A method, system, medium, and product for ultrasonic tomographic transducer design

Abstract

The application discloses a kind of ultrasonic tomography transducer design method, system, medium and product, it is related to ultrasonic imaging technical field, including: with center frequency, element quantity, element bandwidth, element aperture and element elevation angle focusing characteristic as the core physical parameter of transducer, construct annular array ultrasonic tomography acoustic simulation model, by changing the value of each core physical parameter, execute model forward simulation, generate the original radio frequency data of core physical parameter influence on imaging quality;Execute the velocity reconstruction based on curved ray model, to translate the original radio frequency data into velocity image;Quantitative evaluation of the quality of velocity image, to determine the relationship between each core physical parameter and system imaging performance, to generate the transducer design criterion for guiding ultrasonic tomography system;The design method provides a set of transducer parameter combination for velocity imaging system based on ray model, which considers high imaging quality and actual engineering constraint.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ultrasonic imaging, and in particular to an ultrasonic tomography transducer design method, system, medium and product. BACKGROUND

[0002] Ultrasound Computed Tomography (USCT) is an emerging non-invasive biomedical imaging technology. Its core principle is to place the imaging target at the center of an ultrasonic transducer array with a certain geometric configuration, and use a one-transmit-all-receive strategy to collect data, that is, to excite a single element as a transmitting source to generate an ultrasonic pulse, and the remaining elements synchronously record the full wave field signal penetrating the target or reflected by the internal structure. For a large amount of raw radio frequency data from different angles, a suitable inverse problem solving algorithm can be used to reconstruct the registered reflectivity, sound speed and sound attenuation images. This imaging paradigm makes USCT inherit the core advantages of traditional ultrasound, such as no ionizing radiation, good penetration of soft tissue, etc., while overcoming the pain points of traditional ultrasound, such as dependence on operator skills, lack of standardization, easy to miss small lesions, etc., and fundamentally breaking through the inherent limitations of traditional ultrasound in qualitative diagnosis and two-dimensional perspective. It provides unprecedented multi-dimensional acoustic parameter characterization for biological tissues, and inherently supports three-dimensional isotropic volumetric imaging. With its unique imaging capabilities, USCT technology has great application potential in the field of biomedical, especially in the early screening and differential diagnosis of breast cancer, and is expected to play a more important role in future basic research and clinical translation.

[0003] In a typical USCT system, a transducer array with a ring or semi-ring geometry is usually used to collect full-angle data of the imaging target to maximize the completeness of the projection data required for the reconstruction problem, and to achieve high-quality tomographic imaging. However, the final quality of the reconstructed image (covering multiple dimensions such as spatial resolution, quantitative accuracy and artifact level) is fundamentally subject to the physical characteristics of the transducer array. A series of key parameters such as center frequency, number of elements, bandwidth and geometric size not only independently affect the transmission and reception of ultrasonic waves, but also are coupled and constrained in multiple dimensions, forming a complex system parameter design space. Therefore, in order to obtain high-fidelity image reconstruction results in actual biomedical imaging applications, it is crucial to systematically optimize the key parameters of the transducer.

[0004] To improve the imaging quality of USCT systems using ring array transducers for ultrasound data acquisition, several technical solutions and design principles based on specific transducer parameter adjustments have been disclosed by those skilled in the art, some of which have been applied in commercial imaging systems. These solutions primarily optimize the number, spacing, and center frequency of transducer elements. For example, a commonly used approach, based on Nyquist sampling theory, designs the element spacing of the ring array to be less than half the center wavelength to effectively suppress grating lobe artifacts in reflectivity images. Simultaneously, increasing the total number of elements (e.g., 1024 or more) to improve the ultrasound beam coverage density has also proven to be an effective means of reducing sound velocity image noise and improving quantitative accuracy. Further research indicates that increasing the number of elements primarily improves the contrast resolution of reflectivity images rather than spatial resolution, and this performance gain tends to saturate when the number of elements exceeds a certain threshold. This design principle has been applied to the transducer design of the commercial USCT system UltraLucid, effectively improving the system's imaging performance. In addition, another key technical solution is to take advantage of the direct correlation between acoustic imaging resolution and operating wavelength, and directly improve spatial resolution by increasing the center operating frequency of the transducer. This solution has also been successfully applied to the transducer design process of the commercial USCT system SoftVue, achieving a doubling of the acoustic imaging quality of the system.

[0005] While the above-mentioned existing technical solutions have improved the performance of USCT in certain aspects, they all have significant limitations, making it impossible to achieve global optimization of system performance:

[0006] 1. The isolation and one-sidedness of the solutions: Existing solutions all adopt a single-variable optimization approach, ignoring the complex coupling relationships between various physical parameters. For example, increasing the frequency will exacerbate sound attenuation, and decreasing the array element aperture will reduce the signal-to-noise ratio. Therefore, optimizing a single parameter in isolation often sacrifices other performance indicators, and may even lead to a decline in overall performance;

[0007] 2. Lack of consideration for other key parameters: Existing solutions mainly focus on the center frequency and the number of array elements, while lacking effective optimization design methods for equally crucial parameters such as array element bandwidth, three-dimensional aperture (width and height), and elevation focusing characteristics. These insufficiently considered parameters are precisely the key factors leading to image distortion, uneven resolution, and other problems.

[0008] 3. Inability to achieve an effective trade-off between cost and performance: Due to a lack of understanding of the synergistic effects of multiple parameters, existing designs struggle to achieve optimal imaging performance within limited hardware costs. Summary of the Invention

[0009] Based on the technical problems existing in the background technology, this invention proposes a design method, system, medium and product for ultrasonic tomography transducers, providing a set of transducer parameter combinations that take into account both high imaging quality and practical engineering constraints for ray model-based sound velocity imaging systems.

[0010] This invention proposes a design method for an ultrasonic tomographic imaging transducer, comprising:

[0011] Using the center frequency, number of array elements, array element bandwidth, array element aperture, and array element elevation angle focusing characteristics as the core physical parameters of the transducer, an acoustic simulation model for ultrasonic tomography based on multidimensional physical characteristic analysis is constructed. By changing the values ​​of each core physical parameter, the model is forward simulated to generate raw radio frequency data on the impact of the core physical parameters on imaging quality.

[0012] Perform sound velocity reconstruction based on a curved ray model to convert the raw radio frequency data into a sound velocity image;

[0013] The quality of sound velocity images is quantitatively evaluated to determine the relationship between various core physical parameters and system imaging performance, thereby generating transducer design criteria to guide ultrasound tomography systems. Specifically, when quantifying the quality of sound velocity images by changing the number of array elements:

[0014] Starting from the same transmitting element, the physical distance between the sampling rays formed by two adjacent receiving elements at the propagation distance is used as the element interval, where the sampling ray passes through the target sound speed region and does not pass through either side of the target sound speed region;

[0015] By controlling the element spacing by varying the number of elements, the reconstruction results of the critical-size phantom under different element spacings were compared. Based on this, the design criterion for the minimum number of elements was derived: the element spacing should not exceed [a certain value]. , The center wavelength is .

