Sound velocity estimation method and device, storage medium and electronic equipment

Abstract

The invention discloses a sound velocity estimation method and device, a storage medium and electronic equipment, and the method comprises the steps: obtaining an echo signal corresponding to each array element in a linear array ultrasonic probe, and carrying out the beam synthesis of echo signals corresponding to a plurality of array elements at each preset imaging grid point for each candidate sound velocity contained in a preset sound velocity sequence. And for the depth of each layer of grid points in the axial direction, screening out an average sound velocity value corresponding to the depth of the layer of grid points from each candidate sound velocity according to signal coherence. And converting the average sound velocity value into an average slowness value to obtain the change information of the average slowness along with the depth. And converting the average slowness into the expression of local slowness, wherein the local slowness is a weighted combination of a plurality of step functions. And through solving a combination coefficient of the weighted combination, identifying potential sound velocity mutation nodes in the depth direction, obtaining information that the sound velocity changes along with the depth in the layered medium, and accurately determining the local sound velocity of each depth.

Description

Technical Field

[0001] This specification relates to the technical field of ultrasonic positioning, and particularly to a method and device for estimating sound speed, a storage medium, and an electronic device. Background Art

[0002] Medical ultrasound imaging is an indispensable imaging method in clinical diagnosis, but its imaging quality has long relied on the operator's experience judgment and qualitative analysis. Since different tissue types may present similar echo characteristics, this property leads to subjective differences in image interpretation and is prone to the risk of missed diagnosis and misdiagnosis. To improve the specificity of ultrasound imaging, the academic community has proposed measurement techniques for introducing tissue quantitative parameters (such as sound speed, elastic modulus, etc.). Among them, sound speed imaging technology can directly reflect tissue pathological characteristics (such as breast cancer differentiation, fatty liver staging), and can correct the propagation distortion effect of traditional ultrasound, showing significant clinical value.

[0003] Currently, the commonly used sound speed imaging technology is the pulse echo sound speed estimation method, which is mainly based on two principles: one is to iteratively optimize the sound speed parameters through B-mode ultrasound image quality indicators (such as speckle size, brightness), and the other is to estimate the average sound speed (ASS) using the maximum spatial coherence criterion, including the local spatial coherence maximization method and the short-lag spatial coherence technique, etc. Although the existing pulse echo sound speed estimation method can obtain the sound speed estimation value, it only reflects the average propagation speed of ultrasonic waves from the probe to the imaging area and cannot accurately characterize the local sound speed (LSS) distribution of tissues, resulting in errors in ultrasound imaging.

[0004] Based on this, this specification provides a method for estimating sound speed. Summary of the Invention

[0005] This specification provides a method and device for estimating sound speed, a storage medium, and an electronic device to at least partially solve the above problems existing in the prior art.

[0006] This specification adopts the following technical solutions: This specification provides a method for estimating sound speed, including: Obtaining the echo signals corresponding to each element in a linear array ultrasonic probe, where the echo signals are received by all elements; At each preset imaging grid point, for each candidate sound speed included in a preset sound speed sequence, performing beamforming on the echo signals corresponding to multiple elements, and calculating the signal coherence at the same imaging grid point at different transmission times according to the beamforming result; For each layer of grid point depth in the axial direction, screening out the average sound speed value corresponding to the depth of this layer of grid points from each candidate sound speed according to the signal coherence; Convert the average sound velocity values corresponding to the depths of each layer of grid points into average slowness values to obtain the information on the variation of average slowness with depth; Convert the average slowness into an expression of local slowness, where the local slowness is a weighted combination of a number of step functions; Determine the coefficient sequence of the weighted combination based on the average slowness values at the depths of each layer of grid points, and obtain the axial distribution information of the local sound velocity according to the obtained axial distribution information of the local slowness.

[0007] Optionally, at each preset imaging grid point, perform beamforming on the echo signals corresponding to multiple array elements for each candidate sound velocity included in the preset sound velocity sequence, and calculate the signal coherence at the same imaging grid point at different transmission times, specifically including: At each preset imaging grid point, for each candidate sound velocity included in the preset sound velocity sequence, calculate the beam domain image corresponding to the array element under this candidate sound velocity according to the echo signal corresponding to each array element; Determine the coherence factor between the echo signals corresponding to each array element under this candidate sound velocity according to the ratio of the coherent superposition result and the incoherent superposition result of the beam domain images corresponding to each array element under this candidate sound velocity, and the coherence factor is a transverse-axial-sound velocity three-dimensional coherence matrix.

[0008] Optionally, for each layer of grid point depth in the axial direction, screen out the average sound velocity value corresponding to the depth of this layer of grid points from each candidate sound velocity according to the signal coherence, specifically including: Average the transverse-axial-sound velocity three-dimensional coherence matrix in the axial direction to obtain a depth-sound velocity two-dimensional coherence matrix; For each layer of grid point depth in the axial direction, determine the average sound velocity value corresponding to the depth of this layer of grid points based on the candidate sound velocity corresponding to the maximum coherence factor in the corresponding depth of the depth-sound velocity two-dimensional coherence matrix.

[0009] Optionally, convert the average sound velocity values corresponding to the depths of each layer of grid points into average slowness values to obtain the information on the variation of average slowness with depth, specifically including: Respectively convert the average sound velocity values corresponding to the depths of each layer of grid points into average slowness values to obtain the initial slowness values corresponding to the depths of each layer of grid points; Successively for each layer of grid point depth in the axial direction, correct the initial slowness value corresponding to this layer of grid points according to the initial slowness value corresponding to the depth of the upper layer of grid points to obtain the average slowness value corresponding to this layer of grid points; Based on the average slowness values corresponding to the depths of each layer of grid points, obtain the information on the variation of average slowness with depth.

