A method, device, electronic device, and storage medium for ultrasonic imaging

Abstract

The present invention discloses an ultrasonic imaging method, apparatus, electronic device and storage medium. The method includes: converting the received ultrasonic echo signal representing time variation into a beam domain signal representing frequency variation; performing compression processing on the beam domain signal by using a preset sampling matrix to obtain a virtual measurement signal, wherein the preset sampling matrix satisfies the restricted isometry property, and the virtual measurement signal has sparsity; restoring the virtual measurement signal to the full array element state to obtain a virtual frequency domain signal; performing migration processing on the virtual frequency domain signal to obtain an actual frequency domain signal; and converting the actual frequency domain signal into a target image matrix for generating an ultrasonic image. The technical solution provided by the present invention can improve the image quality of ultrasonic imaging to a certain extent.

Description

Technical Field

[0001] The present invention relates to the technical field of ultrasonic imaging, and particularly relates to a method, device, electronic device and storage medium for ultrasonic imaging. Background Art

[0002] Ultrasonic imaging uses an ultrasonic beam to scan the human body, and obtains an image of internal organs by receiving and processing reflected signals. Plane wave imaging provides an ultra-fast ultrasonic imaging method for obtaining plane signals at a very high frame rate. Plane wave imaging simultaneously excites all elements of the transducer. After scattering by tissues, all elements also simultaneously receive echo signals, and an ultrasonic image of the entire area can be obtained through these echo signals. Beamforming is an important part of a plane wave imaging system, and the performance of the beamforming algorithm affects the quality of ultrasonic imaging.

[0003] In the prior art, the beamforming method is mainly the Delay and Sum (DAS) algorithm. The DAS algorithm applies fixed weights to the echo signals after delay processing to reduce signal side lobes. However, the weights of traditional window functions are usually a set of fixed parameters preset according to depth, and do not fully consider the characteristics of the echo signals themselves. Therefore, when the side lobes are attenuated to a certain extent, the main lobe width will also increase, resulting in poor quality of ultrasonic imaging. Summary of the Invention

[0004] In view of this, embodiments of the present invention provide a method, device, electronic device and storage medium for ultrasonic imaging to improve the image quality of ultrasonic imaging.

[0005] On the one hand, the present invention provides an ultrasonic imaging method, which includes: converting the received ultrasonic echo signal representing time change into a beam domain signal representing frequency change; compressing the beam domain signal by using a preset sampling matrix to obtain a virtual measurement signal, where the preset sampling matrix satisfies the restricted isometry property, and the virtual measurement signal has sparsity; restoring the virtual measurement signal to the full element state to obtain a virtual frequency domain signal; performing an offset processing on the virtual frequency domain signal to obtain an actual frequency domain signal; and converting the actual frequency domain signal into a target image matrix for generating an ultrasonic image.

[0006] In one embodiment, converting the received ultrasonic echo signal representing time change into a beam domain signal representing frequency change includes: performing a two-dimensional fast Fourier transform on the ultrasonic echo signal to obtain a beam domain signal.

[0007] In one embodiment, the transformation matrix of the direction vectors constructed from the number of ultrasonic transducer elements includes: equally dividing the range of -90° to 90° into intervals according to twice the number of ultrasonic transducer elements to obtain the element pitch; constructing the direction vectors of different emission directions of the ultrasonic transducer elements according to the element pitch; and generating a transformation matrix according to the direction vectors of different emission directions.

[0008] In one embodiment, solving the projection coefficient vector includes: constructing a cost function, where the cost function is used to represent the sparsity of the projection coefficient vector; constructing a Lagrangian operator according to the cost function, the measurement matrix, the projection coefficient vector, and the virtual measurement signal; and solving the projection coefficient vector by minimizing the Lagrangian operator.

[0009] In one embodiment, performing an offset process on the virtual frequency-domain signal to obtain an actual frequency-domain signal includes: calculating an actual sound speed and actual coordinates according to the angle between the plane wave emitted by the ultrasonic transducer and the ultrasonic transducer; and converting the virtual frequency-domain signal into an actual frequency-domain signal based on the actual sound speed and the actual coordinates.

[0010] In one embodiment, converting the actual frequency-domain signal into a target image matrix for generating an ultrasonic image includes: performing an inverse two-dimensional Fourier transform on the actual frequency-domain signal to obtain a target image matrix for generating an ultrasonic image.

[0011] On the other hand, the present invention also provides an ultrasonic imaging device, which includes: a signal conversion unit configured to convert a received ultrasonic echo signal representing a time change into a beam domain signal representing a frequency change; a signal compression unit configured to perform compression processing on the frequency-domain signal by using a preset sampling matrix to obtain a virtual measurement signal, where the preset sampling matrix satisfies the restricted isometry property, and the virtual measurement signal has sparsity; a signal restoration unit configured to restore the virtual measurement signal to a full-element state to obtain a virtual frequency-domain signal; a signal offset unit configured to perform an offset process on the virtual frequency-domain signal to obtain an actual frequency-domain signal; and a signal inverse conversion unit configured to convert the actual frequency-domain signal into a target image matrix for generating an ultrasonic image.

