A 4D intracardiac ultrasound coherent compound imaging method, device and medium

Abstract

This invention relates to the field of ultrasound imaging technology, and discloses a 4D intracardiac ultrasound coherent composite imaging method, device, and medium. The method includes: acquiring a target transmitted waveform; obtaining an echo dataset after transmitting the target transmitted waveform; obtaining subarray beam signals based on the echo dataset; obtaining beamforming lines based on the subarray beam signals; calculating the beamforming lines using a coherent superposition algorithm to obtain superimposed beamforming lines; obtaining a two-dimensional ultrasound image dataset based on the superimposed beamforming lines; improving the clarity of two-dimensional ultrasound imaging by acquiring and calculating the echo dataset; improving beamforming rate and accuracy by dividing beamforming into two stages; inputting the two-dimensional ultrasound image dataset into a trained cardiac center prediction model to obtain the center coordinates within the target heart; thereby improving the clarity of the 4D ultrasound image.

Description

Technical Field

[0001] This invention relates to the field of ultrasound imaging technology, and in particular to a 4D intracardiac ultrasound coherent composite imaging method, device, and medium. Background Technology

[0002] As one of the core technologies in the field of medical imaging, three-dimensional ultrasound imaging technology acquires multi-angle volume data of the target area through ultrasound phased array, reconstructs and generates a stereo image through computer algorithms, and then reconstructs the stereo image in real time to obtain a 4D ultrasound image, realizing a leap from planar observation to spatial analysis.

[0003] However, existing 4D ultrasound images have low frame rates, resulting in low imaging resolution; moreover, 3D ICE based on rectangular transducer arrays are expensive, have limited imaging arrays, and poor orthogonal plane image quality, making ultrasound imaging unclear; establishing a traditional external coordinate system to create 4D ultrasound images will reduce the quality of 4D ultrasound images.

[0004] Therefore, a new 4D ultrasound imaging method is needed to improve the clarity of 4D ultrasound imaging. Summary of the Invention

[0005] This invention provides a 4D intracardiac ultrasound coherent composite imaging method, device, and medium to improve the clarity of 4D ultrasound images.

[0006] The first aspect of this invention discloses a 4D intracardiac ultrasound coherent composite imaging method, the method comprising: Obtain the echo dataset; wherein the echo dataset is obtained by transmitting a target waveform to the target object; Based on the echo dataset, subarray beam signals are obtained; based on the subarray beam signals, beamforming lines are obtained; the beamforming lines are calculated using a coherent superposition algorithm to obtain superimposed beamforming lines; based on the superimposed beamforming lines, a two-dimensional ultrasound image dataset is obtained. The two-dimensional ultrasound image dataset is fed back into a pre-trained cardiac center prediction model to obtain the center coordinates of the target object; based on the two-dimensional ultrasound image dataset and the center coordinates of the target object, a 4D ultrasound image is obtained. The trained cardiac center prediction model was obtained through the following method: Obtain a cardiac ultrasound image dataset; use the cardiac ultrasound image dataset as a training dataset to train an initial cardiac center prediction model. During the training process, constrain the training of the initial cardiac center prediction model through a morphological constraint function to obtain a trained cardiac center prediction model.

[0007] As an optional implementation, in the first aspect of the invention, the echo dataset is obtained by: Determine the target transmission direction; the target transmission direction is used to describe the direction in which the target waveform is transmitted toward the target object; Based on the target's transmission direction, the target receiving point is obtained; Based on the target receiving point, the target transmitted waveform is transmitted several times to obtain an enhanced transmitted beam; wherein the target transmission direction of the target transmitted waveform is the same in the several transmissions; Based on the enhanced transmitted beam, an enhanced echo is obtained; based on the enhanced echo, an echo dataset is obtained.

[0008] As an optional implementation, in the first aspect of the present invention, the method further includes: The preset ultrasonic phased array is divided into several N-order phased arrays. N subarrays; where N represents a preset positive integer; Binding is performed on all the array elements in the subarray to obtain a binding unit; the binding unit is then bound on a preset control unit to obtain a delay register unit; wherein, the delay register unit is used to enable the transmit or receive channels of all the array elements in the subarray to read the delay value and weighting coefficient from the delay register unit; The subarray delay value is obtained based on the geometric center coordinates and target focal length within the subarray; the subarray delay value and a preset weighting coefficient are fed back into the delay register unit to obtain an improved phased array.