[0010] Optionally, for each depth of the grid points in the axial direction in sequence, based on the initial slowness value corresponding to the depth of the grid points in the previous layer, correct the initial slowness value corresponding to the grid points in this layer to obtain the average slowness value corresponding to the grid points in this layer, specifically including: For each depth of the grid points in the axial direction in sequence, estimate the upper limit of the average slowness of the depth of the grid points in this layer based on the initial slowness value corresponding to the depth of the grid points in the previous layer, the depth value of the grid points in the previous layer, the thickness value of the grid points in this layer, the preset lower limit of the sound speed, and the depth value of the grid points in this layer; estimate the lower limit of the average slowness of the depth of the grid points in this layer based on the initial slowness value corresponding to the depth of the grid points in the previous layer, the depth value of the grid points in the previous layer, the thickness value of the grid points in this layer, the preset upper limit of the sound speed, and the depth value of the grid points in this layer; Use the upper limit of the average slowness and the lower limit of the average slowness as constraint conditions to correct the average slowness value corresponding to the depth of the grid points in this layer.

[0011] Optionally, convert the average sound speed value corresponding to each depth of the grid points into an average slowness value to obtain the information of the average slowness varying with depth, specifically including: Convert the average sound speed value corresponding to each depth of the grid points into an average slowness value; Denoise the average sound speed value corresponding to each depth of the grid points by using a preset regularization matrix and adopting third-order difference regularization; Obtain the information of the average slowness varying with depth according to the denoising result.

[0012] Optionally, convert the average slowness into an expression of local slowness, specifically including: For each depth of the grid points in the axial direction in sequence, represent the average slowness value corresponding to the depth of the grid points in this layer as a weighted combination of the local slowness values corresponding to the depths of the grid points in the previous layer.

[0013] This specification provides a sound speed estimation device, and the device includes: A signal acquisition module that acquires the echo signal corresponding to each element in a linear array ultrasonic probe, and the echo signal is received by all elements; A beam synthesis module that performs beam synthesis on the echo signals corresponding to multiple elements for each candidate sound speed included in a preset sound speed sequence at each preset imaging grid point, and calculates the signal coherence at the same imaging grid point at different transmission times according to the beam synthesis result; An average sound speed determination module that screens out the average sound speed value corresponding to the depth of the grid points in this layer from each candidate sound speed according to the signal coherence for each depth of the grid points in the axial direction; An average slowness determination module that converts the average sound speed value corresponding to each depth of the grid points into an average slowness value to obtain the information of the average slowness varying with depth; A local slowness representation module that converts the average slowness into an expression of local slowness, where the local slowness is a weighted combination of a number of step functions; A solution module that determines the coefficient sequence of the weighted combination based on the average slowness value at each grid point depth, and obtains the axial distribution information of the local sound speed according to the obtained axial distribution information of the local slowness.

[0014] This specification provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the above-mentioned sound speed estimation method is implemented.

[0015] This specification provides an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the program, the above-mentioned sound speed estimation method is implemented.

[0016] At least one of the above technical solutions adopted in this specification can achieve the following beneficial effects: In the sound speed estimation method provided in this specification, by determining the information of the average sound speed varying with depth, and representing the local slowness as a weighted combination of a number of step functions, potential sound speed mutation nodes in the depth direction are identified. By solving the combination coefficients of the weighted combination, the information of the local slowness varying with depth is obtained, and the local slowness is converted into local sound speed, that is, the information of the sound speed varying with depth in the layered medium is obtained. Applying this method can accurately determine the local sound speed at each depth, and thus obtain a more accurate ultrasonic imaging. Description of the Drawings

[0017] The drawings described herein are used to provide a further understanding of this specification, and constitute a part of this specification. The schematic embodiments of this specification and their descriptions are used to explain this specification and do not constitute an improper limitation to this specification. In the drawings: Figure 1 It is a schematic flowchart of a sound speed estimation method in this specification; Figure 2 It is a specific implementation process of a sound speed estimation method in this specification; Figure 3 It is a schematic diagram of the sound domain used in the simulation in this specification; Figure 4 It is a distribution diagram of the average slowness varying with depth obtained by simulation in this specification; Figure 5 It is a distribution diagram of the local sound speed varying with depth obtained by simulation in this specification; Figure 6 It is a schematic diagram of a sound speed estimation device in this specification; Figure 7 Corresponding to what is provided in this specification Figure 1Schematic diagram of an electronic device. Detailed implementation manners

[0018] To make the objectives, technical solutions, and advantages of this specification clearer, the technical solutions of this specification will be clearly and completely described below in conjunction with specific embodiments of this specification and the corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this specification, rather than all the embodiments. Based on the embodiments in this specification, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of this application.

[0019] Existing acoustic velocity imaging technologies are mainly divided into two systems: ultrasound computed tomography (USCT or UCT) and pulse-echo ultrasound imaging. USCT is based on the propagation time of multi-angle ultrasonic waves to invert the acoustic velocity distribution, and a circumferential probe array is required to collect full-angle waveform data. Although this technology can achieve high-precision reconstruction, its complex hardware architecture and strict acquisition conditions limit the popularization of clinical applications.

[0020] In contrast, the pulse-echo ultrasound acoustic velocity estimation technology is more clinically adaptable, but faces the core challenge of missing position information of echo signals. Although such methods can obtain acoustic velocity estimation values, they only reflect the average propagation speed of ultrasonic waves from the probe to the imaging area and cannot accurately characterize the local acoustic velocity (LSS) distribution of tissues.

[0021] To obtain local acoustic velocity parameters and construct an LSS inversion model, a tomographic acoustic velocity imaging method is introduced into pulse-echo acoustic velocity imaging. A typical method is to establish an acoustic velocity tomography model through local echo phase information; or design an average compensation framework to achieve layered acoustic velocity estimation through the mathematical conversion from ASS to LSS. However, its iterative inversion process is vulnerable to noise interference, resulting in insufficient stability. In recent years, although the end-to-end acoustic velocity estimation method based on deep learning has made progress, its characteristic of relying on simulation data training makes its clinical generalization ability questionable. In specific clinical application scenarios (such as liver steatosis assessment), these existing technologies expose two key defects: 1) The stability of acoustic velocity estimation is sensitive to probe configuration parameters (such as transmission frequency, number of array elements), and the accuracy drops significantly when hardware conditions are limited; 2) Traditional tomographic inversion methods ignore the piecewise continuity characteristics of tissue acoustic velocity, resulting in non-physical oscillations in the acoustic velocity distribution in the depth direction.