[0012] On the other hand, the present invention also provides an electronic device, which includes a processor and a memory. The memory is configured to store a computer program, and when the computer program is executed by the processor, the above ultrasonic imaging method is implemented.

[0013] On the other hand, the present invention also provides a computer-readable storage medium, which is configured to store a computer program, and when the computer program is executed by a processor, the above ultrasonic imaging method is implemented.

[0014] The technical solution provided by this application converts the echo signal received by the ultrasonic transducer into the form of the beam domain, then performs compressive sensing on the beam domain to obtain the virtual frequency domain signal, then performs offset processing on the virtual frequency domain signal to obtain the actual frequency domain signal, and finally converts the actual frequency domain signal into the target image matrix for ultrasonic imaging, thereby improving the image quality of ultrasonic imaging to a certain extent during the process of using compressive sensing for echo signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The features and advantages of the present invention will be more clearly understood by referring to the attached drawings. The drawings are schematic and should not be construed as imposing any limitation on the present invention. In the drawings:

[0016] Figure 1 shows a schematic flow diagram of the ultrasonic imaging method in an embodiment of the present invention;

[0017] Figure 2 shows a schematic diagram of the plane wave transmission and reception model and the seismic wave explosion model in an embodiment of the present invention;

[0018] Figure 3 shows a simulation comparison diagram of the imaging results of the frequency domain beam synthesis method based on compressive sensing and the DAS method in an embodiment of the present invention;

[0019] Figure 4 shows the low echo contrast diagram of the frequency domain beam synthesis method (fkCS) based on compressive sensing and the time domain beam synthesis method (tmCS) based on compressive sensing in an embodiment of the present invention;

[0020] Figure 5 shows the bright spot contrast diagram of the frequency domain beam synthesis method (fkCS) based on compressive sensing and the time domain beam synthesis method (tmCS) based on compressive sensing in an embodiment of the present invention;

[0021] Figure 6 shows the schematic diagram of the lateral resolution of the frequency domain beam synthesis method (fkCS) based on compressive sensing and the time domain beam synthesis method (tmCS) based on compressive sensing and its variation with the compression ratio in an embodiment of the present invention;

[0022] Figure 7 shows the simulated B-mode image of the frequency domain beam synthesis method (fkCS) based on compressive sensing and the time domain beam synthesis method (tmCS) based on compressive sensing when the compression ratio is 50% in an embodiment of the present invention;

[0023] Figure 8 shows a schematic diagram of the ultrasonic imaging device in an embodiment of the present invention;

[0024] Figure 9 The structural schematic diagram of the electronic device in an embodiment of the present invention is shown. Specific embodiments

[0025] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part rather than all of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] Plane wave imaging provides an ultrafast ultrasonic imaging method by acquiring plane signals at a very high frame rate, but this will sacrifice image quality and result in a problem of excessive data volume. Compared with the focused wave imaging mode, the plane wave transmission mode greatly improves the frame rate. However, since the transmitted beam is not focused, the signal-to-noise ratio, image resolution, and contrast of the echo signal all deteriorate to a certain extent. In practical applications, the multi-angle compound imaging method is often used to improve the image quality. In this imaging mode, the image quality improves as the angle increases, but the frame rate also linearly decreases. In order to retain the advantages of high frame rate of plane wave imaging as much as possible, different beamforming algorithms can be applied in plane wave imaging to improve the image quality without reducing the frame rate.

[0027] Beamforming is an important part of a medical ultrasonic imaging system, and the performance of the beamforming algorithm affects the quality of ultrasonic imaging. The commonly used beamforming methods currently are the delay and sum (DAS) algorithm and the adaptive beamforming algorithm. The DAS algorithm is the most basic beamforming algorithm. It applies fixed weights to the echo signals after delay processing to reduce signal side lobes. However, the weights of traditional window functions are usually a set of fixed parameters preset according to depth and do not fully consider the characteristics of the echo signals themselves. Therefore, when the side lobes are attenuated to a certain extent, the main lobe width will also increase, that is, the resolution deteriorates. Compared with DAS, the adaptive beamforming algorithm uses the echo signals received by the transducer array to adaptively calculate the dynamic weighting values applied to each element according to environmental changes, so it has better resolution and anti-interference capabilities. Traditional adaptive beamforming has advantages such as high resolution and strong anti-interference capabilities, but at the same time, its problems of high side lobe level and high sensitivity to angular mismatch will greatly reduce the receiving performance. In addition, adaptive beam synthesis involves a large number of matrix operations, which will occupy a large amount of computing space and reduce the computing speed.

[0028] In addition to image quality, high frame rates also bring problems of data loading and memory storage. Compared with MV-based algorithms, the plane wave imaging process introduces the Fourier domain, providing a new direction for beamforming. Compared with beamforming methods in the time domain, beamforming in the Fourier domain may improve computational efficiency by applying the fast Fourier transform without occupying too much computing space. Subsequently, a Fourier domain beamforming process based on the explosion reflection model (ERM) was proposed. This model assumes that the echoes of backscatterers only propagate along the sensor direction. The principle of Stolt's shift in this model is closer to reality, thus improving the imaging quality.