[0016] Furthermore, the quantitative evaluation of the sound velocity image quality involves changing the center frequency of the transducer while keeping other core physical parameters constant, and performing forward simulation and sound velocity reconstruction on the acoustic simulation model corresponding to each center frequency to obtain the design criterion for the center frequency: taking the sound velocity image of the mid-frequency band between the low-frequency band and the high-frequency band.

[0017] Furthermore,

[0018] The design criteria for obtaining the minimum number of array elements are as follows:

[0019] By sampling rays, the internal information of the target sound speed region and the spatial boundary of the target sound speed region are obtained, so as to find that the maximum end physical interval appears near the receiving array element directly opposite the transmitting array element.

[0020] Based on the fact that the size of the maximum physical interval at the end is equal to the spacing between the array elements on the array, a complex ray coverage problem that dynamically changes with spatial position can be simplified and equivalently represented as a constraint problem on the spacing between array elements, wherein the spacing is determined by the total number of array elements and the array diameter.

[0021] A quantitative relationship is obtained to guide the configuration of the number of array elements, and this quantitative relationship is used as the design criterion for the minimum number of array elements.

[0022] Furthermore, the quality of the quantified sound velocity image is evaluated, wherein the array element bandwidth determines the measurement accuracy of the time of flight;

[0023] The design principle for the array element bandwidth is to select wideband transducer array elements.

[0024] Furthermore, the quality of the quantified sound velocity image is evaluated, wherein the array element aperture includes the array element width and the array element height;

[0025] The design principle for the array element width is: the array element width is set at... Within the range.

[0026] Furthermore, the design criteria for the element height are as follows: the optimal element height for the ring array transducer is designed to be around three times the center wavelength.

[0027] Furthermore, when generating raw RF data, the system systematically traverses each core physical parameter point in the physical parameter analysis space and performs acoustic forward simulation on the corresponding sound velocity simulation model to obtain raw RF data corresponding to each core physical parameter.

[0028] A computer system includes a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the method described above.

[0029] A computer-readable storage medium storing a plurality of classification programs, the plurality of classification programs being invoked by a processor to execute the method described above.

[0030] A computer program product includes a computer program that is executed by a processor using the method described above.

[0031] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as ROM, RAM, magnetic disk, or optical disk.

[0032] The advantages of the ultrasound tomography transducer design method, system, medium, and product provided by this invention are as follows: It establishes a parameter optimization design method for ring array ultrasound tomography transducers based on multidimensional physical characteristic analysis and systematic numerical simulation. By deeply decoupling and quantitatively evaluating the inherent trade-offs of each core physical parameter, it achieves the effect of providing an optimal or balanced parameter combination for a ray-model-based sound velocity imaging system that balances high imaging quality with practical engineering constraints. This lays a solid foundation for developing clinical ultrasound tomography equipment with both excellent performance and cost-effectiveness, and accelerates the promotion and application of this imaging technology in the early diagnosis of various major diseases. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of the process of the present invention;

[0034] Figure 2 A block diagram showing the core physical parameters and performance trade-offs of the ring array USCT transducer;

[0035] Figure 3 This is a schematic diagram illustrating the principle of sound velocity reconstruction based on the curved ray model.

[0036] Figure 4 The flowchart shows the sound velocity reconstruction algorithm based on the curved ray model.

[0037] Figure 5 This is a schematic diagram of the parameter distribution of a high-fidelity two-dimensional numerical breast phantom. In the diagram, 'a' represents the positional distribution of various tissue types that may exist in the actual breast tissue simulated by the numerical breast phantom; 'b' shows the acoustic parameter distribution corresponding to each simulated tissue in the phantom, where b1 is the sound velocity distribution of the phantom, b2 is the density distribution of the phantom, and b3 is the sound attenuation coefficient. Location distribution, b4 is the power-law decay exponent. Location distribution;

[0038] Figure 6 Comparison images of numerical breast phantom reconstruction results at different center frequencies are shown. Image a is a comparison image of sound velocity reconstruction results of numerical breast phantom at different center frequencies (1–4MHz). Image b is a quantitative comparison image of the true sound velocity and the one-dimensional sound velocity profile along the central horizontal line (Y=0) of each image in image a. Specifically, image a1 is the numerical breast phantom reconstruction result at a center frequency of 1MHz, image a2 is the numerical breast phantom reconstruction result at a center frequency of 2MHz, image a3 is the numerical breast phantom reconstruction result at a center frequency of 3MHz, image a4 is the numerical breast phantom reconstruction result at a center frequency of 4MHz, image b1 is the reference image of the true sound velocity of the numerical phantom, and image b2 is a quantitative comparison image of the one-dimensional sound velocity profile along the central horizontal line (Y=0) of each image in image a.

[0039] Figure 7 A schematic diagram illustrating the principle of sound speed target discrimination based on geometric relationships;

[0040] Figure 8 Comparison of reconstruction results for the critical-size sound velocity phantom under different element numbers: a) shows the reconstruction result with a center frequency of 1.25 MHz and a diameter of 3.6 mm; b) shows the reconstruction result with a center frequency of 2.50 MHz and a diameter of 1.8 mm. In both sets of experimental results, sub-figures a1 and b1 represent the true sound velocity reference distribution; a2 shows the reconstruction result with a center frequency of 1.25 MHz and a diameter of 3.6 mm, and an element number of 64 (i.e., 2... 6 The reconstruction results are shown in Figure a3, where a3 represents the number of array elements is 128 (i.e., 2) at a center frequency of 1.25 MHz and a phantom diameter of 3.6 mm. 7 The reconstruction results are shown in Figure a4, where a4 represents the number of array elements (256) at a center frequency of 1.25 MHz and a diameter of 3.6 mm. 8 The reconstruction result is shown in Figure a5, where a5 represents the number of array elements is 512 (i.e., 2) at a center frequency of 1.25MHz and a phantom diameter of 3.6mm. 9 The reconstruction result is shown in Figure a6, where a6 represents the number of array elements (1024, i.e., 2^34) at a center frequency of 1.25 MHz and a diameter phantom of 3.6 mm. 10 The reconstruction result is shown in Figure 1; Figure 2 shows the reconstruction result with a center frequency of 2.50MHz and a phantom diameter of 1.8mm, and the number of array elements is 64 (i.e., 2). 6 The reconstruction result is shown in Figure b3, where b3 represents the number of array elements (128) at a center frequency of 2.50 MHz and a diameter phantom of 1.8 mm. 7 The reconstruction results are shown in Figure b4, where b4 represents the number of array elements is 256 (i.e., 2) at a center frequency of 2.50 MHz and a phantom diameter of 1.8 mm. 8 The reconstruction result is shown in image b5, where b5 represents the number of array elements (512) at a center frequency of 2.50 MHz and a phantom diameter of 1.8 mm. 9 The reconstruction result is shown in Figure b6, where b6 represents the number of array elements (1024) at a center frequency of 2.50 MHz and a diameter phantom of 1.8 mm. 10 The reconstruction result image at time );

[0041] Figure 9The simulation results of the impact of array element bandwidth on the sound velocity imaging quality of USCT are shown in Figure a. Figure a compares the reconstructed sound velocity images under different bandwidths (50%, 100%, 150% and ideal infinite bandwidth), where subfigure a1 is the reference sound velocity distribution, a2 is the sound velocity reconstruction result under the condition of 50% array element bandwidth, a3 is the sound velocity reconstruction result under the condition of 100% array element bandwidth, a4 is the sound velocity reconstruction result under the condition of 150% array element bandwidth, and a5 is the sound velocity reconstruction result under the condition of infinite array element bandwidth. Figure b provides supplementary verification of noise robustness, where b1 is the reconstruction error diagram after applying two different noise instances under the condition of 50% array element bandwidth, b2 is the reconstruction error diagram after applying two different noise instances under the condition of 100% array element bandwidth, and b3 is the reconstruction error diagram after applying two different noise instances under the condition of 150% array element bandwidth.