[0022] This urges the industry to urgently develop an adaptive acoustic velocity imaging method that integrates prior constraints, which can improve the clinical applicability and measurement repeatability of LSS estimation while maintaining the simplicity of the pulse-echo system.

[0023] The following will detail the technical solutions provided by the embodiments of this specification in conjunction with the drawings.

[0024] Figure 1 This is a schematic flow diagram of a sound speed estimation method in this specification, specifically including the following steps: S100: Obtain the echo signals corresponding to each element in the linear array ultrasonic probe, and the echo signals are received by all elements.

[0025] This specification does not limit the acquisition method of the linear array ultrasonic probe signals. For example, the synthetic aperture scanning mode of transmitting element by element - receiving by all elements, the mode of transmitting by all elements synchronously - receiving by all elements, etc.

[0026] The reception time of the echo signals of the linear array ultrasonic probe is related to the signal propagation speed. Therefore, the echo signals corresponding to each element are the basic information for subsequent estimation of the sound speed of the ultrasonic signal.

[0027] Taking the synthetic aperture scanning mode as an example, each element serves as an independent emission source in turn to emit broadband ultrasonic pulses, and at the same time, all elements synchronously receive the reflected echo signals.

[0028] For each element included in the ultrasonic probe, take this element as the transmitting element, control this transmitting element to emit an ultrasonic pulse signal. After the ultrasonic pulse signal is emitted, the ultrasonic array switches to the signal reception mode, and all elements receive the echo pulse signals. This echo pulse signal is a three - dimensional data volume including the emission position, reception position, and reception time delay. At the th acquisition, the th element emits an ultrasonic pulse, and the echo pulse signal received by the th element is denoted as , where represents the reception time delay of the signal.

[0029] S102: At each preset imaging grid point, perform beamforming on the echo signals corresponding to multiple elements for each candidate sound speed included in the preset sound speed sequence, and calculate the signal coherence at the same imaging grid point at different emission times according to the beamforming result.

[0030] In ultrasonic imaging, the accuracy of the sound speed has an important impact on the imaging quality. In this method, first, according to the characteristics of the medium to be imaged, such as human tissues and structures, set a preset sound speed range. Then, according to the preset sound speed interval, determine multiple candidate sound speeds included in the preset sound speed range to obtain a sound speed sequence. For example, it can be between the preset sound speed range , with as the sound speed interval, divide it between the preset minimum sound speed and the preset maximum sound speed to obtain a sound speed sequence composed of multiple candidate sound speeds.

[0031] Since this method needs to determine the variation information of the sound speed along the depth in the medium to be imaged, each candidate sound speed is used to match the accurate sound speed of each depth layer of the imaging medium.

[0032] At each preset imaging grid point, for each candidate sound speed included in the preset sound speed sequence, beamforming of the echo signals corresponding to multiple array elements is performed to obtain the beam domain image of the ultrasonic signal.

[0033] Specifically, before beamforming, an imaging grid needs to be preset, and the imaging grid can be set according to the specifications of the ultrasonic probe. For example, taking the probe center of the ultrasonic probe as the reference, the probe width is used as the lateral width of the imaging grid, and the axial range of the ultrasonic probe is the imaging depth of the imaging grid. The grid size can be set according to requirements. The smaller the grid size, the higher the imaging resolution. For example, the grid size can be set to one-fourth of the wavelength of the ultrasonic beam.

[0034] For each candidate sound speed, according to the echo pulse signal collected by the ultrasonic pulse signal emitted by the m-th array element, a beam domain image corresponding to the m-th array element under the candidate sound speed is generated by using a beamforming algorithm.

[0035] This specification does not limit the specific beamforming algorithm used. For example, Delay and Sum Beamforming (DAS), Frequency Domain Beamforming (FDB), Linearly Constrained Minimum Variance (LCMV), etc.

[0036] Taking the DAS algorithm as an example, at the candidate sound speed the beam domain image corresponding to the m-th transmitting array element of the ultrasonic probe can be determined by the following formula:

[0037]

[0038] where and represent the lateral and axial coordinates of the grid point, and represent the lateral and axial coordinates of the -th array element, is the geometric propagation delay, indicating the time for the ultrasonic pulse signal emitted by the m-th transmitting array element to propagate to the imaging grid ) and then continue to propagate and be received by the n-th array element.

[0039] Finally, a full-aperture data set can be generated , the full-aperture data set is a four-dimensional tensor, M is the total number of array elements in the ultrasonic probe. The full-aperture data set is obtained after the collected original echo pulse signal is compensated by time gain, providing complete wave propagation information for subsequent processing.

[0040] Then, based on the beam domain image corresponding to each array element at the candidate sound speed, the coherence factor between the echo signals corresponding to each array element at the candidate sound speed can be calculated. The coherence factor is a three-dimensional coherence matrix of transverse-axial-sound speed, which is used to reflect the coherence between the echo signals corresponding to each array element, that is, the transmission channels.

[0041] Specifically, the coherence factor between the echo signals corresponding to each array element at the candidate sound speed is determined according to the ratio of the coherent superposition result of the beam domain images corresponding to each array element at the candidate sound speed to the incoherent superposition result of the beam domain images corresponding to each array element.

[0042] For each candidate sound speed , calculate the candidate sound speed The beam domain image corresponding to each array element under The coherence factor between them is It represents the horizontal and vertical coordinates of the grid points of the preset imaging grid, and the coherence factor is the three-dimensional coherence matrix of transverse-axial-sound speed.

[0043] Specifically, when the speed of sound is In the case of, the corresponding coherence factor (CF) is calculated according to the following formula:

[0044] In the above formula, the numerator is the coherent superposition result of the beam domain images of each array element, that is, the power of the coherent superposition of the signals between the transmit channels, and the denominator is the incoherent superposition result of the beam domain images of each array element, that is, the sum of the independent powers of each transmit channel.

[0045] S104: for each grid point depth in the axial direction, select an average sound speed value corresponding to the grid point depth from each candidate sound speed according to the signal coherence.