[0029] Compressed sensing (CS) is a new sampling theory. By exploiting the sparse characteristics of signals, it enables signals to be sampled randomly under conditions far less than the Nyquist sampling rate to obtain discrete samples of the signals, and then the signals can be perfectly reconstructed through non-linear reconstruction algorithms. The compressed sensing theory states that as long as a signal is compressible or sparse in a certain transform domain, a high-dimensional signal obtained by transformation can be projected onto a low-dimensional space using an observation matrix that is uncorrelated with the transform basis. Then, by solving an optimization problem, the original signal can be reconstructed with high probability from these few projections, and it can be proven that such projections contain sufficient information for reconstructing the signal. The characteristic of the compressed sensing theory is that it can reduce the number of sampling channels or the signal dimension while maintaining the original characteristics of the signal. However, in traditional methods of applying compressed sensing to time-domain beam synthesis, as the compression rate decreases, the image quality deteriorates rapidly, and there is an obvious contradictory relationship between image quality and data compression rate.

[0030] The embodiments of this specification provide a scenario example of an ultrasonic imaging method. First, perform a two-dimensional fast Fourier transform (FFT) on the received ultrasonic echo signal to transform the echo signal X(x, z = 0, t) received by the transducer into the beam domain G(k x , z = 0, f). Then, determine the size of the transformation matrix according to the number of array elements, and construct the transformation matrix H. To ensure that the projection vector is a sparse vector, the number of rows of the transformation matrix should be less than the number of columns. After dividing the area from -90° to 90° into equal parts, we get N is the number of ultrasonic transducer array elements. Use to construct the direction vector where d is the ultrasonic transducer array element spacing and λ is the ultrasonic wavelength. Construct the transformation matrix Η from the direction vector δ(φ k ).

[0031]

[0032] Projecting the array element received signal onto the transformation matrix yields the projection coefficient vector, and the array element received signal can be expressed as the product of the transformation matrix and the projection coefficient vector at this time.

[0033] G(k x ,0,f) = HS(k x ,0,f)

[0034] G(k x ,0,f) represents the frequency-domain signal after performing a two-dimensional Fourier transform on the transducer received signal, H represents the transformation matrix, and S(k x ,0,f) represents the projection coefficient vector of the transducer received signal G(k x ,0,f) onto the transformation matrix H. Most of the values in the projection coefficient vector are 0 or extremely small and can be approximated as 0, and this vector is the sparse vector.

[0035] Design the sampling matrix Λ = [κ1, κ2, …, κ M , the size of Λ is M×N, where M being less than N can play a role in compressing the frequency domain of the original signal. In this algorithm, the sampling matrix is a random Gaussian matrix. The sampling matrix can downsample the original sampling data to obtain a new sampling matrix G s (k x ,0,f).

[0036] Z(k x ,0,f) = ΛG(k x ,0,f) = ΛHS(k x ,0,f)

[0037] The matrix Z(k x ,0,f) is the downsampled matrix that needs to be recovered by this algorithm, and the dimension of Z(k x ,0,f) is M×L. When the signal is sparse and the sampling matrix has the Restricted Isometry Property (RIP), the original signal can be recovered from Z(k x ,0,f). Since the dimension of Z(k x ,0,f) is much smaller than that of G(k x ,0,f), the compressed sensing theory can play an important role in reducing the sampling frequency, reducing the sampling channels, etc.

[0038] After obtaining the transformation matrix and the sampling matrix, calculate the measurement matrix P from the sampling matrix.

[0039] P = ΛH

[0040] The measurement matrix is used to sample the observed values so that the received signal can be reconstructed. To reconstruct the original signal from the compressed signal, the necessary condition needs to be met: the product of the measurement matrix and the sparse matrix satisfies the RIP. This property ensures a one-to-one mapping relationship from the original space to the sparse space, which requires that the sampling matrix randomly selected from the observation matrix must be non-singular. The measurement matrix can measure the signal to obtain the measurement vector, and then use the reconstruction algorithm to reconstruct the original signal from the measured values. Compressive sampling is performed on the received signal of the array element, and the compressive sampling vector can be obtained through the sampling matrix. When designing the measurement matrix, it is required that during the sparse representation of the signal, the measured values do not affect the information of the original signal, so as to ensure that the signal can be accurately reconstructed. At this time, the downsampling matrix Z(k x ,0,f) can be expressed as:

[0041] Z(k x ,0,f) = PS(k x ,0,f)

[0042] After obtaining the compressive sampling values, the Regularized Multiplicative FOCUSS (RM-FOCUSS) algorithm is used to estimate the projection coefficient vector. To solve for the sparse projection coefficient vector, a cost function J (p) (S) is first constructed. The smaller the cost function, the sparser the signal. Suppose there are L plane wave signals,

[0043]

[0044] When p is closer to 0, the S(t) represented by J (p) (S) is sparser.

[0045] Subsequently, the Lagrangian operator L(S,Ω) is defined,

[0046]

[0047] where ω l is the Lagrangian operator vector, l = 1, 2, … L, and S(t) can be obtained by minimizing L(S,Ω). The specific calculation process is as follows:

[0048]

[0049] After calculating the sparse solution S(k x ,0,f), the restored full array element signal can be obtained through the formula G s (k x ,0,f) = HS(k x ,0,f). At this time, G s (k x ,0,f) is a full array element signal, but after the compression process, it is different from the original frequency domain signal G(kx , 0, f) is different.