[0009] As an optional implementation, in the first aspect of the present invention, obtaining the subarray beam signal based on the echo dataset includes: Obtain the target distance from the target element to the array center in the improved phased array; based on the target distance, obtain the total analog delay corresponding to the target element; Obtain the minimum delay step size in the improved phased array; obtain the total simulation delay series based on the total simulation delay and the minimum delay step size; obtain the total computational delay based on the total simulation delay series and the minimum delay step size. Based on a preset clock frequency, a first analog delay step size is obtained; based on the first analog delay step size and the total computational delay, a first analog delay level is obtained. The first simulation delay is obtained based on the first simulation delay level and the first simulation delay step size; the second simulation delay is obtained based on the first simulation delay and the total computation delay. Based on the first simulated delay and the second simulated delay, the echo dataset is processed using a delay alignment algorithm to obtain an RF signal; the RF signal is then processed using a first simulated weighting algorithm to obtain a subarray beam signal.

[0010] As an optional implementation, in the first aspect of the invention, obtaining the beamforming line based on the subarray beam signal includes: The second analog weighting algorithm is used to process all the subarray beam signals to obtain the beamforming line; The calculation method of the second simulated weighting algorithm is as follows:

[0011]

[0012] In the above formula, The delay applied to the signal of the m-th subarray beam. Let F be the distance from the center of the m-th subarray to the center of the array, and let F be the distance from the center of the m-th subarray to the center of the array. The distance from each array element to the target object, where c is the speed of sound in a specific medium. These are the weighting coefficients at the subarray level. Here, M represents the subarray beam signal, and M is the number of subarrays participating in the global synthesis. This is the total time scale. For beamforming lines.

[0013] As an optional implementation, in the first aspect of the present invention, the step of calculating the beamforming line using a coherent superposition algorithm to obtain the superimposed beamforming line includes: Determine several spatial points on the beamforming line; perform the following operation on each spatial point: Obtain the effective aperture array element number H corresponding to the ultrasonic beam emitted in the improved phased array; move the target receiving point along the array direction by aH array element distances to obtain multiple target moving receiving points corresponding to the spatial point; wherein, a represents a specific numerical value, and H represents a preset positive integer; Beamforming is performed on each of the target mobile receiving points to obtain a complex synthesized signal corresponding to each target mobile receiving point; The complex composite signal is calculated using a coherent superposition algorithm to obtain the calculated amplitude of the complex signal; The amplitude is calculated based on the complex signal corresponding to all the spatial points to obtain the superimposed beamforming line.

[0014] As an optional implementation, in the first aspect of the present invention, constraining the training of the initial cardiac center prediction model through a morphological constraint function during the training process to obtain a trained cardiac center prediction model includes: Based on the cardiac ultrasound image dataset, determine the actual cardiac center coordinates of each cardiac ultrasound image in the cardiac ultrasound image dataset. The cardiac ultrasound image dataset is input into the initial cardiac center prediction model to obtain the initial cardiac center coordinates; the morphological constraint function is used to calculate the morphological deviation between the initial cardiac center coordinates and the actual cardiac center coordinates to obtain the morphological deviation result. When the morphological deviation result is greater than or equal to the preset training threshold, the center correction coefficient is obtained based on the initial cardiac center coordinates. When the morphological deviation result is less than the preset training threshold, the initial heart center coordinate point is output as the target result. The center correction coefficient is fed back into the initial cardiac center prediction model to obtain the trained cardiac center prediction model.

[0015] A second aspect of this invention discloses a 4D intracardiac ultrasound coherent composite imaging device, the device comprising: The data acquisition module is used to acquire an echo dataset; wherein the echo dataset is obtained by transmitting a target waveform to the target object; The data processing module is used to obtain subarray beam signals based on the echo dataset; obtain beamforming lines based on the subarray beam signals; calculate the beamforming lines using a coherent superposition algorithm to obtain superimposed beamforming lines; and obtain a two-dimensional ultrasound image dataset based on the superimposed beamforming lines. An ultrasound imaging module is used to feed the two-dimensional ultrasound image dataset into a pre-trained cardiac center prediction model to obtain the center coordinates of the target object; and to obtain a 4D ultrasound image based on the two-dimensional ultrasound image dataset and the center coordinates of the target object. The trained cardiac center prediction model was obtained through the following method: Obtain a cardiac ultrasound image dataset; use the cardiac ultrasound image dataset as a training dataset to train an initial cardiac center prediction model. During the training process, constrain the training of the initial cardiac center prediction model through a morphological constraint function to obtain a trained cardiac center prediction model.