[0042] Figure 10 The quantitative analysis of the effect of array element bandwidth on the stability of time-of-flight extraction in USCT acoustic imaging shows that the stability of the first arrival time measurement of the received signal gradually becomes more stable and accurate as the bandwidth increases. a1 is a schematic diagram of the TOF extraction results under 50% narrowband conditions, a2 is a schematic diagram of the TOF extraction results under 100% narrowband conditions, and a3 is a schematic diagram of the TOF extraction results under 150% narrowband conditions.

[0043] Figure 11 The images show a comparison of sound velocity reconstruction results for different element widths. The sub-images are as follows: a) reference sound velocity distribution; b) reconstruction result for ideal point-like elements; c) reconstruction result for element width of 1.0 mm (approximately...). The reconstruction result is shown in the image when the array element width is 2.0 mm (approximately 0.5 mm). The reconstruction result image at time );

[0044] Figure 12 A schematic diagram of a three-dimensional numerical phantom used to evaluate the effect of element height on elevation resolution and out-of-plane interference suppression capability;

[0045] Figure 13 The images show the two-dimensional sound velocity reconstruction results of the central section of the three-dimensional phantom at different element heights, along with quantitative comparison results. This demonstrates the trade-off between element height in suppressing sound field divergence interference and avoiding height averaging effects. Image a shows the two-dimensional sound velocity reconstruction results of the central section of the three-dimensional phantom at different element heights. Image b shows the quantitative comparison results of the one-dimensional sound velocity profiles along the central horizontal line (Y=0) in each image of a. Image a1 shows the reference sound velocity distribution of the central section. Image a2 shows the reconstructed image of the central section when the element height is ideally pointlike. Image a3 shows the reconstructed image of the central section when the element height is three times the central wavelength. Image a4 shows the reconstructed image of the central section when the element height is infinitely large (Inf).

[0046] Figure 14 The image shows a comparison of the sound velocity reconstruction results of the three-dimensional phantom under two conditions: non-focused and focused. This is used to verify the effectiveness of elevation angle focusing in improving image quality. In the image, a is the two-dimensional cross-sectional sound velocity reconstruction result corresponding to the non-focused array element; b is the two-dimensional cross-sectional sound velocity reconstruction result corresponding to the focused array element; and c is the quantitative comparison result of the one-dimensional sound velocity profile of a and b along the central horizontal line (Y=0). Detailed Implementation

[0047] The technical solution of the present invention will now be described in detail through specific embodiments. Many specific details are set forth in the following description to provide a thorough understanding of the invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0048] like Figures 1 to 14 As shown, the present invention proposes a design method for an ultrasonic tomographic transducer, comprising:

[0049] Step 1: Using the center frequency, number of array elements, array element bandwidth, array element aperture, and array element elevation angle focusing characteristics as the core physical parameters of the transducer, construct an acoustic simulation model of ring array ultrasonic tomography based on multi-dimensional physical characteristic analysis, and perform forward simulation of the model to generate raw radio frequency data on the influence of each core physical parameter on imaging quality.

[0050] Step 2: Perform sound velocity reconstruction based on the curved ray model to convert the raw radio frequency data into a sound velocity image;

[0051] Step 3: By changing the values ​​of various core physical parameters, the quality of the sound velocity image is quantitatively evaluated to determine the relationship between each core physical parameter and the system's imaging performance, thereby generating transducer design criteria to guide this type of ultrasonic tomographic imaging system. Specifically, when quantitatively evaluating the quality of the sound velocity image by changing the number of array elements:

[0052] Starting from the same transmitting element, the physical distance between the sampling rays formed by two adjacent receiving elements at the propagation distance is used as the element interval, where the sampling ray passes through the target sound speed region and does not pass through either side of the target sound speed region;

[0053] By controlling the element spacing by varying the number of elements, the reconstruction results of the critical-size phantom under different element spacings were compared. Based on this, the design criterion for the minimum number of elements was derived: the element spacing should not exceed [a certain value]. , The center wavelength is .

[0054] This embodiment provides a novel design method for ring array USCT (ultrasound computed tomography) transducers through multi-parameter collaborative optimization. The aim is to find a set of globally optimal or highly balanced transducer parameter combinations that can achieve the best overall quality of sound velocity imaging while meeting engineering constraints. This lays a solid foundation for developing clinical USCT equipment that combines excellent performance and cost-effectiveness, and accelerates its application in the early diagnosis of various major diseases.

[0055] In one embodiment, step one involves constructing an acoustic simulation model for ring array ultrasonic tomography based on multidimensional physical characteristic analysis, using the center frequency, number of array elements, array element bandwidth, array element aperture, and array element elevation focusing characteristics as the core physical parameters of the transducer. The model is then subjected to forward simulation to generate raw radio frequency data on the impact of each core physical parameter on imaging quality. Specifically:

[0056] The key to this embodiment lies in establishing a high-fidelity ultrasonic forward propagation model in acoustic simulation software, and then, through in-depth theoretical analysis, identifying a set of core physical parameters affecting the quality of sound velocity imaging, and clarifying their unique and mutually restrictive influence mechanisms. The core content and trade-offs are as follows: Figure 1 As shown, the specific description is as follows:

[0057] (A1) Center frequency;

[0058] This physical parameter is crucial in determining the core trade-off between system spatial resolution and effective imaging depth. On one hand, increasing the center frequency can compress the width of the first Fresnel zone by shortening the wavelength, thereby improving the theoretical spatial resolution. On the other hand, frequency-dependent acoustic attenuation reduces the signal-to-noise ratio (SNR), interfering with the accurate extraction of time-of-flight (TOF), thus limiting the actual effective resolution and penetration depth. Therefore, the choice of center frequency is a precise trade-off between theoretical resolution and effective SNR.

[0059] Time of Flight (TOF) is a distance measurement technique based on the round-trip time of light pulses. It calculates the distance to an object by measuring the time difference between the emitted and received light and is widely used in high-precision 3D imaging and positioning.

[0060] (A2) Number of array elements;

[0061] This embodiment reveals that, since sound velocity reconstruction relies on time-of-flight (TOF) data rather than wavenumber domain information, the traditional Nyquist sampling theorem is not applicable. Under this physical model, while increasing the number of array elements can improve image accuracy by increasing ray coverage density, this performance gain exhibits a saturation effect, while hardware costs and data processing complexity increase dramatically. Therefore, its design requires a trade-off between the saturation gain of reconstruction quality and engineering costs.