[0050] According to the seismic wave model, the process of ultrasonic imaging can be transformed into a "virtual target" in the imaging area emitting ultrasonic waves to the ultrasonic transducer by itself. After the transducer receives the signal of the virtual target, imaging is performed. During this transformation process, the sound speed c in the original process and the coordinates of the imaging area have both changed. Therefore, it is necessary to find the corresponding relationship between the real imaging target and the "virtual target" and solve for the coordinates of the "virtual target" and the sound speed of the reflected signal.

[0051]

[0052] Among them, θ is the angle between the emitted plane wave and the ultrasonic transducer.

[0053] G s (k x , 0, f) is the result after compressive sensing recovery. Assume ψ θ (x, z, t = 0) is the final imaging signal, φ θ (k x , k z , t = 0) is the frequency-domain signal corresponding to ψ θ (x, z, t = 0), that is, the frequency-domain signal of the image formed by the real imaging target. Then it is the frequency-domain signal of the imaging signal of the "virtual target". Now it is necessary to convert G s (k x , 0, f) into ψ θ (x, z, t = 0) for imaging.

[0054]

[0055] Among them It is known and then φ θ (k x , k z , t = 0) is deduced.

[0056]

[0057] Finally, through two-dimensional inverse Fourier transform the final image ψ θ (x, z, t = 0) is obtained.

[0058] The above is only an example of a scenario provided in this specification and is not intended to limit the present invention. Any modifications, equivalent replacements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

[0059] Please refer toFigure 1 , an embodiment of the present application provides an ultrasonic imaging method, which may include the following steps.

[0060] S110: Convert the received ultrasonic echo signal representing time change into a beam domain signal representing frequency change.

[0061] In some cases, if compressive sensing is directly applied to the time-domain beamforming method, as the compression ratio decreases, the image quality deteriorates rapidly, and there is an obvious contradictory relationship between the image quality and the data compression ratio. Therefore, the received ultrasonic echo signal can be first converted into a beam domain signal representing frequency change.

[0062] In this embodiment, a two-dimensional fast Fourier transform is performed on the received ultrasonic echo signal to transform the echo signal X(x, z = 0, t) received by the ultrasonic transducer into the beam domain G(k x , z = 0, f).

[0063] S120: Compress the beam domain signal by using a preset sampling matrix to obtain a virtual measurement signal; wherein, the preset sampling matrix satisfies the restricted isometry property; the virtual measurement signal has sparsity.

[0064] In this embodiment, the preset sampling matrix consists of M rows, that is, M sampling bases. The preset sampling matrix Λ = [γ1, γ2,..., γ M T . Each sampling base is an N-dimensional vector, that is, γ i = [γ i1 , γ i2 ,..., γ iN (i = 1, 2,..., M). The sampling of the i-th row represents projecting the outputs of all array elements onto this sampling base, corresponding to a compressive sampling point. For example, Z i (t) = γ i1 G1(t) + γ i2 G2(t) +... + γ iN G N (t) represents the received signal of the i-th compressive sampling point corresponding to the compressive sampling. The sampling matrix Λ has a total of M rows, indicating that only M compressive sampling points are required, that is, M array elements are selected from the N array elements of the original array for spatial sampling. The sampling matrix can be a Gaussian random matrix. Of course, the sampling matrix can also be a Hadamard matrix, a sparse random matrix, a partial Fourier matrix, etc. The original sampling matrix data can be downsampled by using the sampling matrix to obtain the virtual measurement signal Z(k x , 0, f).

[0065] Z(k x , 0, f) = ΛG(k​x ,0,f)

[0066] S130: Restore the virtual measurement signal to the full array element state to obtain a virtual frequency domain signal.

[0067] In some cases, through the compressive sensing theory, after reducing the number of sampling channels or reducing the dimension of the signal while maintaining the original characteristics of the signal, it is still necessary to restore the compressed data to obtain a restored signal.

[0068] In this embodiment, determine the size of the transformation matrix according to the number of array elements, construct the transformation matrix H. To ensure that the projection vector is a sparse vector, the number of rows of the transformation matrix should be less than the number of columns. Divide the area from -90° to 90° into equal parts to obtain N is the number of ultrasonic transducer array elements. Use Construct a direction vector where d is the ultrasonic transducer array element spacing, λ is the ultrasonic wavelength, and construct the transformation matrix Η from the direction vector δ(φ k ).

[0069]

[0070] Project the array element received signal onto the transformation matrix to obtain a projection coefficient vector. At this time, the array element received signal can be expressed as the product of the transformation matrix and the projection coefficient vector. By constructing the measurement matrix P = ΛH. Where, Z(k x ,0,f) = PS(k x ,0,f). By solving the projection sparse vector S(k x ,0,f), and then using G s (k x ,0,f) = HS(k x ,0,f) to obtain a virtual measurement signal.

[0071] S140: Perform an offset process on the virtual frequency domain signal to obtain an actual frequency domain signal.