[0016] As an optional implementation, in the second aspect of the invention, the echo dataset is obtained in the following manner: Determine the target transmission direction; the target transmission direction is used to describe the direction in which the target waveform is transmitted toward the target object; Based on the target's transmission direction, the target receiving point is obtained; Based on the target receiving point, the target transmitted waveform is transmitted several times to obtain an enhanced transmitted beam; wherein the target transmission direction of the target transmitted waveform is the same in the several transmissions; Based on the enhanced transmitted beam, an enhanced echo is obtained; based on the enhanced echo, an echo dataset is obtained.

[0017] As an optional implementation, in a second aspect of the invention, the apparatus further includes: The phased array improvement module is used to divide a preset ultrasonic phased array into several N-order phased arrays. N subarrays; where N represents a preset positive integer; Binding is performed on all the array elements in the subarray to obtain a binding unit; the binding unit is then bound on a preset control unit to obtain a delay register unit; wherein, the delay register unit is used to enable the transmit or receive channels of all the array elements in the subarray to read the delay value and weighting coefficient from the delay register unit; The subarray delay value is obtained based on the geometric center coordinates and target focal length within the subarray; the subarray delay value and a preset weighting coefficient are fed back into the delay register unit to obtain an improved phased array.

[0018] As an optional implementation, in a second aspect of the present invention, the data processing module obtains the specific operation mode of the subarray beam signal based on the echo dataset, including: Obtain the target distance from the target element to the array center in the improved phased array; based on the target distance, obtain the total analog delay corresponding to the target element; Obtain the minimum delay step size in the improved phased array; obtain the total simulation delay series based on the total simulation delay and the minimum delay step size; obtain the total computational delay based on the total simulation delay series and the minimum delay step size. Based on a preset clock frequency, a first analog delay step size is obtained; based on the first analog delay step size and the total computational delay, a first analog delay level is obtained. The first simulation delay is obtained based on the first simulation delay level and the first simulation delay step size; the second simulation delay is obtained based on the first simulation delay and the total computation delay. Based on the first simulated delay and the second simulated delay, the echo dataset is processed using a delay alignment algorithm to obtain an RF signal; the RF signal is then processed using a first simulated weighting algorithm to obtain a subarray beam signal.

[0019] As an optional implementation, in a second aspect of the invention, the data processing module obtains the specific operation mode of the beamforming line based on the subarray beam signal, including: The second analog weighting algorithm is used to process all the subarray beam signals to obtain the beamforming line; The calculation method of the second simulated weighting algorithm is as follows:

[0020]

[0021] In the above formula, The delay applied to the signal of the m-th subarray beam. Let F be the distance from the center of the m-th subarray to the center of the array, and let F be the distance from the center of the m-th subarray to the center of the array. The distance from each array element to the target object, where c is the speed of sound in a specific medium. These are the weighting coefficients at the subarray level. Here, M represents the subarray beam signal, and M is the number of subarrays participating in the global synthesis. This is the total time scale. For beamforming lines.

[0022] As an optional implementation, in a second aspect of the present invention, the data processing module uses a coherent superposition algorithm to calculate the beamforming line to obtain a specific operation method for the superimposed beamforming line, including: Determine several spatial points on the beamforming line; perform the following operation on each spatial point: Obtain the effective aperture array element number H corresponding to the ultrasonic beam emitted in the improved phased array; move the target receiving point along the array direction by aH array element distances to obtain multiple target moving receiving points corresponding to the spatial point; wherein, a represents a specific numerical value, and H represents a preset positive integer; Beamforming is performed on each of the target mobile receiving points to obtain a complex synthesized signal corresponding to each target mobile receiving point; The complex composite signal is calculated using a coherent superposition algorithm to obtain the calculated amplitude of the complex signal; The amplitude is calculated based on the complex signal corresponding to all the spatial points to obtain the superimposed beamforming line.