[0046] Because the coherence factor obtained above is a three-dimensional coherence matrix of transverse-axial-sound speed, in order to eliminate the influence of the transverse structure on the sound speed and determine the variation information of the sound speed with depth in the layered medium, the transverse-axial-sound speed three-dimensional coherence matrix of the coherence factor is horizontally averaged according to the total number of grid points in the transverse direction of the preset imaging grid to obtain a two-dimensional coherence matrix of depth-sound speed.

[0047] In the depth-speed two-dimensional coherence matrix, the closer the CF value is to 1, the higher the phase consistency of the signals in each transmission channel is, and the corresponding speed is The closer it is to the true sound speed.

[0048] Therefore, for the depth of each grid point in the axial direction, based on the candidate sound speed corresponding to the maximum coherence factor in the corresponding depth of the depth-sound speed two-dimensional coherence matrix, the average sound speed value corresponding to the depth of this layer of grid points is determined. Based on the average sound speed values corresponding to the depths of each layer of grid points, the information on the variation of the average sound speed with depth is obtained.

[0049] Specifically, along the axial grid depth direction, the candidate sound speed values corresponding to the maximum coherence factor can be searched layer by layer to determine the estimated sound speed corresponding to the depth of each layer of grid. That is, for each grid depth , all candidate sound speeds are traversed , and the candidate sound speed that makes the largest is selected .

[0050] This process can be expressed by the following formula:

[0051] Among them, is the estimated sound speed corresponding to the j-th grid depth, and the estimated sound speed is the average sound speed from the grid top to the j-th grid depth.

[0052] Furthermore, in one embodiment, after obtaining the depth-sound speed two-dimensional coherence matrix, in order to improve the accuracy of the estimated sound speed corresponding to each layer of grid depth determined, bilinear interpolation can be performed on the sound speed based on the depth-sound speed two-dimensional coherence matrix to improve the sound speed resolution. For example, if the sound speed interval in S100 is initially set to 2 m / s, it can be increased to 0.1 m / s after interpolation .

[0053] S106: Convert the average sound speed values corresponding to the depths of each layer of grid points into average slowness values to obtain the information on the variation of the average slowness with depth.

[0054] For the convenience of subsequent cumulative processing of the sound speed in the time dimension, starting from this step, the sound speed is converted into slowness for subsequent operations. Then the information on the variation of the average sound speed with depth obtained above is also correspondingly converted into the information on the variation of the average slowness with depth. Slowness is the reciprocal of speed, with the unit of s / m, and slowness is linearly related to time, which is convenient for subsequent construction of a linear model for solution.

[0055] This step can be expressed by the following formula:

[0056] Among them, is the average slowness corresponding to the j-th layer of grid depth.

[0057] S108: Convert the average slowness into an expression of local slowness, where the local slowness is a weighted combination of a number of step functions.

[0058] Based on the stratified propagation model, establish a sound speed inversion framework, and achieve the accurate calculation of ASS through sparse constraint optimization. First, discretize the axial detection area into a multi-layer structure, and establish a linear mapping model according to the transfer relationship between LSS and ASS.

[0059] For each grid point depth in the axial direction in turn, represent the average slowness value corresponding to the depth of the grid point in this layer as a weighted combination of the local slowness values corresponding to the grid point depths in the previous layer.

[0060] Specifically, according to the following formula, represent the average slowness of the nth layer grid point depth as the weighted sum of the local slownesses corresponding to the grid point depths of the first n layers:

[0061] where, represents the average slowness vector, represents the local slowness vector, is the transformation matrix, , represents the nth layer grid depth.

[0062] It can be seen that the transformation matrix is a lower triangular matrix. For the element in the nth row and kth column of , its value can be expressed as follows:

[0063] Through this transformation matrix T, represent the average slowness of the nth layer as the weighted sum of the local slownesses of the first n layers, is the layer thickness of the nth layer grid depth , so the weight corresponding to the local slowness of each layer is the proportion of the layer thickness .

[0064] S108: Determine the coefficient sequence of the weighted combination through the average slowness value of each grid point depth, and obtain the axial distribution information of the local sound speed according to the obtained axial distribution information of the local slowness.

[0065] Based on the piecewise constant characteristic of the sound speed distribution in medical tissues, construct a sparse representation space composed of step function bases, and represent the local sound speed as a weighted combination of a finite number of step changes.

[0066] The local slowness corresponding to the nth layer grid depth can be expressed by the following formula:

[0067]

[0068] Among them, is a piecewise constant basis function, is a threshold. This threshold can be set as needed, for example, it can be set to two wavelengths.

[0069] Based on the conversion relationship between the local slowness of each grid depth and the piecewise constant basis function, a sparse representation model can be obtained, and its matrix form is .

[0070] Among them, the element in the matrix , and its meaning is whether the grid depth of the j-th layer exceeds the threshold of the i-th basis function . The coefficient matrix represents the sparse coefficients of the local sound speed distribution, The elements of are mostly close to 0, and can represent the amplitude and position information of the piecewise change of the sound speed with depth.

[0071] Up to this step, for the local sound speed of each grid depth, it is the quantity to be solved. By solving the coefficient matrix , a coefficient sequence of the weighted combination of step functions can be obtained, and the axial distribution information of the piecewise constant local sound speed that conforms to the characteristics of biological tissues can be inversely calculated.

[0072] In this specification, the coefficient matrix can be solved by the least squares method. The specific solution method is as follows: First, construct a least squares optimization function: .

[0073] Second, solve the above optimization function, and the coefficient matrix can be obtained. Any existing least squares solution method can be used for solving. In this specification, the fast iterative shrinkage-thresholding algorithm (FISTA) is used as an example to illustrate the solution steps.

[0074] The solution steps are as follows. Initialize .

[0075] Then, iteratively update ( ):

[0076]

[0077]

[0078] Among them, is the soft threshold operation with a threshold of , which is used to sparsify the coefficients. is the iteration step size. is the acceleration factor, which is used to accelerate convergence.

