A Low-Frequency Reverberant Ultrafast Viscoelastic Imaging Method and System for Breast Ultrasound Based on FPINN

Abstract

This invention belongs to the field of medical ultrasound imaging technology, and specifically relates to a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging method and system based on FPINN. The low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging method includes the following steps: acquiring and processing displacement data of the breast tissue to be imaged to obtain frequency domain information; acquiring shear wave velocities at different frequencies based on the frequency domain information, and obtaining elasticity, viscosity, and fluidity data of the breast tissue based on the shear wave velocities at different frequencies; and performing imaging processing using a trained FPINN model based on the obtained elasticity, viscosity, and fluidity data of the breast tissue to obtain the imaging results of the breast tissue. The FPINN model of this invention can quickly convert the input elasticity, viscosity, and fluidity data of the breast tissue into imaging results during the imaging processing, thereby improving the real-time performance and accuracy of the imaging.

Description

Technical Field

[0001] This invention belongs to the field of medical ultrasound imaging technology, and specifically relates to a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging method and system based on FPINN (Fractional Physics-Informed Neural Network). Background Technology

[0002] Breast cancer is a disease caused by the abnormal proliferation and uncontrolled formation of tumors from breast cells. It is one of the three most common cancers worldwide. Further, breast, lung, and colorectal cancers account for 51% of all newly diagnosed cancers in women, with breast cancer accounting for 32%. Despite substantial evidence that early-stage breast cancer is curable in many cases, its incidence and mortality rates continue to rise, most likely due to insufficient screening for early breast tissue lesions using current diagnostic techniques.

[0003] Currently, commonly used clinical methods for breast lesion screening include mammography, breast ultrasound, magnetic resonance imaging (MRI), breast biopsy, thermal imaging, and molecular imaging. Mammography is widely used in large-scale screening due to its low cost; however, it has low sensitivity in dense breast tissue, potentially missing small tumors, and carries the risk of low-dose radiation. More importantly, mammography only provides two-dimensional images, which may lead to missed lesions due to overlapping lesions. Breast ultrasound and thermal imaging are radiation-free and suitable for pregnant women, breastfeeding women, and young women, especially for assessing masses in dense breast tissue. However, their sensitivity and specificity are low, making it difficult to accurately locate or detect deep lesions, and they also carry the risk of missed diagnoses. While MRI and molecular imaging offer high resolution and can clearly display breast tissue structure, their high cost limits their application in large-scale screening. In conclusion, there is an urgent clinical need for a convenient, simple, low-cost, non-invasive, and radiation-free breast lesion screening tool, which is of significant clinical importance for the early detection and treatment of breast cancer.

[0004] In recent years, shear wave elastography (SWE) has developed rapidly. As a non-destructive method for measuring the biomechanical properties of tissues, it has been widely applied in various fields such as liver, breast, and heart. However, existing ultrasound shear wave imaging techniques based on acoustic radiation force or external vibration excitation via a steady-state exciter have certain safety risks. Specifically, acoustic radiation force elastography requires focused ultrasound to act directly on biological tissue, and its safety for diseased tissues is not yet clear. The external vibration excitation method using a steady-state exciter may induce transient large accelerations, thereby generating shock waves that may damage the microenvironment of diseased tissues. In addition, multiple reflections from organ boundaries, tissue inhomogeneities, and mode transitions can generate complex wave fields in time and space, leading to signal interference and superposition, affecting the accuracy of shear wave propagation paths, increasing the difficulty of tissue characteristic assessment, and thus causing uncertainty and error in the results.

[0005] The mechanical properties of soft tissues are complex and cannot be fully described by simple integer-order models. Without fractional-order theory, existing technologies may not accurately characterize the elasticity, viscosity, and other mechanical behaviors of soft tissues such as breast tissue, which in turn affects the imaging effect. This means that existing technologies may not be able to process these data quickly and effectively to generate imaging results, thus affecting the real-time performance and accuracy of imaging. Summary of the Invention

[0006] The purpose of this invention is to provide a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging method and system based on FPINN, to solve one or more of the aforementioned technical problems. In the technical solution provided by this invention, the FPINN model can quickly convert input breast tissue elasticity, viscosity, and fluidity data into imaging results during the imaging process. Compared with existing traditional imaging methods, it does not require complex manual feature extraction and lengthy computational processes. This invention utilizes the parallel computing power of neural networks and the learned mapping relationships to efficiently generate imaging results, improving the real-time performance and accuracy of imaging.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] In a first aspect, this invention provides a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging method based on FPINN, comprising the following steps:

[0009] Displacement data of the breast tissue to be imaged is acquired and processed to obtain frequency domain information; based on the frequency domain information, shear wave velocities at different frequencies are obtained, and elasticity, viscosity, and fluidity data of the breast tissue are obtained according to the shear wave velocities at different frequencies.

[0010] Based on the obtained data on the elasticity, viscosity and fluidity of breast tissue, the trained FPINN model was used for imaging processing to obtain the imaging results of breast tissue.

[0011] In the training process of the FPINN model, the input is sample data of breast tissue elasticity, viscosity and fluidity, and the output is the imaging results of breast tissue obtained according to the viscoelastic properties of different locations inside the breast tissue. The parameters are updated by minimizing the defined loss function, and the training is completed after the preset convergence condition is reached.

[0012] Defined loss function L fPINN The expression is:

[0013] L fPINN =L fPDE +β1L data1 +β2L dat ;

[0014] In the formula, L fPDE L is the model-driven loss. data1 L data2 These represent the data-driven losses from the first and second experiments, respectively; β1 and β2 represent L... data1 L dat Weighting coefficients;

[0015]

[0016] In the formula, The equivalent Lamé coefficient; u is the particle vibration displacement; t is time; f idc The reverberant excitation source is denoted by μ and λ, which are Lamé coefficients; α is the fractional order; ω is the angular frequency of vibration; k is the wave number; x is the spatial position vector; i is the imaginary unit; A j The amplitude of the excitation source is j;

[0017]

[0018] In the formula, u1 is the displacement field predicted by the neural network in the FPINN model; u1 * The displacement field estimated for ultrafast ultrasonic plane wave imaging;

[0019]

[0020] In the formula, c s The shear wave velocity derived from the Lamé coefficients predicted by the neural network in the FPINN model; c s * The shear wave velocity is obtained using the reverberation method.