[0023] As an optional implementation, in the second aspect of the present invention, the specific operation method of constraining the training of the initial cardiac center prediction model through a morphological constraint function during the training process to obtain the trained cardiac center prediction model includes: Based on the cardiac ultrasound image dataset, determine the actual cardiac center coordinates of each cardiac ultrasound image in the cardiac ultrasound image dataset. The cardiac ultrasound image dataset is input into the initial cardiac center prediction model to obtain the initial cardiac center coordinates; the morphological constraint function is used to calculate the morphological deviation between the initial cardiac center coordinates and the actual cardiac center coordinates to obtain the morphological deviation result. When the morphological deviation result is greater than or equal to the preset training threshold, the center correction coefficient is obtained based on the initial cardiac center coordinates. When the morphological deviation result is less than the preset training threshold, the initial heart center coordinate point is output as the target result. The center correction coefficient is fed back into the initial cardiac center prediction model to obtain the trained cardiac center prediction model.

[0024] A third aspect of the present invention discloses an apparatus comprising a memory and a processor, the apparatus comprising: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute some or all of the steps in the 4D intracardiac ultrasound coherent composite imaging method according to any of the first aspects of the present invention.

[0025] The fourth aspect of the present invention discloses a computer storage medium storing computer instructions, which, when invoked, are used to execute some or all of the steps in the 4D intracardiac ultrasound coherent composite imaging method described in any of the first aspects of the present invention.

[0026] Compared with the prior art, the present invention has the following beneficial effects: This invention can acquire the target's transmitted waveform, obtain an echo dataset after transmitting the target's transmitted waveform, obtain subarray beam signals based on the echo dataset, obtain beamforming lines based on the subarray beam signals, calculate the beamforming lines using a coherent superposition algorithm, obtain superimposed beamforming lines, and obtain a two-dimensional ultrasound image dataset based on the superimposed beamforming lines. By acquiring the echo dataset and performing calculations, the clarity of two-dimensional ultrasound imaging is improved. By dividing beamforming into two stages, the beamforming rate and beamforming accuracy can be improved simultaneously. By training a heart center prediction model, the two-dimensional ultrasound image dataset can be input into the trained heart center prediction model to obtain the center coordinates of the target heart, thereby improving the clarity of 4D ultrasound image imaging. Attached Figure Description

[0027] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0028] Figure 1 This is a schematic flowchart of a 4D intracardiac ultrasound coherent composite imaging method disclosed in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a 4D intracardiac ultrasound coherent composite imaging device disclosed in an embodiment of the present invention; Figure 3 This is a schematic diagram of a structure including a memory and a processor device disclosed in an embodiment of the present invention; Figure 4 This is a schematic diagram of transmitting and receiving beamforming disclosed in an embodiment of the present invention. Detailed Implementation

[0029] To enable those skilled in the art to better understand the present invention, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or end that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or ends.

[0031] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0032] This invention provides a 4D intracardiac ultrasound coherent composite imaging method, device, and medium to improve the clarity of 4D ultrasound images. These will be described in detail below.

[0033] Example 1 Please see Figure 1 , Figure 1 This is a schematic flowchart of a 4D intracardiac ultrasound coherent composite imaging method disclosed in an embodiment of the present invention. Figure 1 The described 4D intracardiac ultrasound coherent composite imaging method can be applied to a 4D intracardiac ultrasound coherent composite imaging device, which can be integrated on a cloud server or a local server. This invention does not limit the scope of the invention. Figure 1 As shown, this 4D intracardiac ultrasound coherent composite imaging method may include the following operations: Step 101: Obtain the echo dataset.

[0034] In this embodiment of the invention, the echo dataset can be obtained by transmitting a target waveform to the target object; the echo dataset can be obtained by transmitting and receiving the echo of a sine wave or by transmitting and receiving the echo of a square wave. In this case, the echo dataset itself does not contain spatial focusing information; the target object can be the heart.

[0035] Step 102: Obtain the subarray beam signal based on the echo dataset; obtain the beamforming line based on the subarray beam signal; calculate the beamforming line using the coherent superposition algorithm to obtain the superimposed beamforming line; obtain the two-dimensional ultrasound image dataset based on the superimposed beamforming line.

[0036] In this embodiment of the invention, the subarray beam signal can be a signal obtained by aligning the echo signals in the echo dataset to the same focal point in time, thereby turning the echo dataset into a usable signal.