[0079] After iterating K times, the solution result of the coefficient sequence is obtained .

[0080] The coefficient sequence is obtained. After that, according to , the local slowness matrix can be obtained. By converting the local slowness matrix into local sound velocity, the axial distribution information of the local sound velocity can be obtained.

[0081] According to Figure 1 the method described above, first construct a multi-sound-velocity beamforming system, generate the coherence factor feature maps at different candidate sound velocities through the beamforming algorithm, extract the average slowness based on the maximum coherence criterion, express the local slowness as a linear combination of a finite number of step functions, construct a sparse constraint inversion framework through the compressive sensing theory, transform the sound velocity boundary location into an optimization problem with a regularization term, and use the FISTA algorithm to achieve the efficient solution of the layered interface.

[0082] The standardized full-process design and deterministic algorithm architecture ensure the high consistency of the measurement results. The full-aperture data acquisition scheme suppresses random errors through the spatial synthesis effect of multi-channel signals, eliminates the differences in manual parameter adjustment with the fixed parameter settings, reduces the fluctuation range of the repeated test results, and meets the stringent requirements of clinical diagnosis for repeatability.

[0083] In the above step S102, based on the prior sound velocity constraint mechanism, the effective sound velocity range of the current layer can also be dynamically adjusted according to the upper-layer sound velocity estimation result, and the hard threshold truncation strategy is used to eliminate the abnormal estimated values in the estimated average sound velocity.

[0084] First, the average sound velocity values corresponding to the depths of the grid points of each layer are respectively converted into average slowness values to obtain the initial slowness values corresponding to the depths of the grid points of each layer.

[0085] That is, according to the method in the above step S102, the candidate sound velocity corresponding to each layer grid depth determined based on the maximum coherence factor is used as the initial sound velocity value to be corrected corresponding to each layer grid depth. Then, according to the reciprocal of the initial sound velocity value, the initial sound velocity value is converted into an initial slowness value.

[0086] Then, for each grid point depth layer in the axial direction, based on the initial slowness value corresponding to the grid point depth of the previous layer, the initial slowness value corresponding to the grid points of this layer is corrected to obtain the average slowness value corresponding to the grid points of this layer.

[0087] Specifically, for each grid point depth layer in the axial direction, based on the initial slowness value corresponding to the grid point depth of the previous layer, the depth value of the previous layer grid point, the thickness value of this layer grid point, the preset lower limit value of the sound velocity, and the depth value of this layer grid point, the upper limit value of the average slowness of this layer grid point depth is estimated; based on the initial slowness value corresponding to the grid point depth of the previous layer, the depth value of the previous layer grid point, the thickness value of this layer grid point, the preset upper limit value of the sound velocity, and the depth value of this layer grid point, the lower limit value of the average slowness of this layer grid point depth is estimated. Taking the upper limit value of the average slowness and the lower limit value of the average slowness as constraint conditions, the average slowness value corresponding to the grid point depth of this layer is corrected.

[0088] For the nth layer grid along the depth direction of the preset imaging grid, based on the initial slowness corresponding to the grid depth of the (n - 1)th layer, the initial slowness of the grid depth of the nth layer is corrected to obtain the average slowness corresponding to the grid depth of the nth layer.

[0089] The following is the recurrence relation of the initial slowness with depth:

[0090] Among them, represents the initial slowness corresponding to the grid depth of the nth layer, represents the initial slowness corresponding to the grid depth of the (n - 1)th layer, represents the grid depth of the nth layer, represents the grid depth of the (n - 1)th layer, represents the layer thickness of the nth layer, .

[0091] The numerator in the recurrence relation is represents the total slowness integral of the grid depths of the first (n - 1) layers, and the sum with represents the slowness integral of the current nth layer. The denominator is the total depth of the nth layer grid , and this recurrence relation can represent the overall average slowness from the surface to the current grid depth layer .

[0092] The local slowness of the nth grid depth is the quantity to be estimated. In this embodiment, according to prior knowledge, a sound velocity constraint is introduced to limit the local slowness . Ensure that the local sound velocity of the current layer falls within the range , that is, the local slowness satisfies .

[0093] It can be deduced that for the nth grid depth, when , the sound speed is the slowest, and the upper limit value of the average slowness is obtained, that is, the maximum allowable average slowness . When , the sound speed is the fastest, and the lower limit value of the average slowness is obtained, the minimum allowable average slowness .

[0094] Based on this, the upper limit value of the average slowness and the lower limit value of the average slowness can be used as constraint conditions to construct the hard upper and lower bounds of the average slowness of this grid depth layer. According to the following formula:

[0095] The initial slowness of the nth layer of grid depth is corrected to obtain the average slowness corresponding to the nth layer of grid depth.

[0096] It can be seen from the hard upper and lower bounds that this embodiment is a layer-by-layer dynamic correction process. The initial slowness of each layer of grid depth is corrected depending on the average slowness obtained after correcting the initial slowness of the previous layer, so that the average slowness of the current layer is within a reasonable range and outliers are suppressed.

[0097] Through this embodiment, a prior sound speed constraint mechanism and a dynamic compensation strategy are introduced, effectively overcoming the technical bottleneck that the traditional method is sensitive to hardware parameters. The hard threshold truncation design limits the sound speed estimation values at each depth to a physically reasonable interval, blocking the longitudinal transmission path of errors; the third-order derivative regularization processing ensures the continuity of the sound speed curve by suppressing non-smooth fluctuations. Under extreme conditions such as low-frequency probes or missing data in some channels, this method can still maintain a stable sound speed output, showing stronger environmental adaptability compared with the traditional coherence factor method.

[0098] By iteratively correcting the systematic offset of the average sound speed curve, the estimation accuracy of the shallow layer is significantly improved. The sparse regularization model based on tissue characteristics accurately captures the sound speed mutation interface and shows better interlayer resolution ability in multi-layer medium scenarios.

[0099] In another embodiment, in step S104, the third-order derivative regularization method can also be used to smooth the sound speed curve and suppress noise disturbances.