[0021] A further improvement to the method of the present invention is that,

[0022] In the step of obtaining the shear wave velocity at different frequencies based on the frequency domain information, the approximate wave value is first extracted from the frequency domain information using finite difference, and then the shear wave velocity is calculated based on the extracted approximate wave value.

[0023] The expression for calculating the approximate wave value is as follows:

[0024]

[0025] In the formula, It is an approximate wave value; C is equal to 10 / (Δx) 2 ·Bvv(0)) is a constant; Δx is the value of autocorrelation within a certain time period; Bvv(0) and Bvv(Δx) are the autocorrelation coefficients at Δt=0 and Δt=x;

[0026] The expression for calculating shear wave velocity is:

[0027]

[0028] In the formula, C s ω1 is the shear wave velocity; ω1 is the angular frequency of the shear wave; k is the extracted approximate wave value.

[0029] A further improvement to the method of the present invention is that,

[0030] In the step of obtaining the elasticity, viscosity, and fluidity data of breast tissue based on the shear wave velocity at different frequencies, the KVFD model is used to fit the shear wave velocity to obtain the elasticity, viscosity, and fluidity data of breast tissue.

[0031] A further improvement to the method of the present invention is that,

[0032] In the step of acquiring and processing the displacement data of the breast tissue to be imaged, the breast tissue imaging device based on low-frequency vibration and reverberant shear wave imaging acquires the displacement data of the breast tissue; and the displacement data of the breast tissue is subjected to Fourier transform through autocorrelation function to obtain frequency domain information.

[0033] The breast tissue imaging device based on low-frequency vibration and reverberant shear wave imaging includes a low-frequency resonance structure and an ultrasound detection structure. The low-frequency resonance structure includes a signal generator, a power amplifier, and a loudspeaker. One end of the power amplifier is connected to the signal generator, and the other end is used to connect to the breast tissue to be tested. The loudspeaker is placed on the breast tissue to be tested. The ultrasound detection structure is connected to the breast tissue to be tested. In use, the signal generator generates a continuous sinusoidal signal, which drives the loudspeaker to vibrate and generate shear waves that are transmitted to the breast tissue to be tested through the power amplifier. The ultrasound detection structure obtains radio frequency data by repeatedly emitting ultrasound waves and receiving reflected signals, and then performs beamforming on the radio frequency data to obtain displacement data of the breast tissue.

[0034] A further improvement to the method of the present invention is that,

[0035] The expression for the autocorrelation function is:

[0036] R(x, x) = E(X(t1)*X(t2));

[0037] In the formula, E is the expectation, X(t1) and X(t2) are the signal strengths at different times, and R(x, x) represents the degree of linear correlation between the values ​​of t1 and t2 at any different times.

[0038] In a second aspect, the present invention provides a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging system based on FPINN, comprising:

[0039] The data acquisition module is used to acquire and process the displacement data of the breast tissue to be imaged to obtain frequency domain information; based on the frequency domain information, it acquires the shear wave velocity at different frequencies, and obtains the elasticity, viscosity and flowability data of the breast tissue according to the shear wave velocity at different frequencies.

[0040] The imaging module is used to perform imaging processing based on the obtained elasticity, viscosity and fluidity data of breast tissue using a trained FPINN model to obtain imaging results of breast tissue.

[0041] In the training process of the FPINN model, the input is sample data of breast tissue elasticity, viscosity and fluidity, and the output is the imaging results of breast tissue obtained according to the viscoelastic properties of different locations inside the breast tissue. The parameters are updated by minimizing the defined loss function, and the training is completed after the preset convergence condition is reached.

[0042] Defined loss function L fPINN The expression is:

[0043] L fPINN =L fPDE +β1L data+β2L data2 ;

[0044] In the formula, L fPDE L is the model-driven loss. data1 L data2 These represent the data-driven losses from the first and second experiments, respectively; β1 and β2 represent L... data1 L data2 Weighting coefficients;

[0045]

[0046]

[0047] In the formula, The equivalent Lamé coefficient; u is the particle vibration displacement; t is time; f idc The reverberant excitation source is denoted by μ and λ, which are Lamé coefficients; α is the fractional order; ω is the angular frequency of vibration; k is the wave number; x is the spatial position vector; i is the imaginary unit; A j The amplitude of the excitation source is j;

[0048]

[0049] In the formula, u1 is the displacement field predicted by the neural network in the FPINN model; u1 * The displacement field estimated for ultrafast ultrasonic plane wave imaging;

[0050]

[0051] In the formula, c s The shear wave velocity derived from the Lamé coefficients predicted by the neural network in the FPINN model; c s * The shear wave velocity is obtained using the reverberation method.

[0052] A further improvement of the system of the present invention is that,

[0053] In the step of the data acquisition module to obtain the shear wave velocity at different frequencies based on the frequency domain information, the approximate wave value is first extracted from the frequency domain information using finite difference, and then the shear wave velocity is calculated based on the extracted approximate wave value.

[0054] The expression for calculating the approximate wave value is as follows:

[0055]

[0056] In the formula, It is an approximate wave value; C is equal to 10 / (Δx) 2·Bvv(0)) is a constant; Δx is the value of autocorrelation within a certain time period; Bvv(0) and Bvv(Δx) are the autocorrelation coefficients at Δt=0 and Δt=x;

[0057] The expression for calculating shear wave velocity is:

[0058]

[0059] In the formula, C s ω1 is the shear wave velocity; ω1 is the angular frequency of the shear wave; k is the extracted approximate wave value.

[0060] A further improvement of the system of the present invention is that,

[0061] In the step of obtaining the elasticity, viscosity, and fluidity data of breast tissue based on the shear wave velocity at different frequencies, the data acquisition module uses the KVFD model to fit the shear wave velocity and obtain the elasticity, viscosity, and fluidity data of breast tissue.

[0062] A further improvement of the system of the present invention is that,

[0063] In the step of acquiring and processing displacement data of the breast tissue to be imaged by the data acquisition module, the breast tissue imaging device based on low-frequency vibration and reverberant shear wave imaging acquires the displacement data of the breast tissue; and performs Fourier transform on the displacement data of the breast tissue through autocorrelation function to obtain frequency domain information.