[0037] In this embodiment of the invention, the beamforming line can be obtained by weighted fusion of all subarray beam signals; that is, local beamforming is performed first to obtain subarray beam signals, and then global beamforming is performed to obtain the beamforming line; for example, the synthesis of subarray beam signals can be performed in the transducer, that is, low-power analog delay and weighted summation are implemented in the ASIC chip, and the signal of each element of the subarray is delayed and weighted to obtain a single signal. Beamforming lines can be generated by sampling subarray beam signals via A / D converter and transmitting them to a processing system. The echo signal of each spatial scattering point is then calculated based on the delay superposition. All beamforming lines are aggregated together to obtain a beamforming line dataset. Finally, a two-dimensional ultrasound image dataset is obtained based on the beamforming line dataset with spatial information.

[0038] Step 103: Feed the two-dimensional ultrasound image dataset into the pre-trained heart center prediction model to obtain the center coordinates of the target object; based on the two-dimensional ultrasound image dataset and the center coordinates of the target object, obtain the 4D ultrasound image.

[0039] In this embodiment of the invention, the center coordinate point of the target object can be the regular center of the heart. A three-dimensional ultrasound image and alignment time information are established through this coordinate point to obtain a 4D ultrasound image. By obtaining the center coordinate point of the heart and the enhanced two-dimensional ultrasound image dataset, and then using traditional 3D heart modeling and alignment time information, the 4D ultrasound image can be made clearer. In this embodiment of the invention, the trained cardiac center prediction model can be obtained by: acquiring a cardiac ultrasound image dataset; using the cardiac ultrasound image dataset as a training dataset to train the initial cardiac center prediction model, and constraining the training of the initial cardiac center prediction model through a morphological constraint function during the training process to obtain the trained cardiac center prediction model. The cardiac center prediction model can be a 3D U-Net model, and the trained cardiac center prediction model is used to analyze the coordinates of the target object's center based on the two-dimensional ultrasound image dataset.

[0040] For example, the measured two-dimensional ultrasound image dataset is input into the heart center prediction model to predict the center coordinates of the heart in the body; then a three-dimensional coordinate system is built with the heart center coordinates as the origin; the multi-angle two-dimensional ultrasound images acquired by the improved phased array are repositioned according to their spatial positions, and the gaps between different two-dimensional ultrasound images are filled by interpolation to form 3D heart body data; finally, the cardiac cycle time is superimposed to become a 4D dynamic heart image.

[0041] As can be seen, the embodiments of the present invention can acquire echo datasets; obtain subarray beam signals based on the echo datasets; obtain beamforming lines based on the subarray beam signals; calculate the beamforming lines using a coherent superposition algorithm to obtain superimposed beamforming lines; obtain a two-dimensional ultrasound image dataset based on the superimposed beamforming lines; improve the clarity of two-dimensional ultrasound imaging by acquiring and calculating the echo datasets; improve the beamforming rate and accuracy by dividing beamforming into two stages; and improve the clarity of 4D ultrasound image imaging by training a cardiac center prediction model, which can feed the two-dimensional ultrasound image datasets back to the trained cardiac center prediction model to obtain the center coordinates of the target heart, thereby improving the clarity of 4D ultrasound image imaging.

[0042] In an optional embodiment, the echo dataset can be obtained in the following way: Determine the target launch direction; Based on the target's launch direction, the target receiving point is obtained; Based on the target receiving point, the target transmitted waveform is transmitted several times to obtain an enhanced transmission beam; The enhanced echo is obtained by enhancing the transmitted beam; the echo dataset is obtained by enhancing the echo.

[0043] In this optional embodiment, the target emission direction can be the direction corresponding to a beamline, or it can be a selected emission angle, or it can be a selected focusing depth (such as the location of a heart valve), or it can be a selected fixed beamline, which is the beam emitted towards the target object.

[0044] In this optional embodiment, the target receiving point is obtained according to the target transmission direction. The target receiving point can be the position on the target object that receives the transmitted beam. Specifically, the process can be to first generate the first sinusoidal transmission pulse, transmit a pure sine wave, and transmit the sine wave without modifying the waveform, adding windows, or amplitude modulation. All array elements / subarrays are in phase and transmitted synchronously. The position where the corresponding sine wave arrives can be the target receiving point. Subsequent transmissions of the same sine wave will all arrive at the target receiving point.