[0100] Specifically, through a preset regularization matrix , third-order difference regularization is used to denoise the average slowness corresponding to the nth layer of grid depth. According to the denoising result, the information of the change of the average slowness with depth is obtained.

[0101] For the average slowness of the nth-layer grid depth, third-order difference regularization is used to suppress noise perturbations. The regularization process is specifically expressed as follows:

[0102] Wherein, is the smoothness parameter, is the regularization matrix, The specific form of is as follows:

[0103] What is obtained in this embodiment is the final estimated average slowness-depth curve compensated by prior constraint.

[0104] Next, in combination with Figure 2 The following shows the embodiment process to illustrate the specific implementation manner of the present invention with specific ultrasonic echo experimental data. The embodiment of the present invention uses a simulation method to implement the estimation process of layered sound velocity. Specifically, this method is implemented by reconstructing the sound velocity from the measurement data of the two-dimensional acoustic simulation of a numerical phantom. In the simulation, a linear array ultrasonic transducer is used, which includes 80 uniformly arranged ultrasonic transceiver elements. The center frequency of each element is 2.7 MHz, the width is 0.206 mm, the element spacing is 0.254 mm, and the fractional bandwidth of the element frequency is 74%. These transceiver elements emit sound pulses in sequence and scan the numerical phantom in turn. The entire data acquisition process includes 80 emission steps, and each element emits alone in this way, and all the other transducers are responsible for recording the acoustic wave field signals.

[0105] The sound velocity in the simulation domain is shown in Appendix Figure 3 , Figure 3 which is the schematic diagram of the sound domain used in the simulation in this specification. Figure 3 The left and right in respectively represent two simulation sound domains. z represents the axial direction, x represents the transverse direction. On the left are 2 layers of media (1500 m / s and 1540 m / s from top to bottom), and on the right are 3 layers of media (1500 m / s, 1580 m / s and 1540 m / s from top to bottom). Different media are marked with different shaped backgrounds, and the medium density is 1000 kg / m 3 . To simulate random speckles, Gaussian white noise with a standard deviation of 3 m / s is added to the density. The spatial and temporal grid settings in the simulation are shown in Table 1.

[0106] Table 1 Simulation parameters

[0107] Above, the simulation settings are completed.

[0108] Execute Step 1: In the simulation environment, in accordance with the sequential emission mode, use the ultrasonic probe to sequentially emit ultrasonic pulses and collect the acoustic pressure signals to obtain the RF signals of the echoes. Collect 10 times repeatedly to evaluate the repeatability of the method.

[0109] Execute Step 2: Preset the imaging grid before beam synthesis, with a lateral width of 25 mm and an axial depth of 5 mm - 95 mm. Generate a candidate sound speed sequence based on a preset sound speed range of 1460 m / s - 1620 m / s (step size 2 m / s). Perform DAS beam synthesis on each candidate value, calculate the signal coherence between the transmitting channels, and construct a three-dimensional coherence matrix of lateral - axial - sound speed. Generate a depth - sound speed two-dimensional coherence feature map by axial stratification and lateral averaging, combine the upper-layer sound speed estimation to dynamically constrain the effective sound speed range of the current layer, and use a hard threshold to truncate outliers. Further, use third-order derivative regularization to iteratively smooth the sound speed curve, and the regularization parameter is 1e7. Obtain the average slowness distribution as shown in the appendix Figure 4 , Figure 4 is the distribution diagram of the average slowness varying with depth obtained by simulation in this specification, where the solid line represents the true sound speed value, and each dashed line represents the experimental result of one time. Figures (a) and (b) respectively correspond to Figure 3 the two media in

[0110] Execute Step 3: Construct a sparse representation space composed of step function bases, with an interval of two wavelength intervals, and the regularization parameter is set to 0.01. The final output axial LSS distribution curve is as shown in the appendix Figure 5 , Figure 5 is the distribution diagram of the local sound speed varying with depth obtained by simulation in this specification, where the solid line represents the true sound speed value, and the dashed line represents the estimated result of the local sound speed varying with depth. Figures (a) and (b) respectively correspond to Figure 3 the two media in Figure 5 It can be seen from this appendix

[0111] Figure 6

[0112] Figure 6 This is a schematic diagram of a sound speed estimation device provided in this specification, specifically including: A signal acquisition module 200, configured to obtain the echo signals corresponding to each element in the linear array ultrasonic probe, and the echo signals are received by all elements; The beam synthesis module 202 is configured to perform beam synthesis on the echo signals corresponding to multiple array elements for each candidate sound speed included in the preset sound speed sequence at each preset imaging grid point, and calculate the signal coherence at the same imaging grid point at different transmission times according to the beam synthesis result; The average sound speed determination module 204 is configured to screen out the average sound speed value corresponding to the depth of each layer of grid points in the axial direction from each candidate sound speed according to the signal coherence; The average slowness determination module 206 is configured to convert the average sound speed values corresponding to the depths of each layer of grid points into average slowness values to obtain the information on the variation of the average slowness with depth; The local slowness representation module 208 is configured to convert the average slowness into an expression of local slowness, where the local slowness is a weighted combination of several step functions; The solution module 210 is configured to determine the coefficient sequence of the weighted combination through the average slowness values at the depths of each layer of grid points, and obtain the axial distribution information of the local sound speed according to the obtained axial distribution information of the local slowness.

[0113] Optionally, the average sound speed determination module 204 is specifically configured to, at each preset imaging grid point, for each candidate sound speed included in the preset sound speed sequence, calculate the beam domain image corresponding to the array element under the candidate sound speed according to the echo signal corresponding to each array element, and determine the coherence factor between the echo signals corresponding to each array element under the candidate sound speed according to the ratio of the coherent superposition result and the incoherent superposition result of the beam domain images corresponding to each array element under the candidate sound speed, where the coherence factor is a transverse-axial-sound speed three-dimensional coherence matrix.

[0114] Optionally, the average sound speed determination module 204 is specifically configured to perform averaging on the transverse-axial-sound speed three-dimensional coherence matrix in the axial direction to obtain a depth-sound speed two-dimensional coherence matrix, and for the depth of each layer of grid points in the axial direction, determine the average sound speed value corresponding to the depth of the layer of grid points based on the candidate sound speed corresponding to the maximum coherence factor in the corresponding depth of the depth-sound speed two-dimensional coherence matrix.