[0062] (A3) Array element bandwidth:

[0063] This embodiment clarifies for the first time that in acoustic imaging, the core role of bandwidth is to determine the measurement accuracy of Time of Flight (TOF), which is fundamentally different from traditional structural imaging, which mainly focuses on its impact on axial resolution. Wider bandwidth corresponds to steeper time-domain signal edges, enabling high-precision, noise-insensitive TOF extraction, a physical prerequisite for ensuring the quantitative accuracy of the final acoustic velocity spectrum. Existing technologies lack a systematic consideration of this parameter.

[0064] (A4) Array element aperture (including array element width and height):

[0065] The limited physical aperture deviates from the ideal point source model, leading to multi-dimensional performance trade-offs. The design of the element width (W) requires a trade-off between lateral resolution (requiring a small width to reduce TOF measurement bias) and signal-to-noise ratio (requiring a large width). Excessive width can also cause model mismatch and artifacts due to directional effects. The element height (H) requires a trade-off between elevation resolution (requiring a small height) and suppressing out-of-plane signal contamination (requiring a large height to avoid contaminating TOF data). The core issue is resolving the contradiction between the assumptions of three-dimensional physical propagation and two-dimensional reconstruction.

[0066] (A5) Array element elevation angle focusing characteristics:

[0067] Elevation focusing is an active sound field manipulation technique that concentrates sound energy into a specific focal area, reducing volume averaging effects and improving elevation resolution and signal-to-noise ratio within that focal area, thereby purifying the TOF data used for sound velocity reconstruction. Previous designs did not explicitly use it as an effective way to improve the accuracy of sound velocity quantification. This embodiment innovatively incorporates it as one of the core optimization parameters.

[0068] (A6) Perform multi-parameter forward simulation to generate the raw RF dataset;

[0069] The aim is to generate a series of raw radio frequency (RF) data that reflects the impact of various core physical parameters on image quality. This process involves constructing a multidimensional parameter analysis space that covers preset value ranges for each core physical parameter.

[0070] To evaluate the unique physical effects of different parameters, this method involves designing a dedicated sound velocity phantom to match specific parameters or combinations of parameters, thereby achieving precise decoupling of the effects of each physical parameter.

[0071] Specifically, the geometry and acoustic properties of each dedicated sound velocity phantom are meticulously designed to isolate and amplify a specific image quality metric to be evaluated.

[0072] For example, in discussing the impact of center frequency on image quality, the phantom was set as a numerical breast phantom containing background fat, fibrous glands, and tumor lesions with frequency-dependent acoustic attenuation to simulate the most common disease model in clinical applications of the USCT system. At the same time, the peak source sound pressure level of the transmitting transducer was set to a typical value of 200 kPa, and Gaussian white noise with a standard deviation of 5 Pa was superimposed on each received signal to simulate the inherent noise level of the system, thus more comprehensively considering the impact of center frequency on image quality.

[0073] When discussing the number of array elements, the phantom also includes targets whose size is precisely matched to the limits of the system's physical resolution. Its layout (e.g., a crosshair lattice) and target spacing are specially designed to create a critical scene that can sensitively reflect whether the sampling is sufficient.

[0074] When discussing the effects of element height and elevation-direction focusing, the phantom must be three-dimensional, with the key feature being the inclusion of structures that vary along the elevation direction (Z-axis), such as out-of-plane interference objects located outside the central plane, or a centrally located target with a height matching the theoretical resolution. By observing whether the in-plane reconstructed image is contaminated by out-of-plane objects, or the reconstruction fidelity of the central target in the Z-axis direction, the layer thickness effect and elevation resolution of the system can be directly evaluated.

[0075] This embodiment, by employing the aforementioned goal-oriented dedicated phantom design strategy, ensures that in univariate analysis, the observed changes in image quality can be clearly and unambiguously attributed to the single parameter currently changing, which greatly enhances the scientific rigor and reliability of the parameter optimization conclusions.

[0076] It should be noted that the above-described ring array ultrasound tomography acoustic simulation model in this embodiment is a general concept. The acoustic simulation models constructed in the process of discussing various core physical parameters mentioned in the text are all ring array ultrasound tomography acoustic simulation models. The difference lies in the different phantoms used. The geometric structure and acoustic characteristics of each dedicated phantom are designed to isolate and amplify a specific image quality index to be evaluated.

[0077] Subsequently, the core physical parameter points in the physical parameter analysis space are systematically traversed, and acoustic forward simulations are performed on the corresponding sound velocity simulation models to obtain the corresponding raw radio frequency (RF) data. The final output of this step is a raw RF dataset that corresponds one-to-one with each physical parameter point in the multidimensional parameter space, which is used for subsequent image reconstruction and quantization evaluation.

[0078] In one embodiment, step two, performing sound velocity reconstruction based on a curved ray model to convert the raw radio frequency data into a sound velocity image, specifically involves:

[0079] This embodiment aims to transform the raw radio frequency (RF) data generated in step one into corresponding, quantifiable sound velocity images. The process includes first applying a time-of-flight (TOF) extraction algorithm (e.g., but not limited to, criteria based on signal statistical characteristics analysis) to each set of raw RF data to obtain TOF measurements for reconstruction. Subsequently, the extracted TOF measurements are input into an iterative inverse problem solver based on a nonlinear sound wave propagation model for sound velocity inversion. In a preferred embodiment, the solver is based on a curved ray model to more accurately simulate the propagation path of sound waves in a non-uniform medium. This iterative process continues until the reconstruction results converge. The final output of this step is a series of sound velocity images, each corresponding to a specific set of transducer parameter combinations from step one.

[0080] This embodiment employs an iterative inverse problem solver based on a nonlinear acoustic wave propagation model. A typical solution method is cited here: LA Shepp, and Y. Vardi, "Maximum likelihood reconstruction for emission tomography," IEEE transactions on medical imaging, 1, 113-122 (1982). For details on the curved ray model, please refer to the literature "Research Progress on Ultrasonic Velocity Tomography Algorithm Based on Ray Model."

[0081] like Figure 4 As shown, the sound velocity reconstruction algorithm based on the curved ray model is as follows: First, initialize the grid slowness distribution. Obtain the system matrix based on ray tracing. Compare the theoretical arrival time of the sound wave with the measured arrival time. If the deviation is less than a preset value, convert the slowness distribution to a sound velocity distribution. If the deviation is greater than or equal to the preset value, iteratively update the grid slowness distribution. Return the system matrix obtained from ray tracing and obtain the sound velocity image through iterative optimization. Based on different combinations of core physical parameters of the transducer, a series of corresponding sound velocity images are obtained.