[0072] In some cases, the process of ultrasonic plane wave imaging can be summarized as: the ultrasonic transducer emits ultrasonic waves to the imaging area, the imaging target reflects the ultrasonic signal, the transducer receives the reflected echo signal, and processes the received signal for imaging. According to the seismic wave model, the process of ultrasonic imaging can be transformed into a "virtual target" in the imaging area emitting ultrasonic waves to the ultrasonic transducer by itself, and the transducer performs imaging after receiving the signal of the virtual target.

[0073] Please refer to Figure 2(a), The ultrasonic transducer emits ultrasonic waves to cover the imaging area. The imaging target reflects the ultrasonic signals, and the reflected echo signals return to the ultrasonic transducer. The signal received by the ultrasonic transducer at the ordinate z = 0 is denoted as X(x, z = 0, t). Please refer to Figure 2 (b) As shown, in the seismic wave explosion model, it is first assumed that the imaging target itself emits a signal, which is the emission signal at t = 0, that is, ψ θ (x, z, t = 0). The true ultrasonic imaging result can also be regarded as the imaging target itself emitting an ultrasonic signal to the transducer for imaging. However, when applying the explosion model to plane wave imaging, Stolt's migration is required for coordinate transformation.

[0074] In this embodiment, the sound speed c in the original process and the coordinates of the imaging area have both changed. Therefore, it is necessary to find the correspondence between the true imaging target and the "virtual target", and solve the coordinates of the "virtual target" and the sound speed of the reflected signal.

[0075]

[0076] Among them is the "virtual sound speed" of the ultrasonic signal emitted by the "virtual target", (x s , z s ) are the true coordinates of the imaging target, is the coordinate of the "virtual target". α, β, and γ are transformation coefficients, defined as follows:

[0077]

[0078] In the formula, θ is the angle between the emitted plane wave and the ultrasonic transducer.

[0079] S150: Convert the actual frequency domain signal into a target image matrix for generating an ultrasonic image.

[0080] In this embodiment, G s (k x , 0, f) is the result after compressive sensing recovery. Assume ψ θ (x, z, t = 0) is the final imaging signal, φ θ (k x , k z , t = 0) is the frequency domain signal corresponding to ψ θ (x, z, t = 0), that is, the frequency domain signal of the image formed by the true imaging target. Then it is the frequency domain signal of the imaging signal of the "virtual target". Now it is necessary to convert G s (k x , 0, f) into ψ θPerform imaging at (x, z, t = 0).

[0081]

[0082] Among them It is known Then φ is derived θ (k x , k z , t = 0).

[0083]

[0084] Finally, through two-dimensional inverse Fourier transform The final image ψ is obtained θ (x, z, t = 0).

[0085] Please refer to Figure 3 , Compare the imaging result with a compression ratio of 80% (abbreviation: fkCS_80%) and Delay-and-Sum (DAS) with the frequency-domain beamforming algorithm based on Stolt's f-k migration of the uncompressed data volume (abbreviation: f-k). The calculation method of the compression ratio (abbreviation: R) is as follows:

[0086]

[0087] Among them, N c and N o are the data volumes after and before compression respectively. In the simulation results, the gCNR is calculated using regions A, B, and C respectively to represent the contrast of the low echo region and bright spots in the imaging. The calculation method of gCNR is as follows:

[0088]

[0089] Among them, p t is the probability density distribution of the target region, and p b (x) is the probability density distribution of the background region.

[0090] Calculate the transverse resolution of the scatterer point with coordinates (10, 70) mm in the calculation region D to compare the resolution of different imaging methods. The transverse resolutions of the three methods at point D for the frequency-domain beamforming method based on compressive sensing are 0.5985 mm (DAS), 0.5262 mm (f-k), and 0.5451 mm (fkCS_80%). From the results, it can be seen that the frequency-domain beamforming imaging method improves the transverse resolution by 12% compared to DAS. The imaging result with a compression ratio of 80% only reduces the transverse resolution by 3.6% compared to when it is not compressed. The contrast ratios of the three methods in the low echo region are 1.7482 (DAS), 2.5791 (f-k), and 2.4915 (fkCS_80%). The f-k method improves the low echo contrast by 47.5% compared to DAS. The fkCS_80% method with a compression ratio of 80% reduces the contrast by 3.4% compared to when it is not compressed (f-k). The appearances of the three methods in the bright spot contrast are 2.9104 (DAS), 3.1455 (f-k), and 3.1081 (fkCS_80%). The f-k method improves the low echo contrast by 8% compared to DAS. The fkCS_80% method with a compression ratio of 80% reduces the contrast by 1.2% compared to when it is not compressed (f-k). It can be seen from this that the frequency-domain beamforming method based on compressive sensing has a certain improvement in image quality compared to DAS. And when the imaging channel number is compressed to 80%, the impact on the imaging quality is small. This is because the ultrasonic signal itself has sparsity, and even if some channels of the signal channels are compressed, the original data can still be restored to a certain extent.