[0115] Optionally, the average slowness determination module 206 is specifically configured to convert the average sound speed values corresponding to the depths of each layer of grid points into average slowness values respectively to obtain the initial slowness values corresponding to the depths of each layer of grid points, and sequentially for the depth of each layer of grid points in the axial direction, correct the initial slowness value corresponding to the layer of grid points according to the initial slowness value corresponding to the depth of the upper layer of grid points to obtain the average slowness value corresponding to the layer of grid points, and obtain the information on the variation of the average slowness with depth based on the average slowness values corresponding to the depths of each layer of grid points.

[0116] Optionally, the average slowness determination module 206 is specifically configured to, for each grid point depth in the axial direction in sequence, estimate the upper limit value of the average slowness of the grid point depth according to the initial slowness value corresponding to the grid point depth of the previous layer, the grid point depth value of the previous layer, the thickness value of the grid point of this layer, the preset lower limit value of the sound velocity, and the grid point depth value of this layer, and estimate the lower limit value of the average slowness of the grid point depth according to the initial slowness value corresponding to the grid point depth of the previous layer, the grid point depth value of the previous layer, the thickness value of the grid point of this layer, the preset upper limit value of the sound velocity, and the grid point depth value of this layer, and use the upper limit value of the average slowness and the lower limit value of the average slowness as constraint conditions to correct the average slowness value corresponding to the grid point depth of this layer.

[0117] Optionally, the average slowness determination module 206 is specifically configured to convert the average sound velocity value corresponding to each layer of grid point depth into an average slowness value, perform denoising on the average sound velocity value corresponding to each layer of grid point depth by using third-order difference regularization through a preset regularization matrix, and obtain the information on the change of the average slowness with depth according to the denoising result.

[0118] Optionally, the local slowness representation module 208 is specifically configured to, for each grid point depth in the axial direction in sequence, represent the average slowness value corresponding to the grid point depth of this layer as a weighted combination of the local slowness values corresponding to the grid point depths of the previous layers.

[0119] This specification also provides a computer-readable storage medium storing a computer program, and the computer program can be used to execute the above Figure 1 provided sound velocity estimation method.

[0120] This specification also provides Figure 7 a schematic structural diagram of the electronic device shown. At the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory. Of course, it may also include other hardware required for other services. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to implement the above Figure 1 described sound velocity estimation method. Of course, in addition to the software implementation manner, this specification does not exclude other implementation manners, such as the manner of logical devices or the combination of software and hardware, etc. That is to say, the execution subject of the following processing flow is not limited to each logical unit, and may also be hardware or logical devices.

[0121] The improvement of a technology can be clearly distinguished as either a hardware improvement (e.g., improvement of circuit structures such as diodes, transistors, switches, etc.) or a software improvement (improvement of method processes). However, with the development of technology, many improvements of method processes today can be regarded as direct improvements of hardware circuit structures. Almost all designers obtain the corresponding hardware circuit structure by programming the improved method process into the hardware circuit. Therefore, it cannot be said that an improvement of a method process cannot be implemented with a hardware entity module. For example, a programmable logic device (PLD) (such as a field programmable gate array (FPGA)) is such an integrated circuit whose logic function is determined by the user programming the device. The designer can program by himself to "integrate" a digital system on a single PLD, without having to ask the chip manufacturer to design and fabricate a dedicated integrated circuit chip. Moreover, nowadays, instead of manually fabricating integrated circuit chips, this programming is mostly implemented using "logic compiler" software, which is similar to the software compiler used in program development and writing. The original code before compilation also has to be written in a specific programming language, which is called a hardware description language (HDL), and there is not only one kind of HDL, but many kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware Description Language), Confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), Lava, Lola, MyHDL, PALASM, RHDL (Ruby Hardware Description Language), etc. The most commonly used ones currently are VHDL (Very-High-Speed Integrated Circuit Hardware Description Language) and Verilog. Those skilled in the art should also be aware that as long as the method process is slightly logically programmed with the above-mentioned several hardware description languages and programmed into the integrated circuit, it is easy to obtain the hardware circuit that implements the logical method process.

[0122] The controller can be implemented in any suitable manner. For example, the controller can take the form of, for example, a microprocessor or a processor and a computer-readable medium storing computer-readable program code (such as software or firmware) executable by the (micro)processor, logic gates, switches, an application specific integrated circuit (ASIC), a programmable logic controller, and an embedded microcontroller. Examples of the controller include, but are not limited to, the following microcontrollers: ARC 625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320. The memory controller can also be implemented as part of the control logic of the memory. Those skilled in the art also know that in addition to implementing the controller in the form of pure computer-readable program code, it is entirely possible to logically program the method steps to enable the controller to be implemented in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. to achieve the same function. Therefore, such a controller can be considered a hardware component, and the devices included therein for implementing various functions can also be regarded as the structures within the hardware component. Or even, the devices for implementing various functions can be regarded as either software modules for implementing the method or structures within the hardware component.

[0123] The systems, devices, modules, or units illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0124] For the convenience of description, when describing the above devices, they are described separately as various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0125] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0126] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowchart and / or block diagram can also be implemented. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a means for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0127] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction means, and the instruction means implements the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0128] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one Figure 1 flow or multiple flows and / or blocks Figure 1 or multiple blocks.

[0129] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and a memory.

[0130] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.

[0131] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0132] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0133] It should be understood by those skilled in the art that the embodiments of this specification may be provided as methods, systems or computer program products. Therefore, this specification may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware. Moreover, this specification may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0134] This specification may be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules may be located in local and remote computer storage media, including storage devices.

[0135] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other, and the key point of each embodiment is to illustrate the differences from other embodiments. In particular, for the system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and for the relevant parts, reference can be made to the partial description of the method embodiment.

[0136] The above description is only for the embodiments of this specification and is not intended to limit this specification. For those skilled in the art, various modifications and changes can be made to this specification. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this specification shall be included within the scope of the claims of this application.