[0082] In one embodiment, step three involves quantitatively evaluating the quality of the sound velocity image to determine the relationship between various core physical parameters and the system's imaging performance, thereby generating transducer design criteria to guide ultrasound tomography. Specifically:

[0083] (B1) Center frequency:

[0084] This embodiment establishes a two-dimensional ring array ultrasound tomography acoustic simulation model containing a ring array with a radius of approximately 100 mm and 512 elements. The simulation mesh is sufficient to support the highest frequency to be analyzed, ensuring simulation accuracy. To closely approximate clinical practice, the model also simulates the overall signal-to-noise ratio (SNR) conditions of an actual receiving circuit and designs a... Figure 5 The simulation shown includes a numerical breast phantom (i.e., a dedicated sound velocity phantom corresponding to the center frequency) containing various tissues (such as fat, glands, and lesions) and frequency-dependent sound attenuation. Figure 5 'a' represents a numerical simulation of the location and distribution of various types of tissues that may exist in an actual breast. Figure 5 b then shows Figure 5 The acoustic parameter distribution diagrams for each simulated tissue in 'a' are shown, where... Figure 5 b1 represents the sound velocity distribution of the phantom body; Figure 5 b2 represents the phantom density distribution; Figure 5 b3 is the sound attenuation coefficient. Location distribution; Figure 5 b4 is the power-law decay exponent. The location distribution of the transducer. By accurately simulating the tissue structure and acoustic characteristics of the real breast and realistically reproducing the signal-to-noise ratio constraints in actual imaging scenarios, this simulation can effectively reflect the imaging challenges faced by USCT in typical clinical scenarios. In this high-fidelity simulation environment, the center frequency of the transducer is systematically changed while keeping other physical parameters constant, and forward simulation and sound velocity reconstruction are performed for each frequency point.

[0085] Figure 6 The reconstructed sound velocity images and center profiles at different center frequencies are presented, clearly demonstrating the bidirectional influence of center frequency on image quality. At a low frequency of 1.0 MHz, due to the longer wavelength and limited spatial resolution, the boundaries of simulated lesions in the reconstructed images become blurred, such as... Figure 6 As shown in a1. However, in the high-frequency range of 3.0 MHz and above, increased acoustic attenuation leads to a deterioration in the signal-to-noise ratio and an increase in time-of-flight (TOF) extraction error. Figure 6 The reconstructed images at 3.0 MHz and 4.0 MHz corresponding to a3 and a4 both exhibit significant noise and artifacts, resulting in decreased quantitative accuracy. In contrast, the reconstruction results corresponding to a center frequency of 2.0 MHz ( Figure 6 a2) achieves a good balance between resolution and noise control: its resolution is better than that of 1.0 MHz, while effectively avoiding artifact problems caused by the decrease in signal-to-noise ratio in the frequency band above 3.0 MHz. Figure 6 b1 is the reference sound speed distribution. Figure 6 Figure b2 shows the sound velocity distribution curves of the reconstructed images at each frequency along the horizontal intercept (X direction) at Y=0. FromFigure 6 As can be further seen in b2, both excessively low and excessively high center frequencies cause the reconstructed sound velocity to deviate from the true value. However, in the 2.0–3.0 MHz range, the reconstructed sound velocity matches the true value more closely, reflecting the advantage of this frequency band in terms of imaging accuracy. Therefore, the results show that the system achieves the best balance in the mid-frequency band of 2.0 MHz to 3.0 MHz. Within this range, the conflict between wavelength and sound attenuation is effectively resolved, enabling the reconstructed image to possess the combined advantages of high resolution, high quantitative accuracy, and low artifact levels.

[0086] (B2) Number of array elements:

[0087] In time-of-flight-based sound velocity reconstruction, the traditional Nyquist sampling theorem is no longer a direct theoretical basis for determining the number of array elements. While increasing the number of elements can improve reconstruction quality by increasing ray coverage density, this improvement suffers from a saturation effect and leads to a sharp increase in hardware costs. Therefore, determining an optimal number of array elements is one of the key challenges in USCT system design. To address this, this embodiment introduces the more intuitive physical quantity of physical interval between sampling rays, defined as the physical distance between two adjacent receiving array elements at a specific propagation distance, originating from the same transmitting array element. It also creatively proposes a conjecture for selecting the number of array elements based on geometric relationships and the physical resolution limit: the sampling rays must be able to geometrically resolve the target sound velocity region. Figure 7 As shown, in order to accurately invert the contour of a target sound velocity region (i.e., the sound velocity region of interest), the sampling ray not only needs to have at least one ray passing through the region to obtain its internal information, but also needs rays passing through its sides to define its spatial boundaries.

[0088] The limiting physical resolution of acoustic velocity imaging based on the curved ray model is approximately three times the center wavelength, meaning the smallest detail size that the system can stably resolve is... , The center wavelength is given. Based on this, the following geometric sampling criterion is proposed: In order for the system to reliably reconstruct a wavelength of size , For the smallest resolvable target, the physical distance between the sampling rays should not exceed half the size of the target, i.e. The goal of this embodiment is to ensure that all regions within the ring array are uniformly and fully sampled and reconstructed, rather than optimizing only a specific region (such as the center point). To achieve this goal, the worst-case sampling conditions within the entire imaging field of view must be constrained.

[0089] Geometric analysis reveals that for any given transmission, the physical spacing of the sampling rays reaches its maximum value at the receiving end (i.e., the endpoint of the ray). Further analysis shows that among all possible rays, the largest end physical spacing occurs near the receiving element directly opposite the transmitting element. Therefore, by constraining this worst-case maximum physical spacing, the ray sampling density at all locations within the entire imaging field of view can be guaranteed to meet the requirements. The size of this maximum end physical spacing is equal to the spacing between the array elements on the array, determined by the total number of array elements (N) and the array diameter.

[0090] Based on the above derivation, this embodiment simplifies and equates a complex ray coverage problem that dynamically changes with spatial position to a single characteristic physical quantity (element spacing). The constraint problem of )). It is worth noting that although the constraint condition is similar in form (i.e., it imposes requirements on the spatial sampling interval) to the Nyquist sampling theorem in traditional structural imaging, its physical connotation and derivation path are completely different. The criterion of this embodiment is based on ensuring the geometric resolution of the minimum resolvable physical target of the system, rather than avoiding wavenumber domain aliasing, which is the focus of the Nyquist criterion. The two solve essentially different technical problems. Thus, a new quantitative relationship for guiding the configuration of the number of array elements can be obtained, as shown in Equation (1):

[0091] ,(1);

[0092] in, For the interval of array elements, The center wavelength is .

[0093] This principle directly links macroscopic geometric sampling to the theoretical limit of resolution, providing a new, physically-based guide for the selection of the number of array elements.