[0091] Please refer to Figures 4 to 7 , the simulation results and the analysis results of the image quality parameters of the frequency-domain beamforming method based on compressive sensing (fkCS) and the time-domain beamforming method based on compressive sensing (tmCS). Figure 4 、 Figure 5 and Figure 6 respectively show the trend graphs of the low echo contrast, bright spot contrast, and transverse resolution of the two imaging methods changing with the compression ratio. Figure 7 is the simulated B-mode image of fkCS and tmCS when the compression ratio is 50%. It can be seen from the B-mode image that when the compression ratio is 50%, the image quality of tmCS_50% is very poor, the low echo region is almost covered, and obvious artifacts are generated in the bright spots; while in the imaging result of fkCS_50%, the low echo region and the scatterer points can still be distinguished, and there are slight artifacts in the bright spots, and the degree of artifacts is significantly weaker than that of fkCS. From Figures 4 - 6 it can be seen that as the compression ratio decreases, the contrast and transverse resolution of tmCS decrease faster than those of fkCS. It can be concluded that when the compression ratio is the same, using compressive sensing in the frequency domain results in better image quality than using compressive sensing in the time domain.

[0092] In some cases, it is advisable to select M between 0.3N and 0.8N, that is, the compression ratio is between 30% and 80%, and M should be an integer. Within this range, both the number of data channels can be reduced and good imaging quality can be maintained.

[0093] In one embodiment, converting the received ultrasonic echo signal representing time variation into a beam domain signal representing frequency variation may include: performing a two-dimensional fast Fourier transform on the ultrasonic echo signal to obtain a beam domain signal.

[0094] In this embodiment, the fast Fourier transform is a general term for efficient and fast calculation methods using the computer discrete Fourier transform. It is obtained by improving the algorithm of the discrete Fourier transform according to the characteristics of odd, even, imaginary, real, etc. of the discrete Fourier transform.

[0095] In one embodiment, restoring the virtual measurement signal to the full array element state to obtain a virtual frequency domain signal may include: constructing a measurement matrix; wherein, the virtual measurement signal is expressed as the product of the measurement matrix and a projection coefficient vector; the projection coefficient vector is used to represent the projection of the ultrasonic echo signal on a transformation matrix of a direction vector constructed by the number of ultrasonic transducer array elements; solving the projection coefficient vector; performing a product operation on the transformation matrix and the projection coefficient vector to obtain the virtual frequency domain signal.

[0096] In this embodiment, the measurement matrix is an M×2N matrix. When the measurement matrix satisfies the restricted isometry property, the beam domain signal G(k x ,0,f) at full array elements can be accurately reconstructed from the virtual measurement signal Z(k x ,0,f) by solving the projection coefficient vector S(k x ,0,f). The measurement matrix can be expressed as the product of a preset sampling matrix and a transformation matrix.

[0097] In one embodiment, the transformation matrix of the direction vector constructed by the number of ultrasonic transducer array elements may include: equally dividing the -90° to 90° region range by twice the number of ultrasonic transducer array elements to obtain an element spacing; constructing a direction vector of different emission directions of the ultrasonic transducer array elements according to the element spacing; generating a transformation matrix according to the direction vectors of different emission directions.

[0098] In this embodiment, considering a full array ultrasonic beam with N elements arranged uniformly and the element spacing being (λ is the operating wavelength of the ultrasonic transducer). Now K far-field echo signals are received, and their complex amplitudes and incident angles are S k (t) and θ sk(k = 1, 2, …, K). One of them is the desired signal, and the remaining K - 1 are interference signals. The received signals of each element of the ultrasonic transducer are represented by an N - dimensional vector X(t), X(t) = [x1(t), x2(t), …, x N (t)] T . Then there is

[0099]

[0100] where a(θ sk ) is the directivity appropriate amount of the array in the direction of θ sk (k = 1, 2, …, K).

[0101]

[0102] According to the principle of dividing by equal sin(θ sk ), the airspace from - 90° to 90° is equally divided into 2N parts, obtaining θ1, θ2, …, θ 2N , and the transformation matrix H is constructed with these 2N directivity appropriate amounts.

[0103] H = [a(θ1), a(θ2), …, a(θ 2N )]

[0104] Writing the array received signal vector X(t) in matrix form represented by the transformation matrix H, there is

[0105] X(t) = HS(t)

[0106] where S(t) is the projection coefficient vector of the array element received signal vector X(t) on the transformation matrix H. In some cases, θ sk (k = 1, 2, …, K) are all among θ1, θ2, …, θ 2N , then the projection coefficient vector S(t) has a form similar to S(t) = [0, 0, …, s1(t), 0, …, s K (t), 0, …, 0]. Then only a few elements in the vector S(t) are non - zero, and the rest are zero elements, that is, S(t) is sparse. Therefore, according to the theory of compressive sensing, the array element received signal vector X(t) can be accurately recovered by a reconstruction algorithm after compressive sampling.

[0107] In one embodiment, solving the projection coefficient vector may include: constructing a cost function; where the cost function is used to represent the sparsity degree of the projection coefficient vector; constructing a Lagrangian operator according to the cost function, the measurement matrix, the projection coefficient vector, and the virtual measurement signal; and solving the projection coefficient vector by minimizing the Lagrangian operator.

[0108] In this embodiment, after obtaining the compressive sampling values Z of M compressive sampling elements M×L , the projection coefficient vector S is estimated by using the multi-point regularized underdetermined system focusing solution (RM-FOCUSS) algorithm 2N×L , and then the full-array element received signal vector G is reconstructed according to G = HS S .