Claims

1. A method for estimating the speed of sound, characterized in that Including: Obtaining echo signals corresponding to each array element in a linear array ultrasonic probe, where the echo signals are received by all array elements; At each preset imaging grid point, for each candidate sound speed included in a preset sound speed sequence, performing beamforming on the echo signals corresponding to multiple array elements, and calculating the signal coherence at the same imaging grid point at different transmission times according to the beamforming result; For each layer of grid point depth in the axial direction, screening out the average sound speed value corresponding to the depth of this layer of grid points from each candidate sound speed according to the signal coherence; Converting the average sound speed values corresponding to each layer of grid point depth into average slowness values to obtain information on the variation of average slowness with depth; Converting the average slowness into an expression of local slowness, where the local slowness is a weighted combination of several step functions; Determining the coefficient sequence of the weighted combination through the average slowness value of each layer of grid point depth, and obtaining the axial distribution information of local sound speed according to the obtained axial distribution information of local slowness.

2. The method according to claim 1, wherein At each preset imaging grid point, for each candidate sound speed included in a preset sound speed sequence, performing beamforming on the echo signals corresponding to multiple array elements, and calculating the signal coherence at the same imaging grid point at different transmission times, specifically including: At each preset imaging grid point, for each candidate sound speed included in a preset sound speed sequence, calculating the beam domain image corresponding to this array element at this candidate sound speed according to the echo signal corresponding to each array element; Determining the coherence factor between the echo signals corresponding to each array element at this candidate sound speed according to the ratio of the coherent superposition result and the incoherent superposition result of the beam domain images corresponding to each array element at this candidate sound speed, where the coherence factor is a transverse-axial-sound speed three-dimensional coherence matrix; 3. The method according to claim 2, characterized in that, For each layer of grid point depth in the axial direction, screening out the average sound speed value corresponding to the depth of this layer of grid points from each candidate sound speed according to the signal coherence, specifically including: Averaging the transverse-axial-sound speed three-dimensional coherence matrix in the axial direction to obtain a depth-sound speed two-dimensional coherence matrix; For each layer of grid point depth in the axial direction, based on the candidate sound speed corresponding to the maximum coherence factor in the corresponding depth of the depth-sound speed two-dimensional coherence matrix, determining the average sound speed value corresponding to the depth of this layer of grid points.

4. The method according to claim 1, characterized in that Converting the average sound speed values corresponding to each layer of grid point depth into average slowness values to obtain information on the variation of average slowness with depth, specifically including: Respectively converting the average sound speed values corresponding to each layer of grid point depth into average slowness values to obtain the initial slowness values corresponding to each layer of grid point depth; Successively for each layer of grid point depth in the axial direction, correcting the initial slowness value corresponding to this layer of grid points according to the initial slowness value corresponding to the previous layer of grid point depth to obtain the average slowness value corresponding to this layer of grid points; Based on the average slowness values corresponding to each layer of grid point depth, obtaining information on the variation of average slowness with depth.

5. The method according to claim 4, characterized in that, Successively for each layer of grid point depth in the axial direction, correcting the initial slowness value corresponding to this layer of grid points according to the initial slowness value corresponding to the previous layer of grid point depth to obtain the average slowness value corresponding to this layer of grid points, specifically including: For each depth of the grid points in the axial direction in sequence, estimate the upper limit of the average slowness at the depth of the grid points of this layer based on the initial slowness value corresponding to the depth of the grid points in the previous layer, the depth value of the grid points in the previous layer, the thickness value of the grid points in this layer, the preset lower limit of the sound speed, and the depth value of the grid points in this layer; estimate the lower limit of the average slowness at the depth of the grid points of this layer based on the initial slowness value corresponding to the depth of the grid points in the previous layer, the depth value of the grid points in the previous layer, the thickness value of the grid points in this layer, the preset upper limit of the sound speed, and the depth value of the grid points in this layer. Take the upper limit of the average slowness and the lower limit of the average slowness as constraint conditions to correct the average slowness value corresponding to the depth of the grid points in this layer.

6. The method according to claim 1, characterized in that, Convert the average sound speed values corresponding to the depths of the grid points in each layer into average slowness values to obtain the information of the variation of the average slowness with depth, specifically including: Convert the average sound speed values corresponding to the depths of the grid points in each layer into average slowness values. Denoise the average sound speed values corresponding to the depths of the grid points in each layer by using third-order difference regularization through a preset regularization matrix. Obtain the information of the variation of the average slowness with depth according to the denoising result.

7. The method according to claim 1, wherein Convert the average slowness into the expression of the local slowness, specifically including: For each depth of the grid points in the axial direction in sequence, express the average slowness value corresponding to the depth of the grid points in this layer as a weighted combination of the local slowness values corresponding to the depths of the grid points in the previous layer.

8. An apparatus for estimating the speed of sound, characterized in that, Include: A signal acquisition module that acquires the echo signal corresponding to each element in the linear array ultrasonic probe, and the echo signal is received by all elements. A beamforming module that performs beamforming on the echo signals corresponding to multiple elements for each candidate sound speed included in the preset sound speed sequence at each preset imaging grid point, and calculates the signal coherence at the same imaging grid point at different transmission times according to the beamforming result. An average sound speed determination module that screens out the average sound speed value corresponding to the depth of the grid points in this layer from each candidate sound speed according to the signal coherence for each depth of the grid points in the axial direction. An average slowness determination module that converts the average sound speed values corresponding to the depths of the grid points in each layer into average slowness values to obtain the information of the variation of the average slowness with depth. A local slowness representation module that converts the average slowness into the expression of the local slowness, and the local slowness is a weighted combination of several step functions. A solution module that determines the coefficient sequence of the weighted combination through the average slowness values at the depths of the grid points in each layer, and obtains the axial distribution information of the local sound speed according to the obtained axial distribution information of the local slowness.

9. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, and when the computer program is executed by a processor, it implements the method described in any one of claims 1 to 7 above.

10. An electronic device, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method described in any one of claims 1 to 7 above.

Images