[0094] To rigorously verify the aforementioned innovative geometric sampling criterion, this embodiment designed a series of targeted numerical simulation experiments. The core of these experiments lies in constructing a critical test scenario that allows the system to operate near the theoretical resolution limit. For this purpose, a special cross-shaped lattice numerical phantom was designed and employed. The true sound velocity of the phantom is as follows: Figure 8 As shown in a1 and b1, the core idea of ​​this numerical phantom design is that the feature size (such as diameter) of its internal target is not arbitrarily set, but rather is related to the theoretical resolution limit at the center frequency used in the current simulation (for example, approximately three times the center wavelength). Precise correlation. This critical size design allows the numerical phantom to serve as a metric for directly testing the system's resolution under different sampling conditions. In this simulation, the physical spacing of the sampling rays (i.e., the element spacing) is precisely controlled by varying the number of elements (e.g., from 64 to 1024). ), and compare different The reconstruction results of the critical-size phantom are used to directly verify the effectiveness of the proposed geometric sampling criterion. Figure 8 Figures a2-a6 and b2-b6 show simulation results under two different parameter combinations (center frequencies of 1.25MHz and 2.5MHz, and corresponding critical phantom sizes), where a2 / b2 is the reconstructed image with 64 elements; a3 / b3 is the reconstructed image with 128 elements; a4 / b4 is the reconstructed image with 256 elements; a5 / b5 is the reconstructed image with 512 elements; and a6 / b6 is the reconstructed image with 1024 elements. These results strongly validate the correctness of the geometric sampling criterion proposed above and reveal four different stages of image quality variation with sampling density:

[0095] In areas with severe undersampling ( ,Right now Much larger Because the ray coverage is too sparse, the system struggles to effectively capture the target's boundary and structural information, resulting in severe aliasing or indistinguishable targets in the reconstructed images. For example... Figure 8 As shown in a2, at a center frequency of 1.25MHz and a number of array elements of 64, the target structure exhibits significant aliasing. This phenomenon is... Figure 8 Further verification was obtained in b2 and b3: when the center frequency is 2.5 MHz, the number of array elements is 64 ( Figure 8 b2) and 128 ( Figure 8 When b3), point targets in the cross-shaped dot matrix cannot be clearly distinguished, indicating that the imaging system has lost its ability to resolve fine structures under this sampling condition.

[0096] When the sampling ray spacing approaches the critical sampling region ( ,Right now Approximately equal to When, such as Figure 8 The a3 (center frequency 1.25 MHz, 128 elements) and Figure 8 As shown in b4 (center frequency 2.5 MHz, 256 elements), the point structure in the crosshair array can be clearly identified at this point. The fundamental reason is that, under this critical condition, the sampled rays can simultaneously cover the interior and edge regions of the target, thus achieving effective resolution of the target structure. Further analysis shows that the transition from severe undersampling to the critical sampling range is the key stage where the system's imaging performance is most significantly improved.

[0097] Enter the full sampling area ( ,Right now Less than Subsequently, as the radiation density continues to increase, more radiation passes through the same target area, leading to an improvement in the signal-to-noise ratio and a steady improvement in the accuracy of sound velocity quantification. For example... Figure 8 The a4 (1.25 MHz, 256 elements) and Figure 8 As shown in b5 (2.5 MHz, 512 elements), although image quality continues to improve in terms of signal-to-noise ratio and quantitative accuracy, the improvement in spatial resolution has slowed significantly, indicating that the system performance is gradually approaching the limit under the current configuration.

[0098] Finally, in the redundant sampling area ( ,Right now much smaller Since adjacent rays provide highly repetitive information, further increasing the number of array elements has very limited effect on improving image quality. For example... Figure 8 The a6 (center frequency 1.25 MHz, 1024 elements) and Figure 9 As shown in b6 (center frequency 2.5 MHz, 1024 elements), imaging performance has reached saturation, but hardware costs and computational burdens continue to rise, reflecting a significant decrease in the marginal benefits of system design at this stage. The above analysis indicates that simply pursuing extremely high sampling density (i.e., an extremely large number of elements) is uneconomical and unnecessary.

[0099] Therefore, this embodiment proposes the following optimization design criteria regarding the number of array elements: In order to achieve the best balance between imaging quality and engineering cost, the physical spacing of the sampling rays in the ring array USCT system (array element spacing) It should be optimized to be located near the critical sampling region, i.e. Specifically, the configuration of the minimum number of array elements should ensure... Not greater than This ensures the system can reach its theoretical resolution limit. Based on this, the number of array elements can be appropriately increased according to the requirements of quantitative accuracy, but redundant sampling areas should be avoided to prevent unnecessary cost waste. This principle provides a clear, quantitative, and physically-based guide for selecting the number of array elements in a practical USCT system.

[0100] (B3) Array element bandwidth;

[0101] To quantitatively verify the core argument proposed in this embodiment that the element bandwidth determines the accuracy of time-of-flight (TOF) measurements, a corresponding numerical simulation experiment was designed. This experiment was conducted on a phantom containing multiple circular high-speed targets to evaluate the system's resolution and noise immunity performance at different bandwidths. In the simulation, the relative operating bandwidth of the transducers was systematically varied, examining several typical values ​​from narrowband to wideband, and their sound velocity reconstruction results were compared. Furthermore, to explore the theoretical limits of performance and provide a theoretical benchmark for optimizing this parameter, this embodiment also includes an idealized ultra-wideband case as a performance reference for practical finite-bandwidth designs.

[0102] Figure 9 The simulation results of this embodiment are shown in an exemplary manner, wherein Figure 9 'a' represents the reconstructed sound velocity images under different bandwidths, 'a1' represents the reference sound velocity distribution, and 'a2'-'a5' represent the sound velocity reconstruction results under the conditions of 50%, 100%, 150%, and infinite bandwidth of the array elements, respectively. Figure 9 Figure b provides supplementary verification of noise robustness, where b1-b3 show the reconstruction error maps after applying two different noise instances with element bandwidths of 50%, 100%, and 150%, respectively. Simulation results reveal that increasing element bandwidth can bring about two stages of performance improvement.

[0103] Phase 1 (array element bandwidth increased from 50% to 100%), such as Figure 9 As shown in a2, when the bandwidth is 50%, the reconstructed boundary of the two-dimensional circular phantom exhibits significant blurring and distortion, and the overall image displays noise-like characteristics. After increasing the bandwidth to 100%... Figure 10 (a3), the above phenomenon is significantly suppressed, and the improvement in image quality is mainly reflected in the enhanced noise resistance. Figure 10 Further quantitative analysis revealed the distribution of the first arrival time (TIME) extraction results for the 128 array elements on the opposite side under different bandwidths, demonstrating that the stability of the TIME measurement of the received signal gradually becomes more stable and accurate with increasing bandwidth. Under 50% narrowband conditions ( Figure 9 The arrival time extraction of a1 exhibits severe random fluctuations with a standard deviation of 2.204, indicating that its arrival time extraction is highly susceptible to noise interference, leading to significant structural distortion and artifacts in the reconstructed image. Figure 9 (a2). When the bandwidth is increased to 100%, the extraction stability is significantly improved, the standard deviation decreases to 1.695, the jitter amplitude is significantly reduced, and the corresponding reconstructed image ( Figure 9 The a3 is clearer and has fewer artifacts.