[0064] The breast tissue imaging device based on low-frequency vibration and reverberant shear wave imaging includes a low-frequency resonance structure and an ultrasound detection structure. The low-frequency resonance structure includes a signal generator, a power amplifier, and a loudspeaker. One end of the power amplifier is connected to the signal generator, and the other end is used to connect to the breast tissue to be tested. The loudspeaker is placed on the breast tissue to be tested. The ultrasound detection structure is connected to the breast tissue to be tested. In use, the signal generator generates a continuous sinusoidal signal, which drives the loudspeaker to vibrate and generate shear waves that are transmitted to the breast tissue to be tested through the power amplifier. The ultrasound detection structure obtains radio frequency data by repeatedly emitting ultrasound waves and receiving reflected signals, and then performs beamforming on the radio frequency data to obtain displacement data of the breast tissue.

[0065] A further improvement of the system of the present invention is that,

[0066] The expression for the autocorrelation function is:

[0067] R(x, x) = E(X(t1)*x(t2));

[0068] In the formula, E is the expectation, X(t1) and X(t2) are the signal strengths at different times, and R(x, x) represents the degree of linear correlation between the values ​​of t1 and t2 at any different times.

[0069] Compared with the prior art, the present invention has the following beneficial effects:

[0070] This invention proposes a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging method based on FPINN. By utilizing a trained FPINN model for imaging processing, it can rapidly convert input breast tissue elasticity, viscosity, and fluidity data into imaging results. Compared to existing traditional imaging methods, it eliminates the need for complex manual feature extraction and lengthy computational processes. This invention leverages the parallel computing capabilities of neural networks and learned mapping relationships to efficiently generate imaging results, improving the real-time performance and accuracy of imaging. Explained, during the training phase of the FPINN model, it is trained using a large amount of breast tissue sample data (including known elasticity, viscosity, and fluidity data and their corresponding imaging result labels). The model learns the complex mapping relationship between breast tissue data and imaging results. This data-driven learning approach fully utilizes sample information, improving the model's adaptability to different situations. The FPINN model itself integrates physical information, enabling a better understanding and processing of breast tissue mechanical property data. By combining physical principles with neural networks, it can more accurately reflect the actual structure and state of breast tissue during imaging, avoiding imaging results that may not conform to physical laws due to purely data-driven methods, further improving imaging accuracy. Further interpretively, the fractional-order wave equation is embedded in the neural network as the first loss function L. fPDE The displacement field u1 estimated by ultrafast ultrasonic plane wave imaging * The displacement field u1 predicted by the neural network is used as the second loss function; the shear wave velocity c obtained by the reverberation method is used as the second loss function. s * The shear wave velocity c derived from the Lamé coefficient predicted by the neural network s As the third loss function; the fractional wave equation is embedded in the neural network as the loss function L. fPDEThe first loss function ensures that the neural network's predictions align with physical laws. The fractional-order wave equation more accurately describes wave phenomena in complex media, thus improving prediction accuracy. The second loss function compares the displacement field estimated by ultrafast ultrasonic plane wave imaging with that predicted by the neural network. Since ultrafast ultrasonic plane wave imaging itself has high accuracy, this comparison helps the neural network learn more accurate displacement field prediction capabilities, thereby improving imaging accuracy. The third loss function further constrains the neural network's predictions by comparing the shear wave velocity obtained through reverberation with the shear wave velocity derived from the Lamé coefficients predicted by the neural network, making it more consistent with physical reality and thus improving imaging accuracy. By simultaneously optimizing the three loss functions, the neural network can learn more generalized feature representations across different tasks. This multi-task learning approach helps improve the neural network's generalization ability, maintaining high accuracy under different datasets and conditions. Furthermore, by embedding physical information into the loss functions, the neural network can learn the inherent patterns of the data more quickly during training, accelerating the convergence process. This means the neural network can achieve higher accuracy in a shorter time, contributing to improved real-time imaging.

[0071] In a preferred embodiment of the present invention, displacement data of breast tissue is obtained through specific techniques, and frequency domain information is obtained by processing the displacement data. Based on the frequency domain information, shear wave velocities at different frequencies are obtained. This process requires accurate physical models and calculation methods. According to physical principles, shear wave velocity is related to the mechanical properties of tissue. Shear wave velocity at different frequencies is calculated using precise mathematical formulas and algorithms, thereby providing support for accurately obtaining data on the elasticity, viscosity, and fluidity of breast tissue.

[0072] In a preferred embodiment of the present invention, a breast tissue imaging device based on low-frequency vibration and reverberant shear wave imaging is proposed. A signal generator is set up to generate a continuous sinusoidal signal, and a power amplifier drives a speaker to vibrate and generate shear waves that are transmitted to the breast. At the same time, a high frame rate ultrasound probe is used to collect radio frequency data by emitting ultrasound waves multiple times and receiving reflected signals, which facilitates subsequent observation of the displacement of breast tissue under shear wave excitation. Specifically, the device connects a signal generator and a power amplifier to produce the vibration frequency required for the experiment. A speaker is fixed to the breast tissue, and the ultrasound probe is positioned above the breast. A low-frequency resonance device generates reverberant shear waves through multiple reflections within the breast tissue, causing minute displacements in the tissue. Simultaneously, the ultrasound probe emits high-frame-rate planar ultrasound waves and receives the echo signals to obtain radio frequency (RF) data. Beamforming is performed on the RF data to acquire breast tissue displacement data. This data is then processed to obtain frequency domain information. Based on this information, shear wave velocities at different frequencies are obtained, and the elasticity, viscosity, and fluidity data of the breast tissue are derived from these velocities. This device, utilizing the reverberant signal generated by low-frequency vibration, significantly improves the sensitivity of shear waves in detecting deep, minute lesions and enables viscoelastic fluid parameter imaging of breast tissue, providing comprehensive, high-resolution imaging of the breast tissue. This device offers enhanced safety, with low-frequency vibration having minimal impact on lesions, making it suitable for large-scale clinical application. In addition, the device's low-frequency vibration massage function can not only effectively stimulate shear wave signals, but also further enhance the comfort of the detection experience, ensuring that while conducting accurate screening, more attention is paid to optimizing the patient's experience. Attached Figure Description

[0073] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0074] Figure 1 This is a schematic flowchart of a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging method based on FPINN in an embodiment of the present invention.

[0075] Figure 2 This is a schematic diagram of the structure of a breast tissue imaging device based on low-frequency vibration and reverberant shear wave imaging in an embodiment of the present invention.

[0076] Figure 3 This is a detailed flowchart of a breast tissue imaging method based on low-frequency vibration and reverberant shear wave imaging, as described in an embodiment of the present invention.

[0077] Figure 4This is a schematic diagram of the FPINN model in an embodiment of the present invention;

[0078] Figure 5 This is a schematic diagram showing the particle vibration results at the maximum displacement at different frequencies in an embodiment of the present invention; wherein, Figure 5 (a) is a schematic diagram at 100Hz. Figure 5 (b) is a schematic diagram at 200Hz. Figure 5 (c) is a schematic diagram at 300Hz. Figure 5 The diagram in (d) is at 400Hz. Figure 5 (e) is a schematic diagram at 500Hz. Figure 5 The diagram in (f) is at 600Hz. Figure 5 The diagram in (g) is at 700Hz. Figure 5 The diagram in the middle (h) is at 800Hz. Figure 5 In diagram (i), the frequency is 900 Hz. Figure 5 (j) is a schematic diagram at 1000Hz;

[0079] Figure 6 This is a schematic diagram of the shear wave imaging results of deep micro-abnormal tissues at different frequencies in an embodiment of the present invention; wherein, Figure 6 (a) is a schematic diagram at 100Hz. Figure 6 (b) is a schematic diagram at 200Hz. Figure 6 (c) is a schematic diagram at 300Hz. Figure 6 The diagram in (d) is at 400Hz. Figure 6 (e) is a schematic diagram at 500Hz. Figure 6 The diagram in (f) is at 600Hz. Figure 6 The diagram in (g) is at 700Hz. Figure 6 The diagram in the middle (h) is at 800Hz. Figure 6 In diagram (i), the frequency is 900 Hz. Figure 6 (j) is a schematic diagram at 1000Hz;

[0080] Figure 7 This is a schematic diagram of a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging system based on FPINN in an embodiment of the present invention.

[0081] The explanations of the reference numerals in the figure are as follows:

[0082] 1. Signal generator; 2. Power amplifier; 3. Speaker; 4. Ultrasonic detection structure. Detailed Implementation

[0083] To make the objectives, technical solutions, and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention; obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0084] Based on the technical solutions disclosed in the embodiments of this invention, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this invention. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to these processes, methods, products, or devices.

[0085] Please see Figure 1 This invention provides a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging method based on FPINN, comprising the following steps:

[0086] Step 1: Obtain and process the displacement data of the breast tissue to be imaged to obtain frequency domain information; based on the frequency domain information, obtain the shear wave velocity at different frequencies, and obtain the elasticity, viscosity and flowability data of the breast tissue according to the shear wave velocity at different frequencies.

[0087] Step 2: Based on the obtained data on the elasticity, viscosity and fluidity of breast tissue, the trained FPINN model is used for imaging processing to obtain the imaging results of breast tissue.

[0088] In the training process of the FPINN model, the input is sample data of breast tissue elasticity, viscosity and fluidity, and the output is the imaging results of breast tissue obtained according to the viscoelastic properties of different locations inside the breast tissue. The parameters are updated by minimizing the defined loss function, and the training is completed after the preset convergence condition is reached.

[0089] Defined loss function L fPINN The expression is:

[0090] L fPINN =L fPDE +β1L data1 +β2L dat ;

[0091] In the formula, L fPDE L is the model-driven loss. data1 L data2 These represent the data-driven losses from the first and second experiments, respectively; β1 and β2 represent L... data1 L data2Weighting coefficients;

[0092]

[0093] In the formula, The equivalent Lamé coefficient; u is the particle vibration displacement; t is time; f idc The reverberant excitation source is denoted by μ and λ, which are Lamé coefficients; α is the fractional order; ω is the angular frequency of vibration; k is the wave number; x is the spatial position vector; i is the imaginary unit; A j The amplitude of the excitation source is j;

[0094]

[0095] In the formula, u1 is the displacement field predicted by the neural network in the FPINN model; u1 * The displacement field estimated for ultrafast ultrasonic plane wave imaging;

[0096]

[0097] In the formula, c s The shear wave velocity derived from the Lamé coefficients predicted by the neural network in the FPINN model; c s * The shear wave velocity is obtained using the reverberation method.

[0098] In practical applications, neural networks are usually much faster at making predictions than traditional physical simulation methods. By embedding physical information into neural networks and training them, more efficient computation can be achieved while maintaining high accuracy. This helps to improve the real-time performance of imaging, enabling it to meet the needs of real-time monitoring and diagnosis.

[0099] Please see Figure 1 and Figure 2 The present invention provides a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging method based on FPINN, comprising the following steps:

[0100] Breast tissue displacement data is acquired, and frequency domain information is obtained by processing the breast tissue displacement data. Based on the frequency domain information, shear wave velocity at different frequencies is obtained, and elasticity, viscosity, and fluidity data of breast tissue are obtained based on the shear wave velocity at different frequencies.

[0101] In the exemplary technical solution, breast tissue displacement data is acquired based on a breast tissue imaging device using low-frequency vibration and reverberant shear wave imaging. A structural diagram of the breast tissue imaging device is shown below. Figure 2As shown, the system includes a low-frequency resonance structure and an ultrasound detection structure 4. The low-frequency resonance structure includes a signal generator 1, a power amplifier 2, and a speaker 3. One end of the power amplifier 2 is connected to the signal generator 1, and the other end is connected to the breast tissue to be tested. The speaker 3 is placed on the breast tissue to be tested. The ultrasound detection structure 4 is connected to the breast tissue to be tested. The signal generator 1 and the power amplifier 2 are correctly connected to generate the vibration frequency required for the experiment. The ultrasound probe of the ultrasound detection structure 4 is connected to the breast tissue to be tested via a coupling agent. The low-frequency resonance frequency is between 100Hz and 1000Hz, with a step size of 100Hz. There are several speakers 3, which are evenly arranged on the breast tissue to be tested. In this embodiment, four speakers 3 are fixed in the four directions of the breast tissue (up, down, left, and right) for the experiment. Because the breast tissue is relatively soft and has an irregular surface, soft materials are used to attach and fix the speakers 3 to the breast tissue to ensure that the contact surface fully transmits vibration. These materials include, but are not limited to, silicone pads, medical tape, flexible bandages, flexible supports, gel patches, etc. The power amplifier's gain is set to X2 or X10. It features a high input impedance of 10kΩ and a complete output protection circuit (output overcurrent protection, internal temperature abnormality protection) to ensure stable, reliable, and safe operation. It is small, lightweight, and easy to use, suitable for various environments. The speaker's frequency range is 80Hz–20kHz, with a diameter of 30mm and a total weight not exceeding 50g. Compact and lightweight, it is a speaker capable of generating low-frequency resonance and multi-functional vibration massage therapy, suitable for various environments. This speaker is essentially a low-frequency resonant vibration horn with multi-functional massage effects. During patient testing, it provides a certain massage effect, effectively relieving patient tension and promoting relaxation, which is beneficial for data acquisition. The signal generator is an arbitrary waveform generator with an output power of less than or equal to 60MHz, a vertical resolution of 14 bits, and a sampling rate of 200Msa / s. This signal generator can generate high-quality arbitrary waveforms, ensuring smoother and more detailed signal changes in the time domain.

[0102] Explaining the principle, signal generator 1 generates a continuous sinusoidal signal, which drives speaker 3 to vibrate through power amplifier 2, generating shear waves that are transmitted to the breast tissue. Then, ultrasonic detection structure 4 uses multiple ultrasonic waves to transmit and receive reflected signals to obtain radio frequency data, which facilitates subsequent observation of the displacement of breast tissue under shear wave excitation. Further explaining, when fixing the low-frequency vibrating speaker to the breast tissue for shear wave experiments, the key is to ensure good contact and stable fixation between the vibration source and the tissue, avoiding detachment or large positional displacement during vibration, so as to effectively transmit vibration without damaging the tissue or affecting the normal experimental results.

[0103] Radio frequency data is beamformed to obtain breast tissue displacement data. The breast tissue displacement data is processed to obtain frequency domain information. Based on the frequency domain information, shear wave velocity at different frequencies is obtained. Based on the shear wave velocity at different frequencies, the elasticity, viscosity, and fluidity data of breast tissue are obtained.

[0104] The process of processing breast tissue displacement data to obtain frequency domain information specifically involves:

[0105] Frequency domain information is obtained by performing Fourier transform on breast tissue displacement data using the autocorrelation function.

[0106] The autocorrelation function is expressed as follows:

[0107] R(x, x) = E(X(t1)*x(t2));

[0108] In the formula, E is the expectation, X(t1) and X(t2) are the signal strengths at different times, and R(x, x) represents the degree of linear correlation between the values ​​of t1 and t2 at any different times.

[0109] In this embodiment of the invention, the shear wave velocity at different frequencies is obtained based on frequency domain information. Specifically, the approximate wave value is first extracted from the frequency domain information using finite difference, and then the shear wave velocity is calculated based on the extracted approximate wave value.

[0110] The expression for calculating the approximate wave value is as follows:

[0111]

[0112] In the formula, It is an approximate wave value; C is equal to 10 / (Δx) 2 ·Bvv(0)) is a constant; Δx is the value of autocorrelation within a certain time period; Bvv(0) and Bvv(Δx) are the autocorrelation coefficients at Δt=0 and Δt=x;

[0113] The expression for calculating shear wave velocity is:

[0114]

[0115] In the formula, C s ω1 is the shear wave velocity; ω1 is the angular frequency of the shear wave; k is the extracted approximate wave value.

[0116] The elasticity, viscosity, and fluidity data of breast tissue were obtained based on the shear wave velocity at different frequencies. Specifically, the KVFD model (Kelvin-Voigt Fractional Derivative Model) was used to fit the shear wave velocity to obtain the elasticity, viscosity, and fluidity data of breast tissue.

[0117] The following is combined Figure 3 and Figure 4 Detailed Description: In this embodiment of the invention, the low-frequency resonance device generates reverberant shear waves through multiple reflections in breast tissue, causing minute displacements in the breast tissue. Specifically: a high frame rate imaging method is used to track the displacement changes within the tissue; a Loupas displacement estimator is used to estimate the phase change of the reflected ultrasound echo to calculate the axial displacement, thereby obtaining an estimate of the minute displacement; simultaneously, the probe of the ultrasound detection structure emits high frame rate planar ultrasound waves and receives the echo signals to obtain radio frequency data; the radio frequency data is filtered and beamformed; the beamformation method uses the classic algorithm DAS (Delay-and-Sum) to obtain the synthesized beam; and breast tissue displacement data is obtained based on the synthesized beam.

[0118] Frequency domain information is obtained by performing Fourier transform on breast tissue displacement data using the autocorrelation function.

[0119] Finite difference is used to approximate the extraction of the wavenumber from the frequency domain information. Based on approximate wave number The shear wave velocity is calculated, and a low-frequency resonant device is excited with low-frequency sinusoidal electrical signals of different frequencies to generate reverberant shear wave fields at different frequencies. The propagation velocity of the shear wave at different frequencies is obtained. The above steps are repeated with low-frequency sinusoidal electrical signals of different frequencies to obtain the shear wave velocity at different frequencies.

[0120] Using the KVFD model to fit the shear wave velocity, we can obtain the multi-parameter mechanical properties of the tissue, including elasticity, viscosity, and flowability. Its constitutive equation is: σ(t)=E0ε(t)+ξD α [ε(t)]; This formula describes the constitutive relation of soft tissue, where σ(t) and ε(t) represent stress and strain under external force, respectively, E0 represents the Young's modulus of the tissue, i.e., elasticity, and ξ represents the viscosity of the tissue, i.e., viscosity, in Pa·s. a D a Represents the fractional derivative, 'a' represents the fluidity of the organization, using... Figure 4 From the formula and the KVFD equation, we can obtain

[0121]

[0122] This formula can be used to fit multi-frequency shear wave velocities, thereby achieving viscoelastic flow parametric imaging. By calculating the viscosity, elasticity, and flow properties of tissue, it overcomes the limitations of existing imaging methods that can only perform elasticity measurements of breast tissue, and is of great significance for the screening and diagnosis of early breast tumors.

[0123] In this embodiment of the invention, imaging results of breast tissue are obtained by using a trained FPINN model based on data on the elasticity, viscosity, and flowability of breast tissue. Fractional-order models, compared to integer-order models, can more accurately describe the physical properties of soft matter; therefore, a fractional-order Kelvin-Voigt model (KVFD) is used to describe the constitutive equations. For example... Figure 4 As shown, the Fractional Physical Information Neural Network (FPINN) is a method that combines deep learning with fractional physical modeling to solve fractional partial differential equation (PDE) problems. The key idea of ​​FPINN is to embed fractional physical constraints into the neural network, enabling the network to learn the behavior of the physical system and satisfy the physical equations. The steps for obtaining the FPINN model include: defining the physical model, defining the physical equations and constraints of the system; constructing the neural network, building a neural network structure containing physical information; designing a loss function to train the network: training the network by minimizing the loss function using an optimization algorithm to make it approximate the physical equations and satisfy the constraints; solving the problem: using the trained network to solve the physical problem.

[0124] The FPINN model of this invention uses positions x, y, z and time t as inputs to the neural network, presents the training input data to the neural network, and uses a loss function to compare the output result with the training output data. The loss returned by the function is used to adjust the weights of the network through backpropagation to reduce the loss.

[0125] This invention, FPINN, uses a custom loss function that includes additional loss components to constrain the neural network to produce outputs that conform to the modeled differential equation. It predicts the numerical solution of the partial differential equation under these conditions and generates a corresponding displacement *u* as output. Training FPINN to conform to the requirements of the differential equation requires outputting the corresponding fractional derivative and second derivative, i.e. and These derivatives can be obtained in TensorFlow and PyTorch using the automatic differentiation function for each platform.

[0126] The distinctive feature of the technical solution in this invention lies in the novel design of the Loss function, which comprises three parts:

[0127] Part 1 L fPDE Model-driven loss function:

[0128]

[0129] The second part is the loss function driven by the ultrasound experimental data:

[0130]

[0131] The third part is the loss function driven by the KVFD model and reverberation experimental data:

[0132]

[0133] The loss function of the entire FPINN neural network is a weighted sum of three loss functions:

[0134] L fPINN =L fPDE +β1L data1 +β2L data2 ;

[0135] The technical solution of this invention utilizes FPINN for one-time training, enabling the learning of mechanical parameters at multiple frequencies, and uses the KVFD model to achieve viscoelastic flow parameter imaging. Furthermore, multiple reverberation excitation sources of various frequencies are provided during network training, and the trained network can achieve the learning of physical parameters excited by multiple reverberation sources of various frequencies, greatly improving learning efficiency.

[0136] Based on this, utilize By fitting viscoelastic flow parameters, viscoelastic flow parametric imaging of breast tissue can be achieved using FPINN.

[0137] Figure 5 The diagram shows the particle vibration displacement at the maximum displacement at different frequencies. Figure 5 Figures (a) to (j) show the time-domain response of the vibration system with frequencies ranging from 100 to 1000 Hz. The results show that the initial displacement of the particle is large, and the amplitude gradually decreases over time, exhibiting typical damped vibration characteristics, which are consistent with the attenuation characteristics and propagation law of vibration signals in soft tissue.

[0138] Figure 6 The image shows the results of shear wave imaging of deep, minute abnormal tissues at different frequencies. Figure 6 As can be clearly seen in (a), the shear wave imaging results are significantly worse at a vibration frequency of 100Hz, making it impossible to accurately calculate the shear wave velocity of deep, minute abnormal tissues; from Figure 6 (b) and Figure 6 In diagram (c), it is evident that the shear wave velocity of deep, minute abnormal tissues can be accurately calculated at vibration frequencies of 200Hz and 300Hz, but a strong interference signal is generated in the top region of the soft tissue; from Figure 6 (d) and Figure 6 It can be clearly seen from the middle (h) that when the vibration frequency is 400H~800Hz, the shear wave velocity of deep micro-abnormal tissues can be accurately calculated, and the surrounding interference signal is relatively weak; when the low-frequency vibration signal is further enhanced, such as Figure 6 Zhong (i) and Figure 6As shown in Figure (j), when the frequencies are 900Hz and 1000Hz, interference signals are generated in the surrounding environment, and the shear wave velocity of deep micro-abnormal tissue cannot be accurately calculated. Therefore, based on the experimental results, low-frequency vibrations in the range of 400Hz to 800Hz can effectively enhance the reverberation field strength, thereby effectively distinguishing the shear wave velocity of deep micro-abnormal tissue from the surrounding normal tissue.

[0139] Specifically, an exemplary embodiment of the present invention provides a breast tissue imaging method based on the FPINN model, comprising:

[0140] Step (1) involves acquiring and processing displacement data of breast tissue, including:

[0141] Data Acquisition: Ultrafast ultrasound imaging technology is used to scan breast tissue and acquire displacement data of the tissue at different time points. These data can be represented as a time series displacement field, i.e., u(x,t), where x is the spatial location and t is the time.

[0142] Frequency domain information processing: Perform Fourier transform on the acquired displacement data to convert it from the time domain to the frequency domain, and obtain frequency domain information U(k,ω), where k is the wave number and ω is the vibration angular frequency.

[0143] Shear wave velocity calculation: Based on frequency domain information, the shear wave velocity c_s(ω) at different frequencies is calculated using wave theory. The shear wave velocity is closely related to the elastic, viscous and fluid properties of the tissue.

[0144] Obtain breast tissue property data: Based on the shear wave velocity c_s(ω), combined with the known wave equation and material mechanics model, the elasticity, viscosity and flow data of breast tissue are derived. These data can be expressed as functions of spatial position, namely E(x), ν(x), η(x), etc., where E is the elastic modulus, ν is Poisson's ratio, and η is the viscosity coefficient.

[0145] Step (2): Imaging processing using the FPINN model, including:

[0146] FPINN Model Training: Input Data: Collect a batch of breast tissue elasticity, viscosity, and fluidity sample data as input to the FPINN model; Output Data: Based on the viscoelastic properties of different locations within the breast tissue, obtain the corresponding breast tissue imaging results as output to the FPINN model; By minimizing the loss function, use an optimization algorithm (such as gradient descent) to update the parameters of the FPINN model. When the loss function reaches the preset convergence condition, the training of the FPINN model is complete.

[0147] Imaging processing: The acquired data on the elasticity, viscosity and fluidity of breast tissue are input into the trained FPINN model. Based on the input data, the FPINN model, combined with physical laws and experimental data, predicts the imaging results of breast tissue. The imaging results can be expressed as a function of spatial position, i.e., I(x), where I is the imaging intensity or imaging parameter.

[0148] Specific example parameters: Frequency domain information processing: The Fast Fourier Transform (FFT) algorithm is used for Fourier transform, and the frequency domain range is selected as 0 to 1000 Hz; Shear wave velocity calculation: The shear wave velocity is calculated using the wave equation, considering the influence of the fractional derivative, and the equivalent Lamé coefficient is calculated using the formula. FPINN model training: Input data: 1000 sets of breast tissue elasticity, viscosity, and fluidity sample data; Output data: 1000 sets of corresponding breast tissue imaging results; Loss function weight coefficients: β1 = 0.5, β2 = 0.5; Optimization algorithm: The Adam optimization algorithm is used for parameter updates; Convergence condition: The loss function value is less than 0.01. Imaging processing: The acquired breast tissue characteristic data is input into the FPINN model, and the FPINN model outputs the imaging results of the breast tissue. The imaging results are displayed in grayscale image form, where the grayscale value represents the imaging intensity. Through the above steps and parameter settings, accurate imaging of breast tissue can be achieved, which can provide strong support for the diagnosis and treatment of breast diseases.

[0149] The following are embodiments of the apparatus of the present invention, which can be used to execute embodiments of the method of the present invention. For details not disclosed in the apparatus embodiments, please refer to the embodiments of the method of the present invention.

[0150] Please see Figure 7 In this embodiment of the invention, a low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging system based on FPINN is provided, comprising:

[0151] The data acquisition module is used to acquire and process the displacement data of the breast tissue to be imaged to obtain frequency domain information; based on the frequency domain information, it acquires the shear wave velocity at different frequencies, and obtains the elasticity, viscosity and flowability data of the breast tissue according to the shear wave velocity at different frequencies.

[0152] The imaging module is used to perform imaging processing based on the obtained elasticity, viscosity and fluidity data of breast tissue using a trained FPINN model to obtain imaging results of breast tissue.

[0153] In the training process of the FPINN model, the input is sample data of breast tissue elasticity, viscosity and fluidity, and the output is the imaging results of breast tissue obtained according to the viscoelastic properties of different locations inside the breast tissue. The parameters are updated by minimizing the defined loss function, and the training is completed after the preset convergence condition is reached.

[0154] Defined loss function L fPINN The expression is:

[0155] L fPINN =L fPDE +β1L data1 +β2L da ;

[0156] In the formula, L fPD E is the model-driven loss, L data L data2 These represent the data-driven losses from the first and second experiments, respectively; β1 and β2 represent L... data L data2 The weighting coefficients.

[0157] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging method based on FPINN, characterized in that, Includes the following steps: The displacement data of the breast tissue to be imaged is acquired and processed to obtain frequency domain information; Based on the frequency domain information, shear wave velocities at different frequencies are obtained, and data on the elasticity, viscosity, and fluidity of breast tissue are obtained based on the shear wave velocities at different frequencies. Based on the obtained data on the elasticity, viscosity and fluidity of breast tissue, the trained FPINN model was used for imaging processing to obtain the imaging results of breast tissue. In the training process of the FPINN model, the input is sample data of breast tissue elasticity, viscosity and fluidity, and the output is the imaging results of breast tissue obtained according to the viscoelastic properties of different locations inside the breast tissue. The parameters are updated by minimizing the defined loss function, and the training is completed after the preset convergence condition is reached. Defined loss function L fPINN The expression is: L fPINN =L fPDE +β1L data1 +β2L data2 ; In the formula, L fPDE For model-driven loss, L data1 L data2 These represent the data-driven losses from the first and second experiments, respectively; β1 and β2 represent L... data1 L data2 Weighting coefficients; In the formula, The equivalent Lamé coefficient; u is the particle vibration displacement; t is time; f idc The reverberant excitation source is denoted by μ and λ, which are Lamé coefficients; α is the fractional order; ω is the angular frequency of vibration; k is the wave number; x is the spatial position vector; i is the imaginary unit; A j The amplitude of the excitation source is j; In the formula, u1 is the displacement field predicted by the neural network in the FPINN model; u1 * The displacement field estimated for ultrafast ultrasonic plane wave imaging; In the formula, c s The shear wave velocity derived from the Lamé coefficients predicted by the neural network in the FPINN model; c s * The shear wave velocity is obtained using the reverberation method.

2. The method for low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging based on FPINN according to claim 1, characterized in that, In the step of obtaining the shear wave velocity at different frequencies based on the frequency domain information, the approximate wave value is first extracted from the frequency domain information using finite difference, and then the shear wave velocity is calculated based on the extracted approximate wave value. The expression for calculating the approximate wave value is as follows: In the formula, It is an approximate wave value; C is equal to 10 / (Δx) 2 ·Bvv(0)) is a constant; Δx is the value of autocorrelation within a certain time period; Bvv(0) and Bvv(Δx) are the autocorrelation coefficients at Δt=0 and Δt=x; The expression for calculating shear wave velocity is: In the formula, C s ω1 is the shear wave velocity; ω1 is the angular frequency of the shear wave; k is the extracted approximate wave value.

3. The method for low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging based on FPINN according to claim 1, characterized in that, In the step of obtaining the elasticity, viscosity, and fluidity data of breast tissue based on the shear wave velocity at different frequencies, the KVFD model is used to fit the shear wave velocity to obtain the elasticity, viscosity, and fluidity data of breast tissue.

4. The method for low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging based on FPINN according to claim 1, characterized in that, In the step of acquiring and processing the displacement data of the breast tissue to be imaged, the breast tissue imaging device based on low-frequency vibration and reverberant shear wave imaging acquires the displacement data of the breast tissue; and the displacement data of the breast tissue is subjected to Fourier transform through autocorrelation function to obtain frequency domain information. The breast tissue imaging device based on low-frequency vibration and reverberant shear wave imaging includes a low-frequency resonance structure and an ultrasound detection structure (4); the low-frequency resonance structure includes a signal generator (1), a power amplifier (2), and a loudspeaker (3); one end of the power amplifier (2) is connected to the signal generator (1), and the other end of the power amplifier (2) is used to connect to the breast tissue to be tested; the loudspeaker (3) is used to be placed on the breast tissue to be tested; the ultrasound detection structure (4) is used to connect to the breast tissue to be tested; in use, the signal generator (1) is used to generate a continuous sine wave, and the power amplifier (2) drives the loudspeaker (3) to vibrate to generate a shear wave that is transmitted to the breast tissue to be tested; the ultrasound detection structure (4) is used to obtain radio frequency data by emitting ultrasound multiple times and receiving reflected signals, and to obtain the displacement data of the breast tissue by beamforming the radio frequency data.

5. The method for low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging based on FPINN according to claim 4, characterized in that, The expression for the autocorrelation function is: R(x,x)=E(X(t1)*X(t2)); In the formula, E is the expectation, X(t1) and X(t2) are the signal strengths at different times, and R(x,x) represents the degree of linear correlation between the values ​​of t1 and t2 at any different times.

6. A low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging system based on FPINN, characterized in that, include: The data acquisition module is used to acquire and process the displacement data of the breast tissue to be imaged to obtain frequency domain information; Based on the frequency domain information, shear wave velocities at different frequencies are obtained, and data on the elasticity, viscosity, and fluidity of breast tissue are obtained based on the shear wave velocities at different frequencies. The imaging module is used to perform imaging processing based on the obtained elasticity, viscosity and fluidity data of breast tissue using a trained FPINN model to obtain imaging results of breast tissue. In the training process of the FPINN model, the input is sample data of breast tissue elasticity, viscosity and fluidity, and the output is the imaging results of breast tissue obtained according to the viscoelastic properties of different locations inside the breast tissue. The parameters are updated by minimizing the defined loss function, and the training is completed after the preset convergence condition is reached. Defined loss function L fPINN The expression is: L fPINN =L fPDE +β1L data1 +β2L dat ; In the formula, L fPDE For model-driven loss, L data1 L data2 These represent the data-driven losses from the first and second experiments, respectively; β1 and β2 represent L... dat L data2 Weighting coefficients; In the formula, The equivalent Lamé coefficient; u is the particle vibration displacement; t is time; f idc The reverberant excitation source is denoted by μ and λ, which are Lamé coefficients; α is the fractional order; ω is the angular frequency of vibration; k is the wave number; x is the spatial position vector; i is the imaginary unit; A j The amplitude of the excitation source is j; In the formula, u1 is the displacement field predicted by the neural network in the FPINN model; u1 * The displacement field estimated for ultrafast ultrasonic plane wave imaging; In the formula, c s The shear wave velocity derived from the Lamé coefficients predicted by the neural network in the FPINN model; c s * The shear wave velocity is obtained using the reverberation method.

7. A low-frequency reverberation breast ultrasound viscoelastic ultrafast imaging system based on FPINN according to claim 6, characterized in that, In the step of the data acquisition module to obtain the shear wave velocity at different frequencies based on the frequency domain information, the approximate wave value is first extracted from the frequency domain information using finite difference, and then the shear wave velocity is calculated based on the extracted approximate wave value. The expression for calculating the approximate wave value is as follows: In the formula, It is an approximate wave value; C is equal to 10 / (Δx) 2 ·Bvv(0)) is a constant; Δx is the value of autocorrelation within a certain time period; Bvv(0) and Bvv(Δx) are the autocorrelation coefficients at Δt=0 and Δt=x; The expression for calculating shear wave velocity is: In the formula, C s ω1 is the shear wave velocity; ω1 is the angular frequency of the shear wave; k is the extracted approximate wave value.

8. A low-frequency reverberation breast ultrasound viscoelastic ultrafast imaging system based on FPINN according to claim 6, characterized in that, In the step of obtaining the elasticity, viscosity, and fluidity data of breast tissue based on the shear wave velocity at different frequencies, the data acquisition module uses the KVFD model to fit the shear wave velocity and obtain the elasticity, viscosity, and fluidity data of breast tissue.

9. A low-frequency reverberant breast ultrasound viscoelastic ultrafast imaging system based on FPINN according to claim 6, characterized in that, In the step of acquiring and processing displacement data of the breast tissue to be imaged by the data acquisition module, the breast tissue imaging device based on low-frequency vibration and reverberant shear wave imaging acquires the displacement data of the breast tissue; and performs Fourier transform on the displacement data of the breast tissue through autocorrelation function to obtain frequency domain information. The breast tissue imaging device based on low-frequency vibration and reverberant shear wave imaging includes a low-frequency resonance structure and an ultrasound detection structure (4); the low-frequency resonance structure includes a signal generator (1), a power amplifier (2), and a loudspeaker (3); one end of the power amplifier (2) is connected to the signal generator (1), and the other end of the power amplifier (2) is used to connect to the breast tissue to be tested; the loudspeaker (3) is used to be placed on the breast tissue to be tested; the ultrasound detection structure (4) is used to connect to the breast tissue to be tested; in use, the signal generator (1) is used to generate a continuous sine wave, and the power amplifier (2) drives the loudspeaker (3) to vibrate to generate a shear wave that is transmitted to the breast tissue to be tested; the ultrasound detection structure (4) is used to obtain radio frequency data by emitting ultrasound multiple times and receiving reflected signals, and to obtain the displacement data of the breast tissue by beamforming the radio frequency data.

10. A low-frequency reverberation breast ultrasound viscoelastic ultrafast imaging system based on FPINN according to claim 9, characterized in that, The expression for the autocorrelation function is: R(x,x)=E(X(t1)*X(t2)); In the formula, E is the expectation, X(t1) and X(t2) are the signal strengths at different times, and R(x,x) represents the degree of linear correlation between the values ​​of t1 and t2 at any different times.

Images