[0109] To solve for S 2N×L , the following cost function is constructed:

[0110]

[0111] When p approaches 0, J (p) (S) represents the non-zero columns of S 2N×L , and the smaller J (p) (S) is, the sparser the signal is. Then the Lagrangian operator L(S, Ω) is defined as

[0112]

[0113] where ω l is the Lagrangian operator vector, l = 1, 2, … L, and S(t) can be obtained by minimizing L(S, Ω), and thus the sparse solution S 2N×L can be obtained. It is solved by the regularized M-FOCUSS iterative algorithm

[0114]

[0115] After calculating the sparse solution S(k x , 0, f), through the formula G s (k x , 0, f) = HS(k x , 0, f), the restored full-array element virtual frequency domain signal G s (k x , 0, f) can be obtained

[0116] In one embodiment, performing an offset process on the virtual frequency domain signal to obtain the actual frequency domain signal may include: calculating the actual sound speed and actual coordinates according to the angle between the plane wave emitted by the ultrasonic transducer and the ultrasonic transducer; converting the virtual frequency domain signal into the actual frequency domain signal based on the actual sound speed and the actual coordinates

[0117] In this embodiment, G s (k x , 0, f) is the result after compressive sensing recovery. Assume ψ θ (x, z, t = 0) is the final imaging signal, and φ θ (k x , k z, t = 0) is ψ θ (x, z, t = 0) corresponding frequency domain signal, that is, the frequency domain signal of the image formed by the real imaging target. Then it is the frequency domain signal of the imaging signal of the "virtual target". Now it is necessary to convert G s (k x , 0, f) into ψ θ (x, z, t = 0) for imaging.

[0118]

[0119] Among them It is known and then derive φ θ (k x , k z , t = 0).

[0120]

[0121] Finally, through two-dimensional inverse Fourier transform obtain the final image ψ θ (x, z, t = 0).

[0122] In one embodiment, converting the actual frequency domain signal into a target image matrix for generating an ultrasonic image may include: performing two-dimensional inverse Fourier transform on the actual frequency domain signal to obtain a target image matrix for generating an ultrasonic image.

[0123] In this embodiment, through two-dimensional inverse Fourier transform obtain the final image ψ θ (x, z, t = 0).

[0124] Please refer to Figure 8 , one embodiment of the present application further provides an ultrasonic imaging device, and the ultrasonic imaging device may include: a signal conversion unit, a signal compression unit, a signal recovery unit, a signal offset unit, and a signal inverse conversion unit.

[0125] The signal conversion unit is used to convert the received ultrasonic echo signal representing time change into a beam domain signal representing frequency change.

[0126] The signal compression unit is used to perform compression processing on the beam domain signal by using a preset sampling matrix to obtain a virtual measurement signal; wherein, the preset sampling matrix satisfies the restricted isometry property; the virtual measurement signal has sparsity.

[0127] The signal recovery unit is used to restore the virtual measurement signal to the full array element state to obtain a virtual frequency domain signal.

[0128] A signal offset unit for offsetting the virtual frequency domain signal to obtain an actual frequency domain signal.

[0129] A signal inverse conversion unit for converting the actual frequency domain signal into a target image matrix for generating an ultrasonic image.

[0130] Regarding the specific functions and effects achieved by the ultrasonic imaging device, reference may be made to other embodiments of this specification for comparison and explanation, which will not be elaborated here. Each module in the ultrasonic imaging device can be implemented in whole or in part by software, hardware, and their combination. Each module can be embedded in the processor of the computer device in hardware form or be independent of it, or be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to each of the above modules.

[0131] Please refer to Figure 9 , an embodiment of the present application further provides an electronic device, which includes a processor and a memory. The memory is used to store a computer program. When the computer program is executed by the processor, the above ultrasonic imaging method is implemented.

[0132] Among them, the processor can be a central processing unit (CPU). The processor can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. chips, or a combination of the above various types of chips.

[0133] As a non-transitory computer-readable storage medium, the memory can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the method in the embodiment of the present invention. By running the non-transitory software programs, instructions, and modules stored in the memory, the processor can execute various functional applications and data processing of the processor, that is, implement the method in the above method embodiments.

[0134] The memory may include a program storage area and a data storage area. The program storage area may store an operating system and application programs required for at least one function. The data storage area may store data created by the processor and the like. In addition, the memory may include a high-speed random access memory, and may also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some embodiments, the memory may optionally include a memory remotely disposed relative to the processor, and these remote memories may be connected to the processor through a network. Examples of the above network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0135] An embodiment of the present application further provides a computer-readable storage medium for storing a computer program, which, when executed by a processor, implements the above ultrasonic imaging method.

[0136] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the described embodiments can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it may include the processes of the embodiments of the various methods. Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided in this specification may include at least one of non-volatile and volatile memories. Non-volatile memory may include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical memory, etc. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0137] It should be understood that each process and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the processes and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. 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 device for implementing the specified function in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0138] Among the multiple embodiments described in this specification, a progressive approach is adopted for description. Different embodiments focus on describing the parts that are different from other embodiments. After reading this specification, those skilled in the art can learn about the multiple embodiments in this specification and the multiple technical features disclosed by the embodiments, and can make more combinations. For the sake of concise description, not all possible combinations of the various technical features in the embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered to be within the scope described in this specification.