[0045] In this optional embodiment, the target transmission direction of the target transmitted waveform can be the same in several transmissions. Based on the target receiving point, the target transmitted waveform is transmitted several times to obtain an enhanced transmission beam. The enhanced transmission beam can be a beam transmitted to the same point, increasing the energy at that point. Specifically, in the second transmission, the transmitted sound wave is exactly the same as the sine wave transmitted in the first transmission. During the transmission process, the phase must be the same, the peaks must be aligned with the peaks, the troughs with the troughs, the direction must remain unchanged, and the focus must remain unchanged. The subsequent transmissions are the same as the first and second transmissions. When the transmission reaches a set number of times, the beams focused at the target receiving point can be called the enhanced transmission beam. By receiving the strong signal after multiple coherent enhancement, that is, the echo of the enhanced transmission beam, the echo dataset is obtained.

[0046] As can be seen, in this optional embodiment, for a fixed transmission direction and focusing depth, sinusoidal ultrasonic pulse waves of the same frequency, phase and waveform are continuously transmitted multiple times. The propagation direction, focusing position and timing phase of each pulse wave are kept consistent, so that the ultrasonic waves can achieve spatial coherent superposition in the imaging target area. Thus, without changing the single pulse waveform, the effective radiation energy and penetration depth of the transmitted sound waves can be improved.

[0047] In another alternative embodiment, the method may further include: The preset ultrasonic phased array is divided into several N-order phased arrays. N subarrays; Bind the array elements in all subarrays to obtain binding units; bind the binding units on the preset control unit to obtain delay register units; The subarray delay value is obtained based on the geometric center coordinates within the subarray and the target focal length; the subarray delay value and the preset weighting coefficients are fed back into the delay register unit to obtain the improved phased array.

[0048] In this optional embodiment, the ultrasonic phased array can be divided into multiple N×N subarrays, such as 2×2 or 4×4, where each subarray can be used as an independent equivalent array element, and N represents a preset positive integer. The array elements in all subarrays are bound together to obtain a binding unit. This can be done by binding all physical array elements in the subarray together, and the bound units are collectively referred to as a binding unit.

[0049] In this optional embodiment, the control unit can be a set of delay and weighting control registers. Binding the binding unit to the same set of delay and weighting control registers yields a delay register unit. This delay register unit enables all transmit or receive channels of elements within the subarray to read delay values ​​and weighting coefficients from the delay register unit. The delay and weighting registers share a set of transmit delay registers, receive delay registers, and weighting coefficient registers allocated to each subarray. All transmit / receive channels of elements within the subarray read delay values ​​and weighting coefficients from this same set of registers, achieving shared control.

[0050] In this optional embodiment, the subarray delay value is obtained based on the geometric center coordinates within the subarray and the target focal length. This can be achieved by the system calculating the subarray delay value based on imaging parameters; for example, based on the focal length F and the subarray center coordinates in the imaging parameters. Calculate delay value The specific calculation method is as follows:

[0051] Will and the corresponding weighting coefficients By writing to the control register of the subarray, an improved phased array can be obtained. When it is necessary to update the focusing parameters, such as when the focus of dynamic focusing moves, the system only needs to update the delay value at the center of the subarray. All array elements in the subarray will synchronously apply the new delay, eliminating the need to configure each array element individually, thus improving computational efficiency.

[0052] Furthermore, when the subarray size N is greater than a certain value, such as 5, 7, or 9, a preset correction algorithm can be used to fine-tune the delay of the edge array elements to compensate for the phase error caused by the positional differences within the subarray and ensure focusing accuracy. The correction algorithm can be to automatically match data in a pre-calculated delay correction value lookup table stored in the system. This algorithm is specifically used to add a tiny delay compensation to the array elements at the edge of the subarray to smooth out the phase error and make the focus more accurate.

[0053] Optionally, the data calculation for the delay correction numerical reference table can be performed as follows: A training dataset is obtained, and a preset delay correction calculation model is trained. This model simulates the theoretical phase error of each array element under different N and focal points. Real-world measurements are then taken using a real probe in a water phantom / standard phantom to obtain actual data. The error in the actual data is compared with the theoretical phase error, and a correction loss function is used for constraint correction. When the output of the correction loss function meets a preset correction threshold range, the correction coefficient in the correction loss function is fed back into the correction loss function, resulting in the trained delay correction calculation model. During imaging, the delay compensation amount corresponding to each subarray size, focal position, and array element number is calculated using the delay correction calculation model. The calculation results are automatically compiled into a table, resulting in the delay correction numerical reference table. Through the calculation of the correction algorithm and the training of the delay correction calculation model, the phase error caused by positional differences within the subarray can be compensated, ensuring focusing accuracy and improving the clarity of the final image.