[0104] To further verify the impact of bandwidth on noise robustness, this embodiment also conducted a supplementary experiment. In this experiment, for transducer arrays with three different bandwidths (50%, 100%, and 150%), two sets of random noise with the same standard deviation but different specific sequences were superimposed on their original noiseless radio frequency (RF) signals. Sound velocity reconstruction was then performed on the two sets of signals for each bandwidth. By calculating the difference between the two reconstructed sound velocity images under the same bandwidth, reconstruction error maps corresponding to the three bandwidths were obtained, as shown below. Figure 9 As shown in b1–b3. The results show that, at the same noise level, the reconstruction results of a 50% narrowband transducer ( Figure 9 b1) is most significantly affected by noise, with a wide error distribution; while the reconstruction results of the 150% broadband transducer ( Figure 10 b3) exhibits strong anti-interference capabilities; despite different noise sequences, the reconstructed images show minimal differences and no significant sound velocity deviation. This result strongly validates the significant advantages of broadband transducers in improving the noise resistance performance of sound velocity imaging.

[0105] The second stage (element bandwidth increased from 100% to infinity): At this stage, the change in imaging performance with element bandwidth is mainly reflected in the improvement of sound velocity image resolution. When the bandwidth is further increased from 100% to 150%, the TOF extraction results are as follows... Figure 9 As shown in a2 and a3, the standard deviation continued to decrease from 1.695 to 1.469, indicating that its extraction performance was still being optimized and was approaching stability, and interference from noise of the same level could be effectively suppressed. At this point, a further increase in bandwidth would make the time-domain pulse more compact, better conforming to the high-frequency approximation assumption of the acoustic signal in ray theory, thus leading to a further improvement in reconstruction resolution. Therefore, as... Figure 9 As shown in Figure a4, the reconstruction result with 150% bandwidth indicates that the boundary of the reconstructed target is sufficiently smooth and clear. To further explore the performance limits, this embodiment simulates the case of infinite bandwidth, which has theoretical reference value, and the reconstruction result is as follows. Figure 10 As shown in a5. Although its quantitative sound velocity value is similar to that of the 150% finite bandwidth result, it visually exhibits near-perfect boundary recovery capability and extremely high spatial resolution, while the corresponding TOF extraction results are as follows. Figure 11 As shown in a4, the extraction results are more stable than those of the aforementioned finite bandwidth transducers, with a standard deviation of 0.700. These simulation results not only validate the theoretical trend that wider bandwidth leads to higher resolution, but more importantly, they provide a gold standard for practical transducer design, serving as a measure of the resolution performance achieved by finite bandwidth transducers.

[0106] In summary, the simulation results of this embodiment confirm the correctness of the relevant theoretical analysis in step one: the wider the array element bandwidth, the more stable the time-of-flight (TOF) extraction, the stronger the anti-noise interference capability, and the higher the final sound velocity reconstruction quality.

[0107] Therefore, this embodiment proposes the following design criteria for array element bandwidth: In order to ensure high-precision time-of-flight (TOF) measurement and high-robust sound velocity reconstruction, wide-bandwidth transducer array elements should be selected as much as possible within the limits of manufacturing process and cost.

[0108] (B4) Array element width;

[0109] To verify the impact of element width on sound velocity reconstruction quality and imaging field of view, this embodiment employs a rectangular dot matrix phantom covering the region from the center to the edge. In this embodiment, the element width of the transducer is systematically varied, and the ideal dot matrix, approximately one center wavelength, is examined. and approximately two center wavelengths And so on. Figure 11 Examples of sound velocity reconstruction results with different element widths are shown, where a is the reference sound velocity distribution, b is the reconstruction result of an ideal point element, and c and d are the reconstruction results for an element width of 1.0 mm (approximately 1.0 mm). ) and 2.0 mm (approximately The reconstructed image when the array elements are ideal point-like ( Figure 11 (b) indicates that the structural fidelity of point targets is relatively high. As the array element aperture increases from point-like to 1 mm ( Figure 11 (c) Point targets near the outer edge of the array begin to exhibit tangential distortion, manifested as point structures being compressed tangentially. When the aperture is further increased to 2 mm ( Figure 12 (d) The reconstruction quality is significantly reduced, the tangential distortion in the edge region of the field of view is aggravated, and the sound velocity distribution is obviously distorted.

[0110] This embodiment argues that the fundamental reason for this technical effect lies in the severe mismatch between the directionality introduced by the finite aperture and the isotropic assumption of the reconstruction model. Physically, finite aperture transducers are most sensitive to normal incidence, while the reconstruction of the field of view edges relies on large-angle oblique incidence sound rays with extremely low transmission and reception efficiency. This directionality effect leads to a decrease in the signal-to-noise ratio or even signal loss in the edge region, making it impossible for the reconstruction algorithm to obtain effective time-of-flight (TOF) information, thus causing a sharp deterioration in tangential resolution. Although ideal point array elements can achieve the best model fit, they are unusable in practical engineering due to their excessively low signal-to-noise ratio.

[0111] Therefore, considering the trade-off between model matching (requiring a small aperture) and signal-to-noise ratio (requiring a large aperture), this embodiment proposes the following design criteria: In order to minimize the negative impact of directional effects while ensuring a sufficient signal-to-noise ratio, the element width should be optimized to approximately half to one center wavelength. Within the range.

[0112] (B5) Array element height and elevation focusing characteristics;

[0113] To quantitatively study the core trade-off between suppressing out-of-plane interference and avoiding volume averaging effects in the element height, this embodiment constructs a three-dimensional numerical simulation and designs a numerical phantom specifically for evaluating elevation angle resolution, such as... Figure 13 As shown. Its core design concept and goal lies in incorporating a structure that varies along the elevation direction (Z-axis), such as an out-of-plane artifact located outside the central plane, or a central critical-size target with a height matching the theoretical resolution. By observing whether out-of-plane artifacts appear in the in-plane reconstructed image, or the reconstruction fidelity of the central target in the Z-axis direction, the layer thickness effect and elevation resolution of the system can be intuitively and accurately evaluated. The key point of this phantom design is that it includes a size that matches the system's theoretical resolution limit (e.g., approximately three times the center wavelength). The precisely matched center critical size target, along with two out-of-plane interference bodies, allows it to serve as a metric unit to sensitively evaluate reconstruction fidelity under different element height configurations.

[0114] In this simulation environment, by systematically changing the element height of the transducer, the reconstructed result of its central cross-section is as follows: Figure 13 As shown, the results intuitively reveal the complex impact of element height on image quality: when the element height is too small (such as an ideal point element), the acoustic beam diverges severely in the elevation direction, leading to significant out-of-plane interference in the reconstructed image, specifically manifested as severe artifacts and quantitative distortion. This phenomenon can be observed in... Figure 13 Observed in a2. As the element height increases to ,like Figure 13 As shown in a3, out-of-plane interference is significantly suppressed, and the reconstructed sound velocity value of the central target tends to be more accurate. However, when the element height is further increased, Figure 13 The a4 data shows that although out-of-plane interference continues to decrease, an excessively large aperture will cause a significant volume averaging effect, which will severely blur the sound velocity information in the elevation direction, ultimately leading to a significant reduction in elevation resolution.

[0115] And such Figure 14As shown in the comparison results of the one-dimensional sound velocity intercept in b, the reconstructed sound velocity values ​​for the central target in the reconstruction results corresponding to infinitely small array elements and infinitely high array elements are seriously deviated from the true values. This is because whether the array element height is too small or too large, it will cause the transducer to detect a large number of sound signals far away from the central plane, which will pollute the TOF data used for reconstruction and ultimately lead to obvious distortion and aberration of the reconstructed sound velocity values.