[0139] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or further includes elements inherent in such a process, method, commodity or device. Without further limitation, an element defined by the statement "including one..." does not exclude the existence of another identical element in the process, method, commodity or device including the said element.

[0140] Each of the multiple embodiments in this specification itself focuses on emphasizing the parts that are different from other embodiments, and the embodiments can be explained by comparing with each other. Any combination of the multiple embodiments in this specification by those skilled in the art based on general technical knowledge is covered within the scope disclosed in this specification.

[0141] The above are only the embodiments of this case and are not intended to limit the scope of protection of the claims of this case. For those skilled in the art, various changes and modifications can be made to this case. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of this case shall be included within the scope of the claims of this case.

Claims

1. An ultrasonic imaging method, characterized in that, The method includes: Converting the received ultrasonic echo signal representing time variation into a beam domain signal representing frequency variation; Performing compression processing on the beam domain signal by using a preset sampling matrix to obtain a virtual measurement signal; wherein, the preset sampling matrix satisfies the restricted isometry property; the virtual measurement signal has sparsity; Restoring the virtual measurement signal to the full array element state to obtain a virtual frequency domain signal; Performing an offset processing on the virtual frequency domain signal to obtain an actual frequency domain signal; Converting the actual frequency domain signal into a target image matrix for generating an ultrasonic image; Wherein, the restoring the virtual measurement signal to the full array element state to obtain a virtual frequency domain signal includes: Constructing a measurement matrix; wherein, the virtual measurement signal is expressed as the product of the measurement matrix and a projection coefficient vector; the projection coefficient vector is used to represent the projection of the ultrasonic echo signal on a transformation matrix of direction vectors constructed by the number of ultrasonic transducer array elements; Solving the projection coefficient vector; Performing a product operation on the transformation matrix and the projection coefficient vector to obtain the virtual frequency domain signal.

2. The method according to claim 1, wherein Converting the received ultrasonic echo signal representing time variation into a beam domain signal representing frequency variation includes: Performing a two-dimensional fast Fourier transform on the ultrasonic echo signal to obtain a beam domain signal.

3. The method according to claim 1, wherein The transformation matrix of direction vectors constructed by the number of ultrasonic transducer array elements includes: Dividing the range of -90° to 90° into equal parts according to twice the number of ultrasonic transducer array elements to obtain an element spacing; Constructing direction vectors of different emission directions of the ultrasonic transducer array elements according to the element spacing; Generating a transformation matrix according to the direction vectors of different emission directions.

4. The method according to claim 1, characterized in that Solving the projection coefficient vector includes: Constructing a cost function; wherein, the cost function is used to represent the sparsity degree of the projection coefficient vector; Constructing a Lagrangian operator according to the cost function, the measurement matrix, the projection coefficient vector and the virtual measurement signal; Solving the projection coefficient vector by performing a minimization operation on the Lagrangian operator.

5. According to the method described in claim 1, performing an offset processing on the virtual frequency domain signal to obtain an actual frequency domain signal includes: Calculating an actual sound speed and actual coordinates according to the angle between the plane wave emitted by the ultrasonic transducer and the ultrasonic transducer; Converting the virtual frequency domain signal into an actual frequency domain signal based on the actual sound speed and the actual coordinates.

6. The method according to claim 1, wherein Converting the actual frequency domain signal into a target image matrix for generating an ultrasonic image includes: Performing an inverse two-dimensional Fourier transform on the actual frequency domain signal to obtain a target image matrix for generating an ultrasonic image.

7. An ultrasonic imaging device, characterized in that, The ultrasonic imaging device includes: A signal conversion unit, configured to convert the received ultrasonic echo signal representing time variation into a beam domain signal representing frequency variation; A signal compression unit, configured to perform compression processing on the beam domain signal by using a preset sampling matrix to obtain a virtual measurement signal; wherein, the preset sampling matrix satisfies the restricted isometry property; the virtual measurement signal has sparsity; A signal restoration unit, configured to restore the virtual measurement signal to the full array element state to obtain a virtual frequency domain signal; A signal offset unit for offsetting the virtual frequency-domain signal to obtain an actual frequency-domain signal; A signal inverse conversion unit for converting the actual frequency-domain signal into a target image matrix for generating an ultrasonic image; Wherein, the signal recovery unit is specifically configured to construct a measurement matrix; wherein the virtual measurement signal is expressed as the product of the measurement matrix and a projection coefficient vector; the projection coefficient vector is used to represent the projection of the ultrasonic echo signal on a transformation matrix of a direction vector constructed by the number of ultrasonic transducer elements; Solve the projection coefficient vector; Perform a product operation on the transformation matrix and the projection coefficient vector to obtain the virtual frequency-domain signal.

8. An electronic device, characterized in that, The electronic device includes a processor and a memory, and the memory is used to store a computer program. When the computer program is executed by the processor, the method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium is used to store a computer program. When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images