[0054] As can be seen, in this optional embodiment, by binding the array elements in all subarrays, a binding unit is obtained, which in turn obtains a delay register unit; based on the geometric center coordinates and target focal length in the subarray, the subarray delay value is obtained, and the calculated subarray delay value and preset weighting coefficients are fed back to the delay register unit; in this way, each subarray is allocated a set of transmit delay registers, receive delay registers, and weighting coefficient registers; the transmit / receive channels of all array elements in the subarray read the delay value and weighting coefficients from the same set of registers, realizing shared control and improving the clarity of imaging.

[0055] In yet another optional embodiment, the subarray beam signal is obtained based on the echo dataset, including: Obtain the target distance from the target element to the array center in the improved phased array; based on the target distance, obtain the total simulation delay corresponding to the target element; Obtain the minimum delay step size in the improved phased array; based on the total simulation delay and the minimum delay step size, obtain the total simulation delay series; based on the total simulation delay series and the minimum delay step size, obtain the total computational delay. Based on the preset clock frequency, the first analog delay step size is obtained; based on the first analog delay step size and the total calculation delay, the first analog delay level is obtained. The first simulation delay is obtained based on the first simulation delay series and the first simulation delay step size; the second simulation delay is obtained based on the first simulation delay and the total computational delay. Based on the first and second analog delays, the echo dataset is processed using a delay alignment algorithm to obtain the radio frequency signal; the radio frequency signal is then processed using a first analog weighting algorithm to obtain the subarray beam signal.

[0056] In this optional embodiment, the total analog delay can be the total time required to allow the echo signals of all array elements to simultaneously align with the same focal point. The calculation method for the total analog delay corresponding to the target array element, based on the target distance, can be as follows:

[0057] in, For the first The distance from each array element to the center of the array For the first The first element corresponds to the first The total analog delay of the path, c is the speed of sound in a specific medium, and F is the first... The distance from each array element to the target object.

[0058] In this optional embodiment, the total number of simulation delay levels can be determined by the minimum delay step size. The final focusing accuracy is not affected by the selection of the first simulation delay step size, avoiding the accuracy loss caused by rounding the first simulation delay in traditional schemes. For example, if the total simulation delay is 23.4 ns and the minimum delay step size of the improved phased array is 1 ns, after rounding down, the total number of simulation delay levels is 23. Optionally, the total calculation delay can be calculated by combining the total number of simulation delay levels with the minimum delay step size, which is 23 ns.

[0059] In this alternative embodiment, the minimum delay step can be an inherent parameter and property of the imaging system or an improved phased array.

[0060] In this optional embodiment, the first analog delay can be a delay compensation applied to the majority of the total computation delay that is an integer number of clock cycles. The delay compensation can involve having the signal of this array element wait for several nanoseconds, temporarily excluding it from the summation, and then superimposing it with the signals of other array elements after the allotted time. The first analog delay does not change the final precision; it only serves to efficiently divide the locked high-precision total delay. The specific calculation method for the first analog delay is as follows:

[0061] in, For the selected first simulation delay level, This is the first simulation delay step. This is the first simulation delay.

[0062] Furthermore, the first analog delay step can be calculated by calculating the reciprocal of the clock frequency corresponding to the imaging system, or by calculating the reciprocal of the preset clock frequency in the improved phased array; for example, if the clock frequency corresponding to the imaging system or the improved phased array is 100MHz, then the first analog delay step can be 1 / 100, which is 10ns; the first analog delay level can be obtained by dividing the total calculated delay by the first analog delay step and rounding down.

[0063] In this optional embodiment, the second simulation delay can be a supplementary delay to the fine simulation delay in the total simulation delay; the calculation of the second simulation delay can be done by subtracting the first simulation delay from the total computational delay, which yields the second simulation delay.

[0064] The calculation process for the first and second analog delays can be illustrated by the following example: Set the total delay of the target =23.4ns, minimum delay step =1ns, then the total number of analog delay stages If the level is such that the total computational delay can be calculated using 23ns, then the computational precision must first be determined; the clock frequency used for the first analog delay calculation... =100MHz, the first analog delay step can be... The large integer complement delay of the first analog delay was determined; the first analog delay series... Level, therefore the first analog delay And the remaining decimal places That is the second simulation delay.

[0065] Furthermore, the second simulation delay step size used for the second simulation delay calculation can also be the minimum delay step size. =1ns calculation, then the second analog delay level in the second analog delay Level, then the second analog delay is .

[0066] In this optional embodiment, the delay alignment algorithm may apply a dedicated delay to the echo signal of each array element, so that the echoes from all array elements originating from the same spatial focal point are aligned to the same moment in time, thus obtaining the radio frequency signal; the delay alignment algorithm calculation method may be:

[0067] in, The enhanced echo data corresponding to the i-th array element; It is the sum of the first and second simulation delays corresponding to the i-th array element, that is, the first simulation delay is calculated first, and then the second simulation delay is calculated. This represents the total time scale; It is a radio frequency signal.

[0068] In this optional embodiment, the first simulated weighting algorithm can be calculated in the following ways:

[0069] In the above formula, The weighted summation of the subarray beam signals, which are the delayed and aligned signals within the subarray, yields the locally focused output of a single subarray. These are weighting coefficients, set according to system parameters; For radio frequency signals; N is the number of participating subarrays.

[0070] As can be seen, in this optional embodiment, by calculating the first and second simulated delays and assigning a dedicated delay to each subarray, the local focusing signals of all subarrays arrive at the global focus at the same time, achieving coherent superposition and improving the global imaging resolution. By calculating the radio frequency signal, the echoes from all array elements originating from the same spatial focus are aligned to the same moment in time. The radio frequency signal is processed using the first simulated weighting algorithm to obtain the subarray beam signal. Due to the small size of the subarray, the delay calculation is simple and has high real-time performance, and the local focusing accuracy is high, which can avoid phase errors under large aperture, thereby improving the clarity of the image.

[0071] In yet another alternative embodiment, obtaining the beamforming line based on the subarray beam signals may include: The second analog weighting algorithm is used to process all subarray beam signals to obtain beamforming lines; The calculation method of the second simulated weighting algorithm is as follows:

[0072]

[0073] In the above formula, The delay applied to the signal of the m-th subarray beam. Let F be the distance from the center of the m-th subarray to the center of the array, and let F be the distance from the center of the m-th subarray to the center of the array. The distance from each array element to the target object, where c is the speed of sound in a specific medium. These are the weighting coefficients at the subarray level. Here, M represents the subarray beam signal, and M is the number of subarrays participating in the global synthesis. This is the total time scale. For beamforming lines.

[0074] In this optional embodiment, such as Figure 4 As shown, the beamforming line can be the result of beamforming the subarray signals.

[0075] As can be seen, in this optional embodiment, beamforming lines are obtained through global beamforming, which retains the high precision of the subarray level while covering a larger imaging range; it reduces the computational pressure of directly performing beamforming on all array elements and improves the computational accuracy of the entire beamforming.

[0076] In another optional embodiment, a coherent superposition algorithm is used to calculate the beamforming line to obtain the superimposed beamforming line, including: Determine several spatial points on the beamforming line; perform the following operation for each spatial point: Obtain the effective aperture array element number H corresponding to the ultrasonic beam emitted in the improved phased array; move the target receiving point along the array direction by aH array element distances to obtain multiple target moving receiving points corresponding to the spatial point; where a represents a specific value and H represents a preset positive integer. Beamforming is performed on each target mobile receiving point to obtain the complex synthesized signal corresponding to each target mobile receiving point; The complex signal is calculated using a coherent superposition algorithm to obtain the calculated amplitude of the complex signal; The amplitude is calculated based on all complex signals corresponding to all spatial points to obtain the superimposed beamforming line.

[0077] In this optional embodiment, the effective aperture element number H can be the number of elements participating in focusing during a single focusing operation; 'a' represents a specific numerical value, which can be 1 / 2, 1 / 3, 1 / 4, etc.; the complex synthesized signal can be a record of the signal's intensity and phase, where the phase is the signal's "time offset / waveform position"; for example, the complex synthesized signal can be written as... in, It is the real part. It is the virtual part. It is the imaginary unit; where I can be the signal strength and Q can be the signal phase / offset.

[0078] In this optional embodiment, the coherent superposition algorithm can be to coherently superimpose multiple complex composite signals in the complex domain, and add the real parