A Real-Time Three-Dimensional Ultrasonic Intelligent Imaging Method and System Based on Row and Column Addressing Array

Abstract

This invention relates to the field of ultrasound imaging technology, and provides a real-time three-dimensional ultrasound intelligent imaging method and system based on a row-column addressable array. The method involves transmitting ultrasonic waves to a target medium via a row-column addressable ultrasound transducer array and receiving the echo signals reflected from the target medium. Raw three-dimensional ultrasound data is acquired based on these echo signals. The raw three-dimensional ultrasound data is preprocessed and transformed from the spatial domain to the frequency-wavenumber domain. The vertical wavenumber of the current depth layer is calculated. A phase-shifting operator is constructed based on a set depth step size, and a recursive evolution strategy is used to obtain the predicted wavefield for each depth layer. At each recursive depth step size, the received wavefield of the current depth layer is multiplied by the complex conjugate of the corresponding source wavefield, and integrated with respect to the angular frequency. The integration result is used as an image slice at the current depth. The image slice is transformed back to the spatial domain to generate a three-dimensional high-resolution ultrasound image of the target medium.

Description

Technical Field

[0001] This invention relates to the field of ultrasound imaging technology, and in particular to a real-time three-dimensional intelligent ultrasound imaging method and system based on row and column addressing array for layered complex media. Background Technology

[0002] With the deepening deployment in high-end equipment manufacturing (such as aerospace composite materials and new energy batteries) and next-generation medical imaging, there is an urgent need for three-dimensional, high-precision, and non-destructive visualization of the internal structure of complex targets. Traditional two-dimensional ultrasound slice imaging, lacking spatial topological information in the depth direction, is no longer sufficient to meet the standards of modern precision engineering for quantitative defect assessment. Therefore, three-dimensional ultrasound volume imaging technology based on array transducers has become a research hotspot in the current interdisciplinary field. In the process of promoting the engineering of three-dimensional ultrasound imaging, there is an irreconcilable contradiction between the "curse of dimensionality" at the hardware level and the acquisition of full-aperture information. To obtain a complete three-dimensional sound field, the ideal physical carrier is a two-dimensional fully phased array matrix. However, for an N×N array, the number of front-end transceiver channels explodes exponentially. This hardware architecture not only leads to high equipment manufacturing costs and extremely difficult cable integration, but also easily causes system thermal runaway, making it difficult to deploy on a large scale in industrial fields or compact medical devices. Existing technologies have attempted to reduce dimensionality by introducing sparse arrays, but the undersampling mechanism, which sacrifices physical detection area, inevitably leads to a deterioration in image signal-to-noise ratio, easily causing missed detections of minute delaminations in composite materials or early lesions. In recent years, Row-and-Column Addressing (RCA) array technology has offered a highly valuable engineering solution. It decouples a two-dimensional array into orthogonal transmit rows and receive columns, reducing the number of system channels from N... 2Compression down to 2N while retaining complete physical aperture coverage provides an excellent hardware foundation for high-quality 3D imaging. However, even so, in complex layered media, 3D imaging still encounters a serious "contradiction between computing power and accuracy" at the algorithm level. In most practical applications (such as penetrating multi-layered carbon fiber plates for flaw detection, or penetrating the skull for brain imaging), sound waves propagate between layers with different acoustic impedances, resulting in strong refraction and path distortion according to Snell's law. If the medium is assumed to have a single uniform sound velocity, it will lead to severe misalignment and defocusing of target features. To eliminate the distortion caused by layered media, the industry currently relies on the Delay-Sum (DAS) algorithm in conjunction with spatial ray tracing technology to compensate for travel time errors. However, the fatal problem is that high-precision ray tracing has a huge computational overhead in 3D voxel space; if the complex cross-transmitter geometry model of the RCA array is to be matched, the time required for point-by-point iterative solution is often in the hours. This offline computing mode completely blocks the application of this technology in scenarios with extremely high time sensitivity, such as "online real-time detection" or "intraoperative real-time guidance". To overcome the computational bottleneck of time-domain algorithms, frequency-domain offset technology based on the wave equation has become a new breakthrough. Utilizing the Fast Fourier Transform (FFT) for wavefield extrapolation in the frequency-wavenumber (fk) domain can transform complex geometric tracking into efficient algebraic operations, thereby achieving orders-of-magnitude computational acceleration. However, a review of current technologies reveals that most existing high-efficiency frequency-domain imaging frameworks remain at the two-dimensional cross-sectional stage; even the few studies involving three-dimensional RCA arrays are almost entirely based on the stringent assumption of a "homogeneous and non-refractive medium." In summary, the industry has yet to propose a complete technical architecture that simultaneously leverages the low-cost hardware advantages of RCA arrays, overcomes the refractive distortion of complex multi-layered media, and possesses ultra-high computational efficiency to achieve truly "real-time" three-dimensional volumetric data reconstruction. How to break through this technological barrier is a core engineering problem urgently needing to be solved in this field. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art and provide a real-time three-dimensional ultrasound intelligent imaging method based on row and column addressing array for layered complex media. This method avoids the computational overhead of traditional time-domain voxel-by-voxel ray tracing by processing in the frequency-wavenumber domain, and can achieve efficient, high-precision, real-time three-dimensional ultrasound imaging of layered media.

[0004] To achieve the above objectives, the present invention provides the following technical solution: This invention provides a real-time three-dimensional ultrasound intelligent imaging method based on row and column addressing arrays, comprising: S1, ultrasonic waves are emitted to the target medium through a row and column addressing ultrasonic transducer array, and the echo signal reflected by the target medium is received. Based on the echo signal, the original three-dimensional ultrasonic data is acquired. S2, preprocess the original three-dimensional ultrasound data, and transform the processed original three-dimensional ultrasound data from the spatial domain to the frequency-wavenumber domain to obtain frequency-wavenumber domain data; S3, in the frequency-wavenumber domain, based on the explosion reflector imaging model, the target medium is discretized into multiple depth layers along the depth direction, and the vertical wavenumber of each current depth layer is calculated according to the preset layered sound velocity information corresponding to each depth layer. ; Wherein, the vertical wavenumber This includes the vertical wavenumber used when deriving from the row array elements to the lower region. and the vertical wavenumber used when deriving from the array elements to the lower region The expression for the vertical wavenumber used to derive the received ultrasonic waves from the row array elements to the lower region is:

[0005] The expression for the vertical wavenumber used to derive the received ultrasonic waves from the array elements to the lower region is:

[0006] In the formula, Angular frequency, m For the current layer index, The speed of sound at the current depth layer, respectively along the horizontal x and y Wavenumber components in the direction; S4, based on the set depth step size And for each current depth layer, a phase shift operator is constructed based on the vertical wavenumber of that layer. Based on the phase shift operator, a recursive evolution strategy is used to obtain the predicted wavefield of each depth layer. S5, at each recursive depth step, in the frequency-wavenumber domain, multiply the received wavefield of the current depth layer by the complex conjugate of the corresponding source wavefield, and integrate with respect to the angular frequency to extract the wavefield value corresponding to time zero. The integral result is used as the current depth. The corresponding frequency-wavenumber domain image slice ; S6. Slice the frequency-wavenumber domain images of each depth layer. Transform back into the spatial domain, stack along the depth dimension to generate a three-dimensional high-resolution ultrasonic image of the target medium, and output it to an intelligent detection system for defect or structural analysis.

[0007] Furthermore, the row-column addressed ultrasonic transducer array includes row array elements and column array elements, wherein the row array elements and the column array elements are orthogonally overlapping in space. The row array element layer includes multiple row array elements distributed in a parallel strip shape, and each row array element has a row lead terminal at its end for connecting to an external control circuit. The array element layer includes multiple array elements distributed in parallel strips, and each array element has a column lead endpoint at its end.

[0008] Furthermore, the specific content of step S2 is as follows: The original three-dimensional ultrasound data is subjected to DC component removal and channel normalization. The processed original three-dimensional ultrasound data is then zero-filled in terms of spatial dimensions. Finally, the processed original three-dimensional ultrasound data is transformed from the spatiotemporal domain to the frequency-wavenumber domain using a three-dimensional fast Fourier transform.

[0009] Furthermore, in step S4, the phase shift operator The expression is:

[0010] In the formula, For complex units; The vertical wavenumber is given when the lower region is derived from the row array elements. When the vertical wavenumber is derived from the array elements in the lower region, it is: .

[0011] Furthermore, the specific steps in step S4 for obtaining the predicted wavefield for each depth layer using a recursive evolution strategy based on the phase shift operator are as follows: Combine the total wave field at the current depth with the phase shift operator Multiplying these yields the predicted wavefield for the next depth layer. By proceeding downwards layer by layer, the predicted wavefield for each depth layer is obtained.

[0012] Furthermore, the current depth The corresponding frequency-wavenumber domain image slice The expression is:

[0013] In the formula, The received wave field at the current depth layer is obtained by extrapolating the wave field from the receiving array elements at the layered sound speed; It is the complex conjugate of the emitted wavefield at the current depth layer, obtained by extrapolation of the emitted array element through the wavefield at the layered sound speed.

[0014] Furthermore, the specific steps of step S6 are as follows: depth Slice each frequency-wavenumber domain image Perform a two-dimensional inverse fast Fourier transform to obtain two-dimensional slices of the spatial domain at each depth; By stacking all the two-dimensional spatial slices of all depths along the depth dimension, they are directly combined to form a three-dimensional spatial data volume, which is the final three-dimensional high-resolution ultrasound image.

[0015] In another aspect, the present invention provides a real-time three-dimensional ultrasound intelligent imaging system based on a row-column addressing array, the real-time three-dimensional ultrasound intelligent imaging system comprising: Ultrasonic transceiver module: includes row and column addressing ultrasonic transducer probes and corresponding multi-channel array controllers. Driven by the multi-channel array controller, it directionally emits ultrasonic waves into the target medium and simultaneously receives the echo signals reflected by the target medium. It collects and converts the echo signals and finally outputs the raw three-dimensional ultrasonic data. Data preprocessing and domain transformation module: Removes DC components and normalizes channels from the original 3D ultrasound data, performs zero-filling on the spatial dimensions of the processed original 3D ultrasound data, and transforms the processed original 3D ultrasound data from the spatiotemporal domain to the frequency-wavenumber domain using 3D fast Fourier transform. Vertical wavenumber calculation module: In the frequency-wavenumber domain, based on the explosion reflector imaging model, the target medium is discretized into multiple depth layers along the depth direction, and the vertical wavenumber of each depth layer is calculated according to the preset layer sound velocity information corresponding to each depth layer. Wavefield recursive evolution module: based on the set depth step size And for each current depth layer, a phase shift operator is constructed for that layer based on the vertical wavenumber. A recursive evolution strategy is adopted to evolve the wavefield layer by layer through the phase shift operator, thereby obtaining the predicted wavefield of the current depth layer at each recursive depth step. Imaging slice extraction module: At each recursive depth step, in the frequency-wavenumber domain wavefield data, the received wavefield of the current depth layer is multiplied by the complex conjugate of the corresponding source wavefield, and integrated with respect to the angular frequency to extract the wavefield value corresponding to time zero. The integration result is used as the current depth. The corresponding frequency-wavenumber domain image slice ; Iterate through all recursive depth steps in sequence, repeat the above steps, and finally obtain the frequency-wavenumber domain image slices corresponding to each depth layer. Imaging Output and Intelligent Analysis Module: Slices the frequency-wavenumber domain images of each depth layer. A two-dimensional inverse fast Fourier transform is performed to transform the medium from the frequency-wavenumber domain back to the spatial domain, resulting in two-dimensional spatial domain slices of each depth layer. All two-dimensional spatial domain slices are stacked along the depth direction to generate a three-dimensional high-resolution ultrasonic image of the target medium. Finally, the three-dimensional high-resolution ultrasonic image is output to an intelligent detection system, which performs defect identification and structural analysis on the image to complete the detection task of the target medium. Display module: used to visualize the three-dimensional high-resolution ultrasound image.

[0016] The present invention has the following beneficial effects: (1) The present invention utilizes a row and column addressing ultrasonic transducer array to reduce the number of channels from N² to 2N, which greatly reduces the number of hardware channels required. It effectively solves the problems of poor hardware scalability, complex structure and high cost caused by too many channels in traditional ultrasonic imaging systems, reduces the difficulty of hardware design, manufacturing and deployment of the system, facilitates the miniaturization and lightweight integration of the system, and adapts to more application scenarios.

