-operator variability. Such error could be minimized by electronically rotating the PSF 90° using a fully sampled matrix array transducer [17]. However, a fully sampled matrix array is costly and difficult to manufacture.

To avoid the complicated fabrication process for the fully sampled matrix array, one solution is to use the row-column array which has less complex wire connections. Previous work has reported the use of row-column array for 3D imaging or Doppler imaging, which however may suffer from the insufficient energy for the ARF excitation and different beam shape in azimuth and elevation directions due to the element non-uniformity [18-19]. On the other hand, the previously reported 1.5 D row-column array [20] has been applied in imaging and surgery and can provide the uniform element impedance and sensitivity with the same equal-area rows design [21]. In an in-silico study, Hossain et al. used a row-column array to electronically rotate the PSF for assessing DoA [22-23]. However, the transducer design was unable to provide a 90° rotation electronically, which reduced the sensitivity of assessing the change in DoA. More recently, our group has reported a 1.5D array worked in row-column excitation mode with various elevational element dimension for the DoA assessment. Although orthogonal beam profiles were achieved, the dimension difference caused an electrical impedance mismatch, which made it hard to provide enough pressure output for excitation in orthogonal directions [24]. Therefore, in this paper, we proposed a custom 1.5D array which contains different number of elements in elevational and lateral directions. The array provides the 90° electronic rotation of the ARF excitation PSF with simpler connection compared with 2-D array. In this manuscript, we present the fabrication, in silico and experimental results of the proposed 1.5D array with row-column excitation method.

II. Materials and Methods

A. Transducer Design

A 64×3 elements array was designed to fulfill the need of two identical ARF excitation PSF in orthogonal directions (Fig. 1(a)). 64 rows (i.e., X1, X2...X64) and 3 columns (i.e., Y1, Y2, Y3) were deployed along the x and y axes, respectively. Fig. 1(b) showed a specific view of each element. The thickness of the active layer was half of the wavelength so as that the center frequency was to be 4 MHz. Also, the thickness of the matching layer was designed to become the quarter of the wavelength. Fig. 1(c) showed the array structure of the transducer. The pitch size, along the lateral direction (y), was 230 μm with a kerf of 30 μm. The width of each element, along the elevational direction (x), was 4.98 mm with the same kerf of 30 μm. The aperture size of the total array was 15×15 mm². The material for the active layer is chosen as PZT-5H. Using the KLM model simulation [25], the Al₂O₃/epoxy was chosen as the matching layer for its good impedance match and easy fabricating process. The air bubble/epoxy mixture was chosen as the backing layer to produce a relatively large pressure output. The material properties are listed in Table I. The center frequency (4 MHz), used in this array, was enough to achieve the image depth over 20 mm. To potentiate the acoustic pressure output induced by the array, the matching layer was designed to have the quarter wavelength and the backing layer thickness was to be around 6-times of the wavelength. 64×3 array design was adopted to provide relatively similar beam profiles in orthogonal directions with an easier fabrication process compared with traditional fully sampled matrix array.

Fig. 1.

(a) Schematic of the row-column array. 64 elements and 3 elements were designed in X-array and Y-array respectively. Two different transmit modes (XG and Y2) was designed for the orthogonal ARF excitation to measure PD. The Y2-array is also used for generating ARF tracking PSF. (b) Schematic for the design of single element. (c) Structure of the array with the pitch size in elevational (x) and lateral directions (y)

TABLE I.

Design parameters for the array

	Material	Thickness	Properties
Active layer	PZT-5H	350 μm	Impedance	33.7 MRayl
			Velocity	4350 m/s
			Density	7750 kg/m3
Matching layer	A1203/epoxy	150 μm	Impedance	5.4 MRayl
			Velocity	2700 m/s
			Density	2000 kg/m3
Backing layer	Air-bubble/epoxy	500 μm	Impedance	0.5 MRayl
			Velocity	600 m/s
			Density	840 kg/m3

B. In Silico: ARF Excitation and Tracking Beam

To predict the performance of the array, Field II [26-27], acoustic simulation software was used to simulate the ARF excitation beam. One ARF excitation PSF was generated using Y2 array (Fig. 1) with the focal depth of 20 mm, which covers 1/3 of the array aperture with 64 elements. To have identical and orthogonal to the first ARF excitation PSF using Y2 array, every 21 adjacent elements (X21 to X41) in Y1, Y2 and Y3 array were used for the second ARF excitation as shown in Fig. 1(a), which totally covers 63 (i.e., 21×3) elements. For the ARF excitation, one burst excitation of 100 cycles was applied for pushing the tissue and creating the displacement.

For the ARF tracking PSF, the array works as a conventional 1D array (Y2 array). The operating frequency was 4 MHz with a focal distance of 20 mm. For the tracking process, 21 elements (X21 to X41) on Y2 array were applied with the f-number of 4.32.

C. Transducer Fabrication

The fabrication process was shown in Fig. 2. A PZT-5H plate (TRS Technologies, Inc., State College, PA, USA) was first lapped to 350 μ (i.e., the half of the wavelength). Then a 5/100 nm Ti/Au layer was deposited as the electrodes on both sides. The sample was diced into 3 columns (Y1, Y2, and Y3 in Fig. 1) with a width of 5 mm. Then, each column was diced into 64 elements with a pitch size of 230 μ and kerf of 30 μ by a dicing saw (DAD 320, DISCO HI-TEC AMERICA, INC.). For each kerf, it was fully diced and filled with epoxy to make sure that the crosstalk between adjacent elements was minimized. The flex circuit (Fig. 3) was designed and fabricated with a pitch size of 230 μ and an electrode width of 160 μ. The width of the electrode in the flex circuit was smaller than that in the transducer element (i.e., 200 μm) to minimize the potential of misconnection with the adjacent elements. After dicing, the array was connected with the flex circuit shown in Fig. 2. The Y1 and Y3 array was first aligned and bonded to the two flex circuits respectively using the epoxy resin (EPO-TEK 301, Epoxy Technology Inc. Billerica, MA). A customed jig was applied to maintain the alignment under a microscope and provide the pressure required for the bonding process. In the room temperature, the epoxy was fully cured for 24 hours. Then, the Y2 array was bonded to another flex circuit with the same method. The EPO-TEK 301 was applied to fill the kerfs between every two elements and to assemble the 3 columns with the kerf of 30 μ between adjacent columns. A 150 μ matching layer of Al₂O₃/epoxy with deposited electrodes was bonded to the top of the array with the E-solder 3022 as a common ground. At the backside of the array, air-backing was fabricated by making a composite of air microbubbles (Blatek Inc., State College, PA) and epoxy (Epotek®301, Epoxy Tech. Inc., San Jose, CA) with a volume ratio of 3:1. (0.5 MRayl) to provide a stronger signal for the ARF excitation beam. The array was attached to a 3-D printed fixture and the flex circuit was connected to the PCB board for elements numbering. In the last step, each channel on the PCB board was labeled and welded to a Verasonics Vantage system (Verasonic Inc., Kirkland, WA) connector according to the pin map [28].

Fig. 2.

Schematic for fabrication process of the row-column array (a) Printed the flex circuit. (b) Diced PZT-5H material with bonding to the flex circuit. (c) Assembled the array with epoxy (d) Added the backing layer, matching layer and connected common ground wire

Fig. 3.

Design of the flex circuit with electrodes for array connection

D. Transducer characterization

The impedance, dielectric loss, and capacitance of each element were measured using an impedance analyzer (Agilent 4294A, Agilent Technologies Inc, Santa Clara, CA). A pulser/receiver (5900 PR, Olympus Corp, Waltham, MA) was used for the pulse-echo test of each element in a water tank and an oscilloscope (Agilent DSO7014B, Agilent Technologies Inc, Santa Clara, CA) was used for monitoring and recording the signals. The electrical impedance of the pulser/receiver was set to be 50 Ohm with the bandpass filter from 1 MHz to 10 MHz and no attenuation and gain were applied for the test. The received signal was used to calculate the center frequency and bandwidth. The applied energy was 1 μ and a steel bar (30 mm × 30 mm × 8 mm) was used as a reflective surface with a distance of 3 mm from the transducer surface. To test the pressure output from a single element, the transducers were driven by 10-cycle sinusoidal inputs at 4 MHz using a function generator (AFG3101, Tektronix Inc., Beaverton, OR, USA) connected with an RF amplifier (75A250A RF Amplifier, Amplifier Research Corp., WA, USA). A hydrophone (HNA-0400, Onda Corp., Sunnyvale, CA, USA) was applied to measure the pressure output at a distance of 20 mm. Then, a 256-channel Verasonic Vantage system was used to drive the transducer to assess the beam profile of the array. Fig. 4 showed the experimental set up for beam profile characterization. 5-cycle excitation signal was applied with 50 Vpp and the PRF of 200 Hz. For the Y2 array operation, the 64 elements (X1 to X64, Y2 array) with equivalent elevation and azimuth width of 15 mm and 5 mm, respectively were excited simultaneously. Similarly, the 63 elements (X21 to X41, Y1 to Y3) with equivalent elevation and azimuth width of 5 mm and 15 mm, respectively were excited simultaneously for XG array operation. The beam profiles for Y2 and XG transmit were measured by translating the hydrophone in axial, azimuth, and elevational dimensions.

Fig. 4.

Experiment setup for the beam profile characterization. Left: Schematic of the experiment setup. Right: The photo for the experiment setup. The pressure output was tested in an axial range of 5 mm to 25 mm. The pressure output beam profile was plotted at 20 mm in the orthogonal directions.

E. Experiment: ARFI Imaging of a Custom Phantom

A custom gelatin-graphite phantom was created at a physiologically-relevant acoustic attenuation and speed of sound using a modified version of the recipe described in [29-30]. Gelatin from porcine skin, gel strength 300, Type A (Sigma-Aldrich G2500), Synthetic graphite nanopowder (Sigma- Aldrich 282863), and 1-propanol (Sigma-Aldrich W292818) with the concentration of 3.43 wt%, 9.1 wt%, and 2 wt% was used.

ARF excitation capability was tested on the custom phantom. ARFI imaging was performed by connecting the transducer with the Verasonics Vantage system. The transducer was controlled by the customized Matlab code (The MathWorks, Inc, MA, USA). For the Y2 and XG excitation, 64 (63) elements were applied to push the tissue at the target depth of 20 mm. A 100-cycle burst excitation was applied and the tissue was then pushed by the Y2-beam and XG-beam respectively. Two two-cycle imaging pulses were collected before the excitation pulse. These two cycle imaging pulse was served as a reference tracking pulse and displacement was estimated respect to the reference pulse. Then, the excitation pulse was applied to the array to push the tissue. After that, two-cycle imaging pulse was transmitted to estimate the excitation pulse induced deformation at the focal depth of 20 mm.

The acquired raw channel data were beamformed using a delay and sum beamformer. ARFI excitation pulse-induced motion was measured using one-dimensional axial normalized cross-correlation (NCC) [15][31] with parameters: 4× spline-based upsampling of RF data (natively sampled at 40 MHz), 770-μm kernel length (i.e. 2λ, where λ is the wavelength of the tracking pulse assuming a speed of sound of 1540 m/s), and an 80-μm search region. A linear filter [31] was applied to displacement profiles.

III. Results

A. Array Performance in simulation

The single element pulse-echo simulation result has shown a sensitivity of 32 mV/V with a target at a distance of 3 mm (Fig. 5). The −6 dB bandwidth was 51 %.

Fig. 5.

Pulse-echo response from a single element (a) Simulated result in KLM model (b) Measured result under the 1 μJ excitation at 3 mm. A steel target was applied in both the simulation and the experiment.

The comparison between the XG- and Y2-ARF excitation PSF was shown in Fig. 6. The Y2 and XG- ARF excitation PSF had −6 dB beamwidth of 4 mm in the y-z plane and x-z plane, respectively. In the y-z plane and x-z plane, Y2 and XG- ARF excitation PSF had −6 dB beamwidth of 15 mm respectively. The simulation results showed the approximately same (< 5% difference) but 90° rotated −6 dB beamwidths at 20 mm. Fig. 7 showed the simulation result of the beam spread function at the depth of 20 mm for both XG excitation mode and Y2 excitation mode. Both modes showed relatively uniform distributions in the pressure output within −6 dB, covering an area of about 14×4 mm². Thus, the simulation results show the capability of electronic beam rotation by using the designed 3×64 1.5D array transducer.

Fig. 6.

Simulated results for the excitation beam (a) The schematic for the array orientation (b) Excitation beam profile in orthogonal directions.

Fig. 7.

Beam spread function at the depth of 20 mm (a) XG excitation (b) Y2 excitation

B. Impedance and capacitance measurement

The impedance, dielectric loss, and capacitance of each element were measured using an impedance analyzer (Agilent 4294A, Agilent Technologies Inc, Santa Clara, CA). Fig. 8 indicates a typical electric response obtained by experiment. The average impedance level was 291±43 Ohm at 3.9 MHz in air.

Fig. 8.

(a) Measured impedance and phase of one representative single element (b) Measured capacitance and loss of all the 192 elements.

C. Transducer characterization

For the pulse-echo test, the applied energy was 1 μ and a steel bar (30 mm × 30 mm × 8 mm) was used as a reflective surface with a distance of 3 mm from the transducer surface. As shown in Fig. 5(b), the central frequency was 3.39±0.3 MHz with the −6 dB bandwidth of 52.4±10.9%. The peak-to-peak voltage can reach up to 203±25 mV. Compared with the simulation results, the waveform has some distortion which may be caused by the variance of backing layer uniformity. The pressure output at 20 mm was measured with the Y2-pushing mode and XG-pushing mode respectively. The −6 dB area was estimated for the beam distribution and compared with the simulated results in Fig. 7, which showed ARF excitation PSF had higher asymmetry. The DoA assessed as the PD ratio depends on the asymmetry of the ARF excitation PSF. The sensitivity of detecting anisotropic features using ARFI PD based DoA increases with the higher asymmetry of the ARF excitation PSF [12] [23]. The measured beam profile has been shown in Fig. 10. The beam profile from the Y2-pushing and XG-pushing has shown a similar distribution of 14 × 4 mm² with a 90° rotation. Fig. 9(a) and (b) showed the pressure output from one element and the XG-excitation mode, respectively. The pressure output from the Y2-group pushing was slightly larger than that from XG-pushing under the same excitation condition (Peak-negative pressure: 0.62 MPa vs. 0.59 MPa). This could be predicted in the numerical simulation; the beam profile with the Y2 excitation mode showed a relatively high-pressure output along the centerline. Position error in the motion stage, manually driven, could affect the accuracy in the beam profile.

Fig. 10.

(a) Measured beam profile with XG-excitation beam (b) Measured beam profile with Y2-excitation beam at distance of 20 mm.

Fig. 9.

Measured pressure output from the (a) single element under 50 V excitation with 5 cycles at 7.5 mm (b) XG-excitation mode under 50 V excitation with 1 cycle at 20 mm.

D. Phantom Experiment

The experimental setup and corresponding B-mode of the phantom are shown in Fig. 11. The B-mode scanning was first acquired with the Y2 array. The imaging from the hydrophone tip has been shown in Fig. 12 clearly as a bright spot due to the strong reflection. Fig. 12 was transformed into binary imaging with a threshold of −6 dB. Then the dimension of the spot was measured as 1.76 mm by 3.61 mm, which is close to the actual lateral size of the transducer head (3.5 mm).

Fig. 11.

Experimental set up and B-mode imaging of the phantom. The reflection from the bottom of the container can be shown as the white line in the B-mode imaging.

Fig. 12.

B-mode imaging of the hydrophone head and the dimension estimation in the binary imaging with a threshold of −6 dB.

Fig. 13 showed the measured displacement profiles in the phantom at a focal depth of 20 mm. By adjusting the input voltage to 75 Volt for the Y2 pushing and 80 Volt for the XG pushing, the same pressure output (0.62 MPa) was acquired in the orthogonal direction and peak displacements of 1.4 μ have been reached in both directions.

Fig. 13.

Measured displacement profile with (a) Y2-excitation (b) XG-excitation mode. For each test, 8 elements were applied for tracking the peak displacement and shown in different colors.

IV. Discussion

This work presented a 1.5D array with the row-column excitation mode for the measurement of tissue anisotropy which can generate a similar beam profile in the orthogonal directions without rotating the transducer. The phantom test has shown the capability of the array in reaching the same PD with an electrical 90° beam rotation.

The pressure output from the Y2-group pushing is slightly larger than that from XG-pushing under the same excitation condition and this has been predicted in the simulation. However, since the beam profiles were similar in the orthogonal directions, the same peak displacement could be achieved by adjusting the excitation signal in orthogonal directions.

However, this array still has a problem of insufficient energy output for reaching a large PD in stiffer tissues. To compare the performance of the array with the commercial transducers, the peak negative pressure output (PNP), mechanical index (Mi), and the spatial-peak temporal-average intensity (I_spta) have been measured and calculated. For the commercial linear array (Siemens VF7-3 linear array transducer, Siemens Medical solutions UsA, inc., Ultrasound Division) working at the ARF excitation, the pushing PNP could reach 1.35 MPa at 4.0 MHz, which corresponding to an Mi of 0.55. For the row-column array, the pushing PNP is 0.61 MPa at 3.39 MHz, which corresponding to an Mi of 0.33. As for the ispta, the commercial array could reach 2.24 W/cm² while the row-column array was 0.57 W/cm² under 75 Volt input with 100 cycles excitation and PRF of 0.2 kHz. The relatively lower pressure output gives more error when the row-column array was applied in the tissue or phantom with a larger attenuation coefficient.

The relatively lower pressure output of the row-column array may due to the following reasons. Compared with the traditional linear array, the 1.5D array has a relatively smaller element size (15 mm×0.4 mm vs. 5 mm × 0.23 mm), which corresponding to a larger electrical impedance. Besides, the packaging of the array has induced more noise as the passivation layer was not perfect due to the experiment limitation. Furthermore, the commercial linear array usually has more elements (128 elements compared with 64 elements), which can have a better tracking beam profile and smaller jitters. However, since the current linear array rotation has been mostly carried out manually with the B-mode guidance, the potential misalignment may strongly affect the results. If the transducer was not properly aligned along or across the tissue, the measured results will reflect the mixture of the transverse and longitudinal moduli and artificially reduced the degree of anisotropy. Therefore, the 1.5D array with electronic PSF rotation may improve the tissue anisotropy estimation compared to the manual rotation of a linear array transducer. Considering its good uniformity of the beam profile in the orthogonal directions, the next step is to meet the requirement of in vivo use.

Orthogonal rotation of the PSF was achieved using our proposed 1.5D array (Fig. 7 and Fig. 10). Though we used only three elevational elements, 90° rotation of the PSF was achieved using the row-column excitation method. These orthogonally oriented ARF excitation pulse will generate different displacements at two orientations in the anisotropic materials. As demonstrated in the previous study [12], PD is different at two orthogonal orientations because when the long axis of the ARF excitation PSF was oriented along versus across the AoS, it predominantly interrogated transverse versus longitudinal shear modulus, respectively. Then, the degree of anisotropy (DoA) was assessed as the ratios of peak displacement (PD) achieved when an asymmetrical ARF excitation PSF was oriented along verses across the axis of symmetry (AoS) of the transversely isotropic material [12].

One limitation of the study was the use of the same imaging or tracking pulse PSF for both orthogonal excitation pulse. Note, the imaging pulse was used to track the ARF excitation pulse induced displacement. The imaging PSFs are impactful to DoA assessment because they determine the degree to which PD is underestimated due to shearing artifacts under the tracking PSF [32]. The use of the same imaging PSF for the orthogonal ARF excitation PSF may induce different amounts of jitter and displacement underestimation at two orientations. Future works will explore the feasibility of using orthogonal PSF for the imaging pulse.

To improve the pushing and tracking performance, a harder piezoelectric material may be applied for sustaining higher electrical input. The prototyped transducer utilized an air/epoxy backing, having a relatively low acoustic impedance, to intensify the acoustic pressure output induced by the array. This may compromise the image resolution with a relatively long ring-down, yet we could achieve a relatively high signal-to-noise ratio using such a low-impedance backing. Yet the −6 dB bandwidth of single-element pulse-echo test reached 51%, which was within the ranges of previously used commercial transducers for ARFI imaging [8] (53% and 55% for Siemens VF7-3 and 9L4 transducers, respectively). In future work, different aperture sizes with various backing materials will be also considered to improve the pressure output meanwhile keep the high tracking performance. Moreover, the degree of distortion in the measured waveforms was distinct at each element in the array, which may be due to the uneven distribution of the acoustic impedance in the backing layer. This could be improved in the future with a more uniform backing layer by adjusting the centrifuge method during the mixing process. The center frequency of the fabricated array shifted to the lower frequency by 11% compared to that of the design, which may due to the inaccurate matching layer thickness in the fabrication process. During fabrication, the matching layer was first lapped and then attached with epoxy to the array. The thickness control could be improved by bonding and lapping the matching layer before attaching the flex circuit. It should be noticed that the aperture size of this array is much smaller compared with traditional commercial arrays. The pressure output of the newly reported array can be further improved to reach a commercial standard by increasing the element number in each Y-array and the elevational width. Compared with a fully sampled matrix which contained n×n elements, the current n×3 array can still be assembled with three 1-D flex circuit, which is much easier for the wiring process and pin map design. To make the array a more compact structure, the customed PCB board with a space-saving flex circuit will be considered in the future. A relatively high k_t would be helpful for the array to widen the frequency bandwidth. A higher tracking frequency will have a smaller PSF size which will reduce the jitter and displacement underestimation [32]. Other ways to reduce the jitter and displacement underestimation in the tracking of displacement was to use more elements for tracking. One on-going solution is using the composite material which can also improve the acoustic impedance match from the array to the tissue. Future works will investigate the different ways to reduce the tracking error.

V. Conclusions

This paper presents the design, fabrication, and characterization of a row-column array. The array has a central frequency of 3.37 MHz with an averaged bandwidth of 54%. The Y-pushing and XG-pushing method has been reported based on the row-column array. By rotating the pushing beam electronically, it can provide the same beam mapping in orthogonal directions, which has shown great potential in anisotropy measurement. With the in vitro test, the array has shown its ability in inducing the same peak displacement in orthogonal directions with an isotropic phantom. To improve the performance, a higher kt material and thicker backing layer could be applied to improve the bandwidth.

Biography

Huaiyu Wu studied mechanical engineering at Shangdong University, Jinan, China, and received the B.Sc. and M.Sc. degrees in 2013 and 2016, respectively. He is currently pursuing the Ph.D. degree in mechanical engineering at the Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, USA. His research interests focus on ultrasound transducer design for contrast imaging and intravascular sono-thrombolysis.

Md Murad Hossain earned his B.Sc and M. Sc degree in electrical engineering from the Islamic University of Technology, Bangladesh and George Mason University, USA in 2009 and 2014, respectively. He completed his Ph.D. degree in biomedical engineering at the joint Department of Biomedical Engineering at the University of North Carolina – Chapel Hill and North Carolina State University in 2019 with a focus on ultrasound elastography. He is now a post-doctoral research scientist in the Biomedical Engineering department at Columbia University. His research interests include acoustic radiation force imaging, shear wave elastography, viscoelasticity, and medical signal processing.

Howuk Kim received the B.S. and M.S. in mechanical engineering from Inha University, Incheon, South Korea, in 2007 and in 2009, respectively, and the Ph.D. degree in mechanical engineering from North Carolina State University in 2020. He was with Hyundai Kefico Corp., South Korea from 2008 to 2015, as a research engineer. He is currently working as a Postdoctoral Research Scholar at the Micro/Nano Engineering Laboratory, North Carolina State University. His broad research interests involve high temperature piezoelectric sensors, ultrasonic wave propagation, SHM/NDT technique, and biomedical ultrasound transducers.

Caterina M. Gallippi earned a B.S.E. in Electrical Engineering and a certificate in Engineering Biology from Princeton University in 1998. She completed her Ph.D. in Biomedical Engineering at Duke University in 2003 with a focus on ultrasonic imaging. She is currently a Professor and Director of Graduate Studies in the Joint Department of Biomedical Engineering at the University of North Carolina at Chapel Hill and North Carolina State University, NC, USA. Her research interests include radiation force imaging, adaptive signal filtering, multidimensional motion tracking, and magneto-motive ultrasound.

Dr. Xiaoning Jiang is a Dean F. Duncan Distinguished Professor of Mechanical and Aerospace Engineering and a University Faculty Scholar at North Carolina State University. He is also an Adjunct Professor of Biomedical Engineering at North Carolina State University and the University of North Carolina, Chapel Hill. Dr. Jiang received his BS, MS and Ph.D. degrees from Shanghai Jiao-tong University (1990), Tianjin University (1992) and Tsinghua University (1997), respectively. He received his Postdoctoral training from the Nanyang Technological University (1996-1997) and the Pennsylvania State University (1997-2001). He was the Chief Scientist and Vice President at TRS Technologies, Inc. prior to joining NC State in 2009. Dr. Jiang is the author and co-author of two books, 6 book chapters, 9 issued US Patents, 120 peer reviewed journal papers and over 100 conference papers on piezoelectric ultrasound transducers, ultrasound for medical imaging and therapy, drug delivery, ultrasound NDT/NDE, smart materials and structures and M/NEMS. Dr. Jiang is a member of the technical program committee for a few international conferences including IEEE Ultrasonics Symposium (TPC-5), SPIE Smart Structures and NDE, ASME IMECE, IEEE NANO and IEEE NMDC. He is the NanoAcoustics Technical Committee Chair for IEEE NTC, IEEE NTC Distinguished Lecturer (2018 and 2019), an editorial board member for the journal Sensors, a senior associate editor for the ASME Journal of Engineering and Science in Medical Diagnostics and Therapy, and Co-Editor-in-Chief of IEEE Nanotechnology Magazine. Dr. Jiang is an ASME Fellow.