A 4D ultrasound imaging system and method based on zoned electrodes and compressed sensing

Abstract

The application discloses a 4D ultrasonic imaging system and method based on a partition electrode and compressed sensing. A backing, a piezoelectric ceramic and a matching layer are sequentially arranged in the shell, a positive electrode of the piezoelectric ceramic serves as an acoustic radiation surface and faces a target area through the matching layer, a negative electrode of the piezoelectric ceramic is provided with the backing for absorbing reflected waves, the piezoelectric ceramic is electrically connected with a control circuit through a cable penetrating out of the shell, a spatial phase mask is arranged outside the shell and contacts a side of the matching layer away from the piezoelectric ceramic, and is used for modulating a spatial phase distribution of a transmitted acoustic wave; the acoustic radiation surface of the piezoelectric ceramic is divided into a plurality of electrode areas, each electrode area is electrically connected with the control circuit respectively, the control circuit is used for controlling independent excitation and reception of each electrode area, and the negative electrode of the piezoelectric ceramic is grounded. Compared with a traditional phased array device, the application greatly reduces the complexity and operation cost of the device, and effectively solves the problem that the prior art is difficult to rotate or move the mask to increase additional signal measurement values.

Description

Technical Field

[0001] This invention relates to the field of ultrasonic nondestructive testing technology, and in particular to a 4D ultrasonic imaging system and method based on partitioned electrodes and compression sensing. Background Technology

[0002] Ultrasonic imaging is widely used in industrial inspection and biomedicine. By moving the ultrasonic probe, an image of the cross-section of the object being inspected can be obtained. The principle is that the ultrasonic probe can excite ultrasonic pulse signals (typically with a frequency of 1–20 MHz) while simultaneously receiving the reflected echo from the target area. By analyzing the time between pulse emission and echo, the location of the imaged target can be estimated. Common ultrasonic phased array probes are mainly made of piezoelectric ceramics arranged in an array. When a voltage is applied to the material, the piezoelectric crystals vibrate and generate short ultrasonic pulses of several cycles. When the relevant array elements are activated with an appropriate delay, the emitted sound waves can propagate in a narrow beam, thus focusing on the point of interest. Through these focused beams, the entire volume can be scanned. When receiving signals, a similar delay can be used to selectively amplify the signal from scatterers on the beam of interest. This focusing technique, which uses delayed signals in both transmission and / or reception, is called beamforming and has been extensively studied and applied for many years.

[0003] 4D imaging technology (real-time 3D dynamic imaging), as an innovative achievement integrating multiple disciplines, has become a cutting-edge research direction. In the field of industrial non-destructive testing, 4D ultrasound has been applied to the real-time 3D localization of internal defects in materials. In the biomedical field, 4D ultrasound combined with Doppler hemodynamic analysis makes it possible to perform kinematic analysis of heart valves and quantitative evaluation of tumor angiogenesis. However, current real-time 3D dynamic imaging technologies rely on beamforming optimization of phased array ultrasound probes and iterative parallel computing architectures (post-processing the individual signals received by each array element after analog-to-digital conversion). This often requires dozens of sensor elements, complex hardware, and cumbersome signal processing logic. Recent discoveries in compressed sensing suggest that signal structures can be used to alleviate the burden imposed by traditional phased array ultrasound imaging requirements.

[0004] The core of compressed sensing lies in disrupting the phase uniformity of sound waves during transmission and reception, thereby projecting 3D object information onto a set of incoherent basis functions. Specifically, a phase-encoded mask is placed in front of the ultrasonic probe, altering the time it takes for the ultrasonic waves to propagate into the target medium. This disrupts the phase uniformity of the sound waves, generating complex spatiotemporal interference patterns. Therefore, for each location (each pixel) in 3D space, there will be a unique ultrasonic signal from that location. The echo signals from all pixels are collected by the ultrasonic probe after passing through the mask, resulting in a compressed measurement. The real scene signal can then be reconstructed through the inverse kinematics of this compressed measurement. However, due to limitations in temporal bandwidth and the specific phase distribution of the mask, it cannot be guaranteed that the echo signals from all pixels are uncorrelated. To address this, current researchers use mask rotation or movement to add additional signal measurements, but this is difficult to implement and execute in practical detection. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a 4D ultrasound imaging system and method based on partitioned electrodes and compression sensing.

[0006] The technical solution adopted in this invention is:

[0007] The present invention includes a housing, a cable, a backing, a piezoelectric ceramic, a matching layer, a spatial phase mask, and a control circuit. The backing, piezoelectric ceramic, and matching layer are sequentially installed inside the housing. The positive electrode surface of the piezoelectric ceramic serves as the sound radiation surface and faces the target area. The negative electrode surface of the piezoelectric ceramic is covered with a backing for absorbing reflected waves. The piezoelectric ceramic is electrically connected to the control circuit through a cable passing through the housing. The spatial phase mask is installed outside the housing and contacts the side of the matching layer away from the piezoelectric ceramic, and is used to modulate the spatial phase distribution of the transmitted sound waves.

[0008] The acoustic radiation surface of the piezoelectric ceramic is divided into L×L electrode regions, each of which is electrically connected to a control circuit. The control circuit is used to control the independent excitation and reception of each electrode region, and the negative electrode of the piezoelectric ceramic is grounded.

[0009] It also includes a compression sensing linear module, which is electrically connected to the control circuit and is used to reconstruct a 4D ultrasound image of the target area;

[0010] It also includes a locking knob, which is located on the upper surface of the housing and is used to secure cables passing through the housing.

[0011] The spatial phase mask has the same dimensions as the piezoelectric ceramic and is divided into L×L mask regions. Each mask region has a different thickness, and the thickness relationship between the mask regions is set to satisfy the condition that the spatial phase difference between the transmitted waves from any two mask regions is greater than or equal to 0 and less than or equal to π. The spatial phase difference between the transmitted waves from two mask regions is set according to the following formula:

[0012] φ = arg(P1 - P2)

[0013] Where φ represents the spatial phase difference of the transmitted wave, arg() represents the complex argument, and P1 and P2 are the complex representations of the transmitted waves in the two mask regions, respectively.

[0014] Both the backing and the matching layer are made of epoxy resin mixed with tungsten powder. The mass ratio of tungsten powder to epoxy resin in the backing is (2-4):1, and the mass ratio of tungsten powder to epoxy resin in the matching layer is 1:(3-5). The thickness of the matching layer is equal to one-quarter of the wavelength of the center frequency of the piezoelectric ceramic.

[0015] The acoustic radiation surface of the piezoelectric ceramic is divided into several electrode regions in any of the following ways: laser cutting, dicing machine cutting, or sputtering deposition of a metal electrode layer on the piezoelectric ceramic.

[0016] The acoustic radiation surface of the piezoelectric ceramic is provided with a multi-electrode flexible circuit board or a needle-shaped electrode.

[0017] The 4D ultrasound imaging method includes the following steps:

[0018] S1. Perform wavefield scanning on the target region to obtain the system matrix, and then establish a linear system model based on the system matrix;

[0019] S2. Randomly control the electrode area of ​​the piezoelectric ceramic to make the piezoelectric ceramic emit ultrasonic waves. The ultrasonic waves are modulated by a spatial phase mask and reflected by the target area to obtain an echo signal.

[0020] S3. The echo signal passes through the spatial phase mask and is received by the piezoelectric ceramic to form a compressed signal, which is then recorded by the control circuit and transmitted to the linear system model.

[0021] S4. Repeat steps S2 and S3 multiple times, randomly controlling the electrode region of the piezoelectric ceramic each time, until the linear system model receives S compression signals. The linear system model reconstructs the target region based on the S compression signals to obtain a 4D ultrasonic image.

[0022] The wavefield scanning is performed according to the following method:

[0023] A hydrophone is used to collect ultrasonic waves modulated by a spatial phase mask in a plane perpendicular to the direction of ultrasonic wave propagation. The ultrasonic waves modulated by the spatial phase mask are then processed to obtain wavefield data. Subsequently, the wavefield data is propagated to planes at different propagation distances using the angular spectrum method to obtain wavefield data at different propagation distances. Finally, a system matrix is ​​constructed based on the wavefield data at different propagation distances.

[0024] The linear system model reconstructs the 4D ultrasound image of the target region according to the following formula:

[0025] u = Hv + n

[0026] Where u represents the compressed signal, v represents the reconstructed 4D ultrasound image signal, n represents the noise signal, and H is the system matrix.

[0027] The beneficial effects of this invention are:

[0028] This invention achieves dynamic three-dimensional imaging using a single ultrasonic transducer with only electrode partitions and a phase mask. Compared with complex and expensive phased array systems and other signal acquisition methods, it significantly reduces equipment costs and greatly increases the portability of the equipment. By randomly activating different areas each time to acquire multiple sets of "multiple transmit and one receive" signals, it effectively solves the problem that existing technologies cannot easily rotate or move the mask to increase additional signal measurement values. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the assembly of the 4D ultrasound imaging system.

[0030] Figure 2 This is an exploded view of the 4D ultrasound imaging system.

[0031] Figure 3 This is a schematic diagram of the electrode wiring of the 4D ultrasound imaging system.

[0032] Figure 4 This is a schematic diagram of the spatial phase mask structure of the 4D ultrasound imaging system.

[0033] Figure 5 This is a schematic diagram of the compressed sensing linear module of the 4D ultrasound imaging method.

[0034] Figure 6 This is a flowchart of the working process of the 4D ultrasound imaging method.

[0035] Figure 7 This is a schematic diagram of a simulated defect in Embodiment 1 of the 4D ultrasound imaging method.

[0036] Figure 8 This is an image showing the imaging result of Embodiment 1 of the 4D ultrasound imaging method.

[0037] Figure 9 This is a schematic diagram of a simulated defect in Embodiment 2 of the 4D ultrasound imaging method.

[0038] Figure 10 This is an image showing the imaging result of Embodiment 2 of the 4D ultrasound imaging method.

[0039] Reference numerals: 1 housing, 2 locking knob, 3 cable, 4 backing, 5 piezoelectric ceramic, 6 matching layer, 7 spatial phase mask. Detailed Implementation

[0040] The present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments of the present invention include, but are not limited to, the following embodiments.

[0041] In the description of this invention, it should be understood that the terms "hollow", "top", "bottom", "inner side", "outer side", etc., which indicate orientation or positional relationship, are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the components or elements referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as limiting this invention.

[0042] like Figure 1 and Figure 2 As shown, the 4D ultrasound imaging system includes a housing 1, a cable 3, a backing 4, a piezoelectric ceramic 5, a matching layer 6, a spatial phase mask 7, and a control circuit. The backing 4, the piezoelectric ceramic 5, and the matching layer 6 are sequentially disposed inside the housing 1. The positive electrode surface of the piezoelectric ceramic 5 serves as the sound radiation surface and emits ultrasonic waves toward the target area. The negative electrode surface of the piezoelectric ceramic 5 is covered with the backing 4 for absorbing reflected waves. The piezoelectric ceramic 5 is electrically connected to the control circuit through the cable 3 passing through the housing 1. The spatial phase mask 7 is installed outside the housing 1 and contacts the side of the matching layer 6 away from the piezoelectric ceramic 5, and is used to modulate the spatial phase distribution of the transmitted sound waves.

[0043] The acoustic radiation surface of the piezoelectric ceramic 5 is divided into L×L electrode regions. Each electrode region is electrically connected to the control circuit. The control circuit is used to control the independent excitation and reception of ultrasonic waves in each electrode region. Therefore, the control circuit can achieve simultaneous excitation of a single electrode region or multiple electrode regions. The negative electrode surface of the piezoelectric ceramic 5 is grounded.

[0044] The backing 4 absorbs the reflected wave, which is the reflected wave of the ultrasonic wave emitted from the piezoelectric ceramic 5 when it hits the backing 4. The backing 4 prevents the wave from returning to the piezoelectric ceramic 5.

[0045] Electrode layers are disposed on both sides of the piezoelectric ceramic 5, and the electrode layers are electrically connected to the drive output terminal of the control circuit via cable 3. The control circuit can be connected to an oscilloscope to observe and analyze the compressed signal.

[0046] The acoustic radiation surface and control circuit of the piezoelectric ceramic 5 are different from those of the phased array ultrasonic transducer. It does not have complex logic circuits and does not control the excitation timing of each array element through electronic delay. The electrode area achieves multi-array element coordinated control such as one-to-all transmission and multiple-to-all transmission through the control circuit. The structure is simpler than the existing 4D ultrasonic imaging system.

[0047] Cable 3 includes several lines. Each electrode area is connected to the control circuit using a separate line. The negative electrode of the piezoelectric ceramic 5 is connected to the ground terminal in the control circuit using a line. Therefore, the number of lines in cable 3 is equal to the number of electrode areas plus one.

[0048] The 4D ultrasound imaging system also includes a compression sensing linear module, which is electrically connected to the control circuit and is used to receive the compression signal transmitted by the control circuit to reconstruct the 4D ultrasound image of the target area.

[0049] It also includes a locking knob 2, which is located on the upper surface of the housing 1 and is used to fix the cable 3 that passes through the housing 1.

[0050] The spatial phase mask needs to be divided into regions according to the number of electrode partitions. For one region of the mask, the modulated transmitted sound field of an incident plane wave with frequency f can be expressed as follows: v a For the sound velocity of the mask material, k a The wave velocity of the mask material is represented by k. a =2πf / v a The wavenumber of the material is given, and the transmitted sound field differs at units of different thicknesses (h).

[0051] like Figure 4 As shown, the spatial phase mask 7 is the same size as the piezoelectric ceramic 5, and the spatial phase mask 7 is divided into L×L mask regions in the same way as the acoustic radiation surface electrode region of the piezoelectric ceramic 5. The thickness of each mask region is different, and the thickness relationship between each mask region is set to satisfy that the spatial phase difference between the transmitted waves of any two mask regions in the spatial phase mask 7 is greater than or equal to 0 and less than or equal to π. The spatial phase difference of the transmitted waves is set according to the following formula:

[0052] φ = arg(P1 - P2)

[0053] Where φ represents the spatial phase difference of the transmitted wave, arg() represents the complex argument, and P1 and P2 represent the complex representations of the transmitted waves in the two mask regions, respectively.

[0054] z = r(cosθ + i sinθ), where r is the modulus of z, r = |z|; θ is the complex argument of z, θ = arg(z).

[0055] The transmitted wave is a wave modulated by the spatial phase mask 7, and the spatial phase difference of the transmitted wave is the phase difference between the two transmitted waves.

[0056] Once the spatial phase mask 7 is fabricated, it can be directly mounted in front of the partitioned electrode single probe. The spatial phase of the mask can be randomly distributed. The spatial phase mask 7 can be made by subtractive manufacturing, 3D printing, or using a phase change polymer material with varying sound velocity on the same plate, but is not limited to this method. In this case, the phase uniformity of the propagating wave is disrupted, thus forming a random deterministic interference pattern in the medium. Generally, additive or subtractive manufacturing methods such as 3D printing and area drilling can be used to prepare spatial phase grinding with randomly heighted regions. It is important to note that to ensure good sound wave transmission efficiency, the selected material must have a small sound attenuation coefficient and an acoustic impedance that is well-matched to water.

[0057] Specifically, the spatial phase mask 7 is divided into several mask regions in the same way as the electrode regions of the acoustic radiation surface of the piezoelectric ceramic 5. That is, the number and size of the mask regions divided by the spatial phase mask 7 are the same as the number of electrode regions divided by the acoustic radiation surface of the piezoelectric ceramic 5, and the mask regions divided by the spatial phase mask 7 correspond one-to-one with the electrode regions divided by the acoustic radiation surface of the piezoelectric ceramic 5.

[0058] Both the backing 4 and the matching layer 6 are made of epoxy resin mixed with tungsten powder. The mass ratio of tungsten powder to epoxy resin in the backing 4 is (2-4):1, and the mass ratio of tungsten powder to epoxy resin in the matching layer 6 is 1:(3-5). The thickness of the matching layer 6 is equal to one-quarter of the wavelength of the center frequency of the piezoelectric ceramic 5.

[0059] A matching layer of uniform thickness is applied evenly using a spin coating method. Generally, the preferred thickness of the matching layer is one-quarter of the wavelength of the piezoelectric crystal's center frequency in water.

[0060] The acoustic radiation surface of the piezoelectric ceramic 5 is divided into several electrode regions in any of the following ways: laser cutting, dicing machine cutting, or sputtering deposition of a metal electrode layer on the piezoelectric ceramic 5, i.e., adding a partition mask when sputtering the original electrode metal coating.

[0061] The piezoelectric ceramic 5 requires the use of a laser or other dicing equipment to partition one side of its electrodes. Each partitioned electrode must be connected using a separate circuit, while the electrodes on the other side of the wafer can share a single circuit. Generally, the positive electrode of the piezoelectric ceramic 5 is preferred for electrode partitioning, while the negative electrode can be connected to a common ground circuit. It is important to note that to ensure that each partitioned area can be individually excited, the surface electrodes of the piezoelectric ceramic 5 can be divided at a certain depth.

[0062] like Figure 3The diagram shows the control circuit for a single probe with partitioned electrodes. Taking a single-probe transducer with 64 electrode partitions as an example, the positive lead of each probe region is individually connected to an analog electronic switch on the controller, while the negative lead is connected to ground. The analog switch of the controller can independently and quickly control the signal switching on and off. Furthermore, this enables simultaneous excitation of a single probe in a single region or multiple regions, while also providing a BNC interface for connecting a signal generator to receive and transmit echo signals.

[0063] The acoustic radiation surface of the piezoelectric ceramic 5 is provided with a multi-electrode flexible circuit board or needle-shaped electrode, and the negative electrode of the piezoelectric ceramic 5 is grounded by a ground wire.

[0064] The 4D ultrasound imaging system follows the core concept of compressed imaging: projecting object / image information onto a single measurement value using a set of incoherent functions, and then decoding the retrieved measurement value to form an image. Based on this, the scheme includes a partitioned electrode single probe that modulates emitted sound waves using a random mask, thereby disrupting the phase of the wave field. The phase delay mask ensures that each region in the image is uniquely identifiable in the compressed measurement. By randomly activating different regions each time and acquiring multiple sets of "multiple transmit, one receive" signals, dynamic three-dimensional imaging can be achieved through real-time signal post-processing.

[0065] like Figure 6 As shown, the 4D ultrasound imaging method includes the following steps:

[0066] S1. Use a hydrophone to scan the wave field of the target area to obtain the system matrix H, and then establish a linear system model based on the system matrix H.

[0067] The 4D ultrasound imaging system can be applied to underwater ultrasound detection and medical ultrasound, etc. The target area refers to the spatial area that needs to be 4D imaged, which can be underwater objects or human tissues.

[0068] Since the acoustic impedance of the human body and water is similar, the system matrix obtained in water can still be used in medical ultrasound applications.

[0069] S2. Randomly control the electrode area of ​​the piezoelectric ceramic 5 so that the piezoelectric ceramic 5 emits ultrasonic waves. After the ultrasonic waves are modulated by the spatial phase mask 7, they are reflected by the target area to obtain the echo signal.

[0070] S3. Due to the phase modulation effect of the mask, the reflected signals from different spatial locations will be superimposed in an incoherent manner. The echo signal passes through the spatial phase mask 7 and is received by the piezoelectric ceramic 5 to form a compressed signal, which is then recorded by the control circuit and transmitted to the linear system model.

[0071] S4. Repeat steps S2 and S3 multiple times, and randomly control the electrode area of ​​the piezoelectric ceramic 5 each time until the linear system model receives S compression signals, so that the linear system model has enough information to reconstruct. The linear system model reconstructs the target area based on the S compression signals to obtain the 4D ultrasound image of the target area.

[0072] Wavefield scanning is performed using the following method:

[0073] A small hydrophone is used to collect ultrasonic waves modulated by the spatial phase mask 7 in a plane near the surface of the spatial phase mask 7 and perpendicular to the direction of ultrasonic wave propagation. The ultrasonic waves modulated by the spatial phase mask 7 are then processed to obtain wavefield data. Subsequently, the wavefield data is propagated to planes at different propagation distances using the angular spectrum method to obtain wavefield data at different propagation distances. Then, a system matrix is ​​constructed based on the wavefield data at different propagation distances.

[0074] In wavefield scanning, the planes are all perpendicular to the direction of ultrasonic wave propagation.

[0075] Before imaging, the block matrix H needs to be acquired to understand the entire spatiotemporal wavefield. Furthermore, a small hydrophone can be used to acquire pulse signals in a plane close to the mask surface and perpendicular to the ultrasonic propagation axis. Then, the recorded wavefield is propagated to any plane parallel to the recording plane using angular spectroscopy.

[0076] Specifically, the transmitted sound field collected using a hydrophone is as follows:

[0077] p0(x,y,z=0)=e (iφ0)

[0078] Where p0(x,y,z=0) represents the initial transmitted sound field distribution on the z=0 plane, and represents the state of the sound field at the z=0 plane.

[0079] The spectrum of this plane can then be calculated using Fourier transform:

[0080] P0(k x ,k y ,z=0)=∫∫p0(x,y,z=0)dxdy

[0081] Wherein, P0(k) x ,k y (z=0) is the frequency spectrum of the transmitted sound field, represented in the frequency domain on the z=0 plane, k x k y It is the spatial wavenumber of the spectrum in the x and y directions.

[0082] Then, the transmitted sound field propagates along the positive z-axis, and the spectrum is multiplied by the phase conversion factor:

[0083] H(kx ,k y ,z=l)=e (ikzl)

[0084] The spectral distribution on any plane z = l can then be obtained:

[0085] P ’ (k x ,k y ,z=l)=P0(k x ,k y ,z=0)*H(k x ,k y ,z=l)

[0086] The sound field distribution on this plane can then be obtained through inverse Fourier transform:

[0087] p ’ (x,y,z=l)=∫∫P ’ (k x ,k y ,z=l)dk x dk y

[0088] Where p0 is the initial sound field collected by the hydrophone when z = 0, φ0 is the phase of the initial sound field collected by the hydrophone when z = 0, and H is the phase conversion factor used to adjust the spectrum after propagation.

[0089] According to the reciprocity theorem, the pulse echo signal can be obtained by self-convolving the propagating hydrophone signal itself. Therefore, the pulse echo ultrasonic signal at each point in 3D space can be calculated, and after multiple random switching, the acquired and calculated signals can fill column H.

[0090] The linear system model reconstructs the 4D ultrasound image of the target region according to the following formula:

[0091] u = Hv + n

[0092] Where u represents the compressed signal, i.e. the actual detection signal, v represents the reconstructed 4D ultrasound image signal, i.e. the unknown real image, n represents the noise signal, and H is the system matrix, i.e. the pulse echo ultrasound signal at each point in the original spatiotemporal wave field.

[0093] Specifically, the system matrix H in the linear system model needs to be obtained in advance. After the linear system model is established, the image reconstruction can be completed by inversely solving the vector v based on the compressed signal u. By randomly activating different regions each time to collect multiple sets of "multiple transmissions and full reception" signals, multiple random excitations can eliminate the ambiguity of the reconstructed image and greatly improve the signal-to-noise ratio of the detected target, thereby achieving efficient three-dimensional imaging of the detected target. Based on the rapid switching characteristics of several electrode regions on the acoustic radiation surface, multiple continuous random acquisitions and processing can be performed within a certain period of time. Combined with probe movement and acquisition, dynamic three-dimensional imaging can be achieved.

[0094] Unlike the standard beamforming technique in traditional phased array ultrasound, incoherent wavefront emission can be achieved by randomly exciting several regions of a single probe. Correspondingly, the echo is also received by the piezoelectric ceramic 5 after mask modulation. Therefore, the received compressed signal is considered as a linear combination of point scatterer echo signals, thus establishing a linear imaging model. The 3D scene contains N pixels, u is the actual detection signal (i.e., the received compressed signal), and v represents the reconstructed 4D ultrasound image signal (i.e., the unknown real image). Figure 5 As shown, the linear imaging model is given by the following equation: u = Hv + n, where u is an M-dimensional column vector, H is an M×N-dimensional block matrix determined by the medium of the target region, v is an N-dimensional column vector, and n is an M-dimensional column vector whose samples represent zero-mean additive white Gaussian noise with a noise variance of σ. e 2 .

[0095] Due to limitations in time bandwidth and mask thickness distribution, it cannot be guaranteed that the echo signals from all regions are uncorrelated. To introduce more diversity between regions, S groups of different randomly excited echoes are selected to obtain additional measurements containing new information. The structure of the block matrix H is as follows:

[0096] H = [H1, H2, H3…H] S ] T

[0097] H r =[h r,1 ,h r,2 …,h r,N ]

[0098] Where H r Let M be a K×N matrix, r∈{1,…,S}, representing the pulse echo response time series of length K at pixel n∈{1…N} in the image excited by each random switching of different array elements. Therefore, M=KS, and u is also a column vector combination of S K-dimensional column vectors.

[0099] After establishing the linear imaging model, image reconstruction requires solving for the inverse vector v. Comparing the measured value M with the number of pixels N to be reconstructed, the linear system is underdetermined (M < N) at this time, and there are infinitely many possible solutions Satisfy Here we choose to use the minimized cost function to find The optimal solution of: Matrix inversion can be implemented and calculated by the preconditioned conjugate gradient method (PCG), and finally we can solve To characterize the accuracy of image reconstruction, the peak signal-to-noise ratio is defined as:

[0100]

[0101] The mean square error is:

[0102]

[0103] Among them, PSNR represents the peak signal-to-noise ratio, and MSE represents the mean square error.

[0104] Example 1:

[0105] First, choose a piezoelectric ceramic sheet with a frequency of 5 MHz. Use a laser marking machine to divide the electrodes on one side into 16x16 areas. After the division, use a flexible FPC circuit board with 256 contacts to weld the electrodes. At the same time, weld the negative electrode on the other side to the common negative connection of the circuit board. Then fill the backing and assemble the piezoelectric ceramic sheet into the housing, and spin-coat the matching layer, controlling the thickness of the matching layer to be about one-fourth of the wavelength of this frequency in water, that is, about 75 um.

[0106] Furthermore, use 3D printing technology to prepare a phase delay mask (between 0 and h) with 256 randomly distributed high and low levels. The sound speed of the resin selected for 3D printing is 2350 m / s, and the sound speed of water for comparison is 1450 m / s. When the thickness h is 0.85 mm, the phase delay of the incident wave can be modulated to π. Subsequently, paste the delay mask on the surface of the probe using a common water-based ultrasonic couplant.

[0107] Furthermore, after completing the assembly of the partitioned electrode single-probe system, immerse the probe in water at this time and use a hydrophone to collect the wave field near the mask, and record the transmitted sound field at this time as p0(x,y,z = 0). Use the angular spectrum method to obtain the sound field on any plane as P ’ (k x ,k y ,z = l). Furthermore, we perform self-convolution on the collected hydrophone signal itself to obtain the pulse echo signal. Randomly switch the transmission of 100 regions among them multiple times and use all regions for collection, and then fill the block matrix H in the linear system.

[0108] Such as Figure 7A simulated target area is set up, and then a probe system is used to scan the target area. Different areas are switched multiple times for transmission, and echoes are received from all areas. Solving the linear equations completes the image reconstruction of the target area. For example... Figure 8 After switching regions and acquiring data multiple times, the peak signal-to-noise ratio of the detected image was 17.2123 dB, which clearly identified the simulated defects. Dynamic 3D imaging can be completed by moving the image to other regions in real time.

[0109] Example 2:

[0110] First, a 5MHz piezoelectric ceramic sheet was selected. A laser marking machine was used to divide one side of the electrode into a 12x12 area. After division, a flexible FPC circuit board with 144 contacts was used for electrode soldering. Simultaneously, the negative electrode on the other side was soldered to the common negative terminal connection of the circuit board. Next, the piezoelectric ceramic sheet was assembled into the housing, a backing was filled, and a matching layer was spin-coated. The thickness of the matching layer was controlled to be one-quarter of the wavelength of this frequency in water, approximately 75µm.

[0111] Furthermore, 144 phase delay masks (between 0 and h) with unevenly distributed depths were machined on the flat plate using a CNC machining center. The selected resin material had a sound velocity of 2350 m / s, compared to 1450 m / s for water. With a thickness h of 0.85 mm, the incident wave phase delay could be modulated to π. The delay masks were then adhered to the probe surface using a common water-based ultrasonic coupling agent.

[0112] Furthermore, after assembling the partitioned electrode single-probe system, the probe is immersed in water, and a hydrophone is used to collect the wave field near the mask. The transmitted sound field at this time is recorded as p0 (x,y,z=0). The sound field on any plane is obtained using the angular spectrum method as P. ’ (k x ,k y (z = l). Furthermore, we perform self-convolution on the acquired hydrophone signal itself to obtain the pulse echo signal. We repeatedly and randomly switch between 50 regions for transmission and acquire data from all regions, subsequently filling the block matrix H in the linear system.

[0113] like Figure 9 A simulated target area is set up, and then a probe system is used to scan the target area. Different areas are switched multiple times for transmission, and echoes are received from all areas. Solving the linear equations completes the image reconstruction of the target area. For example... Figure 10 After switching regions and acquiring data multiple times, the peak signal-to-noise ratio of the detected image was 14.1282 dB, which clearly identified the simulated defects. Dynamic 3D imaging can be completed by moving the image to other regions in real time.

[0114] The innovation of this invention lies in:

[0115] 1) By dividing the acoustic radiation surface of the piezoelectric ceramic 5 into L×L electrode regions, and using a control circuit to control the independent excitation and reception of each electrode region, the acoustic radiation surface of the piezoelectric ceramic 5 and its control circuit in this structure differ from those of a phased array ultrasonic transducer. It does not possess complex logic circuits, and does not control the excitation timing of each array element through electronic delay. Instead, the electrode regions achieve multi-element coordinated control, such as single-emission and multi-emission / multi-emission, through the control circuit. The structure is simpler than that of existing 4D ultrasonic imaging systems.

[0116] 2) At present, researchers increase additional signal measurement values ​​by rotating or moving the mask. However, this invention divides the spatial phase mask 7 into L×L mask regions with different thicknesses. It is used in conjunction with the electrode region of the acoustic radiation surface of the piezoelectric ceramic 5. By randomly activating different regions each time, multiple sets of "multiple transmit and one receive" signals are collected, which effectively solves the problem that it is difficult to rotate or move the mask to increase additional signal measurement values ​​in the existing technology.

[0117] 3) The present invention proposes that the 4D ultrasound image of the target area can be reconstructed using the following formula: u=Hv+n, which is simpler and can clearly identify simulation defects compared to the existing methods for reconstructing the 4D ultrasound image of the target area.

[0118] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. A 4D ultrasound imaging system based on partitioned electrodes and compression sensing for acquiring four-dimensional ultrasound images of a target area, characterized in that: it includes a shell (1), a cable (3), a backing (4), a piezoelectric ceramic (5), a matching layer (6), a spatial phase mask (7), and a control circuit. The backing (4), the piezoelectric ceramic (5), and the matching layer (6) are sequentially installed inside the shell (1). The positive electrode surface of the piezoelectric ceramic (5) serves as the sound radiation surface and faces the target area. The negative electrode surface of the piezoelectric ceramic (5) is provided with a backing (4) for absorbing reflected waves. The piezoelectric ceramic (5) is electrically connected to the control circuit through the cable (3) passing through the shell (1). The spatial phase mask (7) is installed outside the shell (1) and contacts the side of the matching layer (6) away from the piezoelectric ceramic (5) for modulating the spatial phase distribution of transmitted sound waves. The acoustic radiation surface of the piezoelectric ceramic (5) is divided into L×L electrode regions. Each electrode region is electrically connected to the control circuit. The control circuit is used to control the independent excitation and reception of each electrode region. The negative electrode of the piezoelectric ceramic (5) is grounded. The spatial phase mask (7) has the same size as the piezoelectric ceramic (5) and is divided into L×L mask regions. The thickness of each mask region is different, and the thickness relationship between each mask region is set to satisfy that the spatial phase difference between the transmitted waves of any two mask regions in the spatial phase mask (7) is greater than or equal to 0 and less than or equal to π. The spatial phase difference between the transmitted waves of the two mask regions is set according to the following formula: in, Let represent the spatial phase difference of the transmitted wave, arg() represent the complex argument, and P1 and P2 represent the complex representations of the transmitted waves in the two mask regions, respectively.

2. The 4D ultrasound imaging system based on partitioned electrodes and compression sensing according to claim 1, characterized in that: It also includes a compression sensing linear module, which is electrically connected to the control circuit and is used to reconstruct a 4D ultrasound image of the target area; It also includes a locking knob (2), which is located on the upper surface of the housing (1) and is used to fix the cable (3) passing through the housing (1).

3. The 4D ultrasound imaging system based on partitioned electrodes and compression sensing according to claim 1, characterized in that: Both the backing (4) and the matching layer (6) are made of epoxy resin mixed with tungsten powder. The mass ratio of tungsten powder to epoxy resin in the backing (4) is (2-4):1, and the mass ratio of tungsten powder to epoxy resin in the matching layer (6) is 1:(3-5). The thickness of the matching layer (6) is equal to one-quarter of the center frequency wavelength of the piezoelectric ceramic (5).

4. A 4D ultrasound imaging system based on partitioned electrodes and compression sensing according to claim 1, characterized in that: The acoustic radiation surface of the piezoelectric ceramic (5) is divided into several electrode regions in any of the following ways: laser cutting, dicing machine cutting, or sputtering deposition of a metal electrode layer on the piezoelectric ceramic (5).

5. A 4D ultrasound imaging system based on partitioned electrodes and compression sensing according to claim 1, characterized in that: The acoustic radiation surface of the piezoelectric ceramic (5) is provided with a multi-electrode flexible circuit board or a needle electrode.

6. A 4D ultrasound imaging method applied to a 4D ultrasound imaging system based on partitioned electrodes and compression sensing as described in any one of claims 1-5, characterized in that, Includes the following steps: S1. Perform wavefield scanning on the target region to obtain the system matrix, and then establish a linear system model based on the system matrix; S2. Randomly control the electrode area of ​​the piezoelectric ceramic (5) so that the piezoelectric ceramic (5) emits ultrasonic waves. After the ultrasonic waves are modulated by the spatial phase mask (7), they are reflected by the target area to obtain the echo signal. S3. The echo signal passes through the spatial phase mask (7) and is received by the piezoelectric ceramic (5) to form a compressed signal, which is then recorded by the control circuit and transmitted to the linear system model. S4. Repeat steps S2 and S3 multiple times, and randomly control the electrode area of ​​the piezoelectric ceramic (5) each time until the linear system model receives S compression signals. The linear system model reconstructs the target area based on the S compression signals to obtain the 4D ultrasound image of the target area.

7. The 4D ultrasound imaging method of a 4D ultrasound imaging system based on partitioned electrodes and compression sensing according to claim 6, characterized in that: The wavefield scanning is performed according to the following method: A hydrophone is used to collect ultrasonic waves modulated by a spatial phase mask (7) in a plane perpendicular to the direction of ultrasonic wave propagation. The ultrasonic waves modulated by the spatial phase mask (7) are then processed to obtain wave field data. Subsequently, the wave field data is propagated to planes with different propagation distances using the angular spectrum method to obtain wave field data with different propagation distances. Then, a system matrix is ​​constructed based on the wave field data with different propagation distances.

8. The 4D ultrasound imaging method of a 4D ultrasound imaging system based on partitioned electrodes and compression sensing according to claim 6, characterized in that: The linear system model reconstructs the 4D ultrasound image of the target region according to the following formula: u=Hv+n Where u represents the compressed signal, v represents the reconstructed 4D ultrasound image signal, n represents the noise signal, and H is the system matrix.

Images