[0116] Comprehensive analysis shows that, based on the principle of matching system performance, the optimal element height in this embodiment should match the system's limiting physical resolution in the elevation direction. Therefore, an element height approximately three times the center wavelength is selected. Using the size of the array element as its height allows for the optimal balance between these two conflicting objectives.

[0117] To further actively improve elevation resolution, this embodiment also verifies the effectiveness of elevation focusing. For example... Figure 14 As shown, by comparing non-focused (corresponding to) Figure 14 a) and focus (corresponding to) Figure 14 From the reconstruction results in cases b), it is clear that after introducing elevation angle focusing, the out-of-plane sound velocity artifact is suppressed, the reconstruction distortion range is reduced, and the quantitative value of sound velocity is closer to the true value. ​ The c represents the comparison result of the quantitative sound velocity reconstruction value of the X-axis intercept at the center position of the Y-axis. This more intuitively and clearly shows that the quantitative sound velocity reconstruction value is closer to the true value after introducing elevation angle focusing. This fully verifies the analytical conclusion that focusing the array element in the elevation direction can significantly suppress the divergence of the sound beam in the elevation direction, thereby effectively improving the elevation angle resolution.

[0118] In summary, element height and elevation focusing are key to optimizing USCT 3D imaging performance. Based on the above analysis, this embodiment proposes the following design principles: 1) The optimal element height of the ring array transducer should be designed to be approximately three times the center wavelength; 2) Based on this, elevation focusing characteristics should be introduced to further reduce the imaging layer thickness and suppress out-of-plane interference, thereby obtaining higher elevation resolution.

[0119] Through in-depth theoretical analysis and systematic numerical simulation verification of each core physical parameter (B1 to B5), this embodiment ultimately forms a set of transducer design guidelines for high-performance, ray-model-based ring array ultrasonic tomography velocity imaging systems. These guidelines aim to solve the complex, multi-dimensional performance trade-offs among various parameters, providing a scientific basis for optimizing imaging quality while meeting practical engineering constraints.

[0120] The design principles proposed in this embodiment are based on the core idea of ​​shifting from simply pursuing the limits of a single performance indicator to seeking the optimal balance among multiple performance indicators. For example, the choice of center frequency is no longer about blindly pursuing high resolution, but about finding a balance between resolution and signal-to-noise ratio; the configuration of the number of array elements also breaks free from the constraints of traditional Nyquist theory, turning to a new approach based on the matching of the system's physical resolution and geometric sampling. These principles together constitute a synergistic design framework that can effectively guide the development of next-generation ring array USCT systems.

[0121] Compared with the prior art, this embodiment has the following advantages:

[0122] Systematic and comprehensive: Unlike existing technical solutions that only analyze a single parameter, this embodiment comprehensively evaluates all key parameters of the transducer and their interactions, revealing its inherent coupling and trade-off mechanisms, and providing a more comprehensive and reliable design basis.

[0123] High design efficiency and accuracy: Through high-fidelity simulation and quantitative evaluation, this embodiment can predict the imaging performance under different parameter designs at a lower cost and higher efficiency before the product prototype is manufactured, avoiding expensive and time-consuming physical experiment iterations and improving the accuracy and success rate of the design.

[0124] Highly instructive: This embodiment outputs a set of optimal or balanced parameter combinations, rather than scattered patterns, which can directly guide engineering practice, shorten the transducer development cycle, and is expected to further improve the sound velocity imaging quality of the USCT system.

[0125] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A design method for an ultrasonic tomographic imaging transducer, characterized in that, include: Using the center frequency, number of array elements, array element bandwidth, array element aperture, and array element elevation angle focusing characteristics as the core physical parameters of the transducer, an acoustic simulation model of ring array ultrasonic tomography based on multi-dimensional physical characteristic analysis is constructed. By changing the values ​​of each core physical parameter, the model is forward simulated to generate raw radio frequency data on the impact of the core physical parameters on imaging quality. Perform sound velocity reconstruction based on a curved ray model to convert the raw radio frequency data into a sound velocity image; The quality of sound velocity images is quantitatively evaluated to determine the relationship between various core physical parameters and system imaging performance, thereby generating transducer design guidelines to guide ultrasound tomography systems. The transducer design criteria are as follows: (1) When quantifying the quality of the sound velocity image by changing the value of the number of array elements, the specific steps are as follows: Starting from the same transmitting element, the physical distance between the sampling rays formed by two adjacent receiving elements at the propagation distance is used as the element interval, where the sampling ray passes through the target sound speed region and does not pass through either side of the target sound speed region; By controlling the element spacing by varying the number of elements, and comparing the reconstruction results of the critical-size phantom under different element spacings, the design criterion for the minimum number of elements was obtained: the element spacing should not exceed [a certain value]. , The center wavelength; (2) Design criteria for center frequency: Take the sound velocity image of the mid-frequency band between the low-frequency band and the high-frequency band; (3) Design criteria for the array element bandwidth: Select wideband transducer array elements; (4) The array element aperture includes the array element width and the array element height. The design principle for the array element width is: the array element width is set at... Within the range; the design criteria for the element height: the optimal element height of the ring array transducer is designed to be three times the center wavelength.

2. The transducer design method according to claim 1, characterized in that, The quantitative evaluation of the sound velocity image quality involves changing the center frequency of the transducer while keeping other core physical parameters constant, and performing forward simulation and sound velocity reconstruction on the acoustic simulation model corresponding to each center frequency to obtain the design criteria for the center frequency.

3. The transducer design method according to claim 1, characterized in that, The design criteria for obtaining the minimum number of array elements are as follows: By sampling rays, the internal information of the target sound speed region and the spatial boundary of the target sound speed region are obtained, so as to find that the maximum end physical interval appears near the receiving array element directly opposite the transmitting array element. Based on the fact that the size of the maximum physical interval at the end is equal to the spacing between the array elements on the array, a complex ray coverage problem that dynamically changes with spatial position can be simplified and equivalently represented as a constraint problem on the spacing between array elements, wherein the spacing is determined by the total number of array elements and the array diameter. A quantitative relationship is obtained to guide the configuration of the number of array elements, and this quantitative relationship is used as the design criterion for the minimum number of array elements.

4. The transducer design method according to claim 1, characterized in that... The quality of the quantized sound velocity image is evaluated, wherein the array element bandwidth determines the measurement accuracy of the flight time.

5. The transducer design method according to claim 1, characterized in that, When generating raw RF data, the system systematically traverses each core physical parameter point in the physical parameter analysis space and performs acoustic forward simulation on the corresponding sound velocity simulation model to obtain the raw RF data corresponding to each core physical parameter.

6. A computer system comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the method according to any one of claims 1-5.

7. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a plurality of classification programs, which are used by a processor to execute the method as described in any one of claims 1-5.

8. A computer program product, comprising a computer program, characterized in that, The computer program is executed by a processor according to any one of claims 1-5.

Images