[0017] (2) This invention constructs an accurate vertical wavenumber expression in the frequency-wavenumber domain, combines a layer recursive wavefield extrapolation scheme, multiplies the received wavefield of the current depth layer with the complex conjugate of the corresponding source wavefield, and integrates with respect to the angular frequency to extract the wavefield value corresponding to time zero, and uses the integration result as the current depth. The corresponding frequency-wavenumber domain image slices fundamentally avoid the large number of redundant voxel-by-voxel calculations in traditional methods, significantly reducing the computational overhead of the system; at the same time, they effectively improve the resolution and imaging quality of ultrasound imaging.

[0018] (3) This invention provides a clear and feasible solution for real-time three-dimensional imaging of multi-layer structures, making up for the shortcomings of existing technologies in the field of real-time imaging of multi-layer structures. Its advantages of high speed, high precision and low cost enable it to be widely used in the fields of clinical diagnosis and industrial non-destructive evaluation. It provides a high-speed in-situ three-dimensional characterization tool for such fields, helping to implement rapid and accurate clinical diagnosis and efficient non-destructive testing of industrial products. It has extremely high practical value and promotion prospects. Attached Figure Description

[0019] Figure 1 This is a flowchart of the three-dimensional ultrasound intelligent imaging process in Example 1.

[0020] Figure 2 This is a general view of a row-column addressed ultrasonic transducer array.

[0021] Figure 3 yes Figure 2 A schematic diagram of the distribution of the array elements in the center row.

[0022] Figure 4 yes Figure 2 A schematic diagram of the distribution of array elements in the middle row.

[0023] Figure 5 This is a schematic diagram of the system modules in Embodiment 2.

[0024] Figure 6 It is a three-dimensional imaging image in a complex layered medium using the traditional DAS method.

[0025] Figure 7 It is a three-dimensional imaging image in a complex layered medium using Example 1. Detailed Implementation

[0026] The technical solution of the present invention will be further described in detail below with reference to specific embodiments. However, these embodiments are not intended to limit the present invention. Any similar structures and similar variations of the present invention should be included in the protection scope of the present invention. The commas in the present invention all indicate the relationship between and. The English letters in the present invention are case-sensitive.

[0027] Example 1 like Figure 1 As shown, this embodiment provides a real-time three-dimensional ultrasound intelligent imaging method based on row and column addressing arrays, including: S1, ultrasonic waves are emitted to the target medium through a row and column addressing ultrasonic transducer array, and the echo signal reflected by the target medium is received. Based on the echo signal, the original three-dimensional ultrasonic data is acquired. like Figures 2-4 As shown, the row and column addressing ultrasonic transducer array 10 includes a row array element layer 12 and a column array element layer 14. The row array element layer 12 and the column array element layer 14 are orthogonally overlapped in space to form a functional grid 16 for transmitting and receiving ultrasonic signals.

[0028] To achieve independent addressing control of signals, such as Figure 3 As shown, the row array element layer 12 includes multiple row array elements distributed in a parallel strip shape, and each row array element has a row lead terminal 18 at its end for connecting to an external control circuit; as shown Figure 4 As shown, the array element layer 14 includes multiple array elements distributed in a parallel strip shape, and each array element has a column lead endpoint 20 at its end; through the row lead endpoint 18 and the column lead endpoint 20, the entire array can be controlled with 2N channels. In the transmission phase, each row electrode is activated sequentially to emit ultrasonic waves; in the reception phase, the echo signals of all column electrodes are collected simultaneously to form a complete transmission-reception dataset.

[0029] S2 involves preprocessing the raw 3D ultrasound data and transforming it from the spatial domain to the frequency-wavenumber domain to obtain frequency-wavenumber domain data. This step leverages the efficient computational characteristics of FFT to lay the foundation for subsequent frequency domain processing and is a crucial step in achieving real-time imaging. The specific steps are as follows: The original three-dimensional ultrasound data is subjected to DC component removal and channel normalization. The processed original three-dimensional ultrasound data is then zero-filled in terms of spatial dimensions. Finally, the processed original three-dimensional ultrasound data is transformed from the spatiotemporal domain to the frequency-wavenumber domain using a three-dimensional fast Fourier transform.

[0030] S3, in the frequency-wavenumber domain, based on the explosion reflector imaging model, the target medium is discretized into multiple depth layers along the depth direction, and the vertical wavenumber of each current depth layer is calculated according to the preset layered sound velocity information corresponding to each depth layer. ; Wherein, the vertical wavenumber This includes the vertical wavenumber used when deriving from the row array elements to the lower region. and the vertical wavenumber used when deriving from the array elements to the lower region ; The expression for the vertical wavenumber used to derive the received ultrasonic waves from the row array elements to the lower region is:

[0031] The expression for the vertical wavenumber used to derive the received ultrasonic waves from the array elements to the lower region is:

[0032] In the formula, Angular frequency, m For the current layer index, The speed of sound at the current depth layer, respectively along the horizontal x and y The expression takes into account the refraction effect of sound waves at the interface of different media layers, providing an accurate theoretical basis for subsequent wavefield extrapolation and ensuring high efficiency while maintaining imaging quality.

[0033] S4, based on the set depth step size The vertical wavenumber of each current depth layer is used to construct a phase shift operator for that layer. Based on the phase shift operator, a recursive evolution strategy is used to obtain the predicted wavefield of each depth layer. The extrapolation result of each layer is used as the input for the extrapolation of the next layer. Through this recursive processing, the wavefield of the entire volume is reconstructed. This process is carried out in the frequency domain, and the computational efficiency is much higher than that of the time domain method.

[0034] Among them, the phase shift operator The expression is:

[0035] In the formula, For complex units; The vertical wavenumber is given when the lower region is derived from the row array elements. When the vertical wavenumber is derived from the array elements in the lower region, it is: .

[0036] The specific steps for obtaining the predicted wavefield at each depth layer are as follows: Combine the total wave field at the current depth with the phase shift operator Multiplying these yields the predicted wavefield for the next depth layer. By proceeding downwards layer by layer, the predicted wavefield for each depth layer is obtained.

[0037] S5, at each recursive depth step, in the frequency-wavenumber domain, multiply the received wavefield of the current depth layer by the complex conjugate of the corresponding source wavefield, and integrate with respect to the angular frequency to extract the wavefield value corresponding to time zero. The integral result is used as the current depth. The corresponding frequency-wavenumber domain image slice This avoids the intermediate process of converting back to the spatiotemporal domain, further improving computational efficiency and serving as an important guarantee for achieving real-time imaging.

[0038] By iterating through all recursive depth steps and repeating the above steps, the frequency-wavenumber domain image slices corresponding to each depth layer are finally obtained.

[0039] Among them, the current depth The corresponding frequency-wavenumber domain image slice The expression is:

[0040] In the formula, The received wave field at the current depth layer is obtained by extrapolating the wave field from the receiving array elements at the layered sound speed; It is the complex conjugate of the emitted wavefield at the current depth layer, obtained by extrapolation of the emitted array element through the wavefield at the layered sound speed.

[0041] S6. Slice the frequency-wavenumber domain images of each depth layer. The process involves transforming the image back to the spatial domain, stacking it along the depth dimension to generate a three-dimensional high-resolution ultrasonic image of the target medium, and then outputting it to an intelligent detection system for defect or structural analysis. The specific steps are as follows: depth Slice each frequency-wavenumber domain image Perform a two-dimensional inverse fast Fourier transform to obtain two-dimensional slices of the spatial domain at each depth; By stacking all the two-dimensional spatial slices of all depths along the depth dimension, they are directly combined to form a three-dimensional spatial data volume, which is the final three-dimensional high-resolution ultrasound image.

[0042] The reconstructed 3D ultrasound images are output to a display or storage device for subsequent clinical diagnosis or industrial testing and analysis. The entire process can be completed in approximately one second on a consumer-grade personal computer, achieving true real-time imaging.

[0043] The intelligent detection system is an automated measurement, control and analysis platform specifically adapted for three-dimensional high-resolution ultrasonic image processing, defect identification and structural analysis. Its core is used to receive three-dimensional high-resolution ultrasonic images output by the three-dimensional full-focus ultrasonic imaging module. Through built-in intelligent algorithms and analysis models, it completes defect identification and structural integrity analysis of the target medium's interior and surface, and finally autonomously completes the non-destructive testing task of the target medium.

[0044] Example 2 like Figure 5 As shown, this embodiment provides a real-time three-dimensional ultrasound intelligent imaging system based on a row and column addressing array. The real-time three-dimensional ultrasound intelligent imaging system includes: Ultrasonic transceiver module: includes row and column addressing ultrasonic transducer probes and corresponding multi-channel array controllers. Driven by the multi-channel array controller, it directionally emits ultrasonic waves into the target medium and simultaneously receives the echo signals reflected by the target medium. It collects and converts the echo signals and finally outputs the raw three-dimensional ultrasonic data. Data preprocessing and domain transformation module: Removes DC components and normalizes channels from the original 3D ultrasound data, performs zero-filling on the spatial dimensions of the processed original 3D ultrasound data, and transforms the processed original 3D ultrasound data from the spatiotemporal domain to the frequency-wavenumber domain using 3D fast Fourier transform. Vertical wavenumber calculation module: In the frequency-wavenumber domain, based on the explosion reflector imaging model, the target medium is discretized into multiple depth layers along the depth direction, and the vertical wavenumber of each current depth layer is calculated according to the preset layer sound velocity information corresponding to each depth layer. Wavefield recursive evolution module: based on the set depth step size And for each current depth layer, a phase shift operator is constructed for that layer based on the vertical wavenumber. A recursive evolution strategy is adopted to evolve the wavefield layer by layer through the phase shift operator, thereby obtaining the predicted wavefield of the current depth layer at each recursive depth step. Imaging slice extraction module: At each recursive depth step, in the frequency-wavenumber domain, the received wavefield of the current depth layer is multiplied by the complex conjugate of the corresponding source wavefield, and integrated with respect to the angular frequency to extract the wavefield value corresponding to time zero. The integration result is used as the current depth. The corresponding frequency-wavenumber domain image slice ; Iterate through all recursive depth steps in sequence, repeat the above steps, and finally obtain the frequency-wavenumber domain image slices corresponding to each depth layer. Imaging Output and Intelligent Analysis Module: Slices the frequency-wavenumber domain images of each depth layer. A two-dimensional inverse fast Fourier transform is performed to transform the medium from the frequency-wavenumber domain back to the spatial domain, resulting in two-dimensional spatial domain slices of each depth layer. All two-dimensional spatial domain slices are stacked along the depth direction to generate a three-dimensional high-resolution ultrasonic image of the target medium. Finally, the three-dimensional high-resolution ultrasonic image is output to an intelligent detection system, which performs defect identification and structural analysis on the image to complete the detection task of the target medium. Display module: used to visualize the three-dimensional high-resolution ultrasound image.

[0045] To verify the effectiveness of the real-time three-dimensional ultrasound intelligent imaging method proposed in this application, a comparative experiment was conducted with the traditional time-domain delay summation (DAS) beamforming method. The verification was carried out from two dimensions: imaging quality and computational efficiency. The results are as follows: from Figure 6 As can be seen, the imaging results exhibit significant background noise, blurred target structure edges, low detail recognition, and unclear contours of layered interfaces and internal defects, failing to accurately reflect the true structural characteristics of the target medium; from Figure 7 As can be seen, the clarity of the imaging results is significantly improved, background noise is effectively suppressed, the detail distinguishability of layered interfaces and internal structures is significantly improved, the target contour is sharp and the contrast is high, and the three-dimensional structure of the target medium can be accurately restored. The comparison shows that the real-time three-dimensional ultrasound intelligent imaging method proposed in this application has a much better imaging quality than the traditional DAS method.

[0046] Furthermore, traditional DAS methods employ time-domain voxel-by-voxel ray tracing imaging logic, resulting in high computational complexity and overhead, which cannot meet the real-time imaging requirements of 3D ultrasound. This application addresses this by using frequency-wavenumber domain layer recursive wavefield extrapolation, multiplying the received wavefield of the current depth layer by the complex conjugate of the corresponding source wavefield, and integrating with respect to the angular frequency to extract the wavefield value corresponding to time zero. The integration result is then used as the current depth. The corresponding frequency-wavenumber domain image slice This proposed solution fundamentally reduces computational complexity, improving computational speed by 2-3 orders of magnitude compared to traditional DAS methods. It enables the reconstruction of 3D volume data on a consumer-grade personal computer within approximately one second, truly achieving real-time 3D ultrasound imaging of layered media. Based on a comparison of imaging quality and computational efficiency, this application significantly reduces computational overhead while maintaining 3D imaging accuracy and improving image detail recognition. It overcomes the technical bottlenecks of poor imaging quality and insufficient real-time performance inherent in traditional DAS methods, providing a feasible technical solution for high-speed, high-precision 3D ultrasound imaging of layered media.

[0047] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

Claims

1. A real-time three-dimensional ultrasound intelligent imaging method based on row and column addressing array, characterized in that, include: S1, ultrasonic waves are emitted to the target medium through a row and column addressing ultrasonic transducer array, and the echo signal reflected by the target medium is received. Based on the echo signal, the original three-dimensional ultrasonic data is acquired. S2, preprocess the original three-dimensional ultrasound data, and transform the processed original three-dimensional ultrasound data from the spatial domain to the frequency-wavenumber domain to obtain frequency-wavenumber domain data; S3, in the frequency-wavenumber domain, based on the explosion reflector imaging model, the target medium is discretized into multiple depth layers along the depth direction, and the vertical wavenumber of each current depth layer is calculated according to the preset layered sound velocity information corresponding to each depth layer. ; Wherein, the vertical wavenumber This includes the vertical wavenumber used when deriving from the row array elements to the lower region. and the vertical wavenumber used when deriving from the array elements to the lower region ; The expression for the vertical wavenumber used to derive the received ultrasonic waves from the row array elements to the lower region is: The expression for the vertical wavenumber used to derive the received ultrasonic waves from the array elements to the lower region is: In the formula, Angular frequency, m For the current layer index, The speed of sound at the current depth layer, respectively along the horizontal x and y Wavenumber components in the direction; S4, based on the set depth step size And for each current depth layer, a phase shift operator is constructed based on the vertical wavenumber of that layer. Based on the phase shift operator, a recursive evolution strategy is used to obtain the predicted wavefield of each depth layer. S5, at each recursive depth step, in the frequency-wavenumber domain, multiply the received wavefield of the current depth layer by the complex conjugate of the corresponding source wavefield, and integrate with respect to the angular frequency to extract the wavefield value corresponding to time zero. The integration result is used as the current depth. The corresponding frequency-wavenumber domain image slice ; S6. Slice the frequency-wavenumber domain images of each depth layer. Transform back into the spatial domain, stack along the depth dimension to generate a three-dimensional high-resolution ultrasonic image of the target medium, and output it to an intelligent detection system for defect or structural analysis.

2. The real-time three-dimensional ultrasound intelligent imaging method based on row and column addressing array according to claim 1, characterized in that, The row-column addressing ultrasonic transducer array includes row array elements and column array elements, which are orthogonally overlapping in space. The row array element layer includes multiple row array elements distributed in a parallel strip shape, and each row array element has a row lead terminal at its end for connecting to an external control circuit. The array element layer includes multiple array elements distributed in parallel strips, and each array element has a column lead endpoint at its end.

3. The real-time three-dimensional ultrasound intelligent imaging method based on row and column addressing array according to claim 1, characterized in that, The specific content of step S2 is as follows: The original three-dimensional ultrasound data is subjected to DC component removal and channel normalization. The processed original three-dimensional ultrasound data is then zero-filled in terms of spatial dimensions. Finally, the processed original three-dimensional ultrasound data is transformed from the spatiotemporal domain to the frequency-wavenumber domain using a three-dimensional fast Fourier transform.

4. The real-time three-dimensional ultrasound intelligent imaging method based on row and column addressing array according to claim 1, characterized in that, Phase shift operator in step S4 The expression is: In the formula, For complex units; The vertical wavenumber is given when the lower region is derived from the row array elements. When the vertical wavenumber is derived from the array elements in the lower region, it is: .

5. The real-time three-dimensional ultrasound intelligent imaging method based on row and column addressing array according to claim 4, characterized in that, The specific steps in step S4 for obtaining the predicted wavefield for each depth layer using a recursive evolution strategy based on the phase shift operator are as follows: Combine the total wave field at the current depth with the phase shift operator Multiplying these yields the predicted wavefield for the next depth layer. By proceeding downwards layer by layer, the predicted wavefield for each depth layer is obtained.

6. The real-time three-dimensional ultrasound intelligent imaging method based on row and column addressing array according to claim 1, characterized in that, Current depth The corresponding frequency-wavenumber domain image slice The expression is: In the formula, The received wave field at the current depth layer is obtained by extrapolating the wave field from the receiving array elements at the layered sound speed; It is the complex conjugate of the emitted wavefield at the current depth layer, obtained by extrapolation of the emitted array element through the wavefield at the layered sound speed.

7. The real-time three-dimensional ultrasound intelligent imaging method based on row and column addressing array according to claim 1, characterized in that, The specific steps of step S6 are as follows: depth Perform a two-dimensional inverse fast Fourier transform on each frequency-wavenumber domain image slice to obtain two-dimensional spatial domain slices at each depth. By stacking all the two-dimensional spatial slices of all depths along the depth dimension, they are directly combined to form a three-dimensional spatial data volume, which is the final three-dimensional high-resolution ultrasound image.

8. A real-time three-dimensional ultrasound intelligent imaging system based on row and column addressing array, characterized in that, The real-time three-dimensional ultrasound intelligent imaging system is applicable to the real-time three-dimensional ultrasound intelligent imaging method according to any one of claims 1-7, and the real-time three-dimensional ultrasound intelligent imaging system comprises: Ultrasonic transceiver module: includes row and column addressing ultrasonic transducer probes and corresponding multi-channel array controllers. Driven by the multi-channel array controller, it directionally emits ultrasonic waves into the target medium and simultaneously receives the echo signals reflected by the target medium. It collects and converts the echo signals and finally outputs the raw three-dimensional ultrasonic data. Data preprocessing and domain transformation module: Removes DC components and normalizes channels from the original 3D ultrasound data, performs zero-filling on the spatial dimensions of the processed original 3D ultrasound data, and transforms the processed original 3D ultrasound data from the spatiotemporal domain to the frequency-wavenumber domain using 3D fast Fourier transform. Vertical wavenumber calculation module: In the frequency-wavenumber domain, based on the explosion reflector imaging model, the target medium is discretized into multiple depth layers along the depth direction, and the vertical wavenumber of each current depth layer is calculated according to the preset layer sound velocity information corresponding to each depth layer. Wavefield recursive evolution module: based on the set depth step size And for each current depth layer, a phase shift operator is constructed for that layer based on the vertical wavenumber. A recursive evolution strategy is adopted to evolve the wavefield layer by layer through the phase shift operator, thereby obtaining the predicted wavefield of the current depth layer at each recursive depth step. Imaging slice extraction module: At each recursive depth step, in the frequency-wavenumber domain, the received wavefield of the current depth layer is multiplied by the complex conjugate of the corresponding source wavefield, and integrated with respect to the angular frequency to extract the wavefield value corresponding to time zero. The integration result is used as the current depth. The corresponding frequency-wavenumber domain image slice ; Iterate through all recursive depth steps in sequence, repeat the above steps, and finally obtain the frequency-wavenumber domain image slices corresponding to each depth layer. Imaging Output and Intelligent Analysis Module: Slices the frequency-wavenumber domain images of each depth layer. A two-dimensional inverse fast Fourier transform is performed to transform the medium from the frequency-wavenumber domain back to the spatial domain, resulting in two-dimensional spatial domain slices of each depth layer. All two-dimensional spatial domain slices are stacked along the depth direction to generate a three-dimensional high-resolution ultrasonic image of the target medium. Finally, the three-dimensional high-resolution ultrasonic image is output to an intelligent detection system, which performs defect identification and structural analysis on the image to complete the detection task of the target medium. Display module: used to visualize the three-dimensional high-resolution ultrasound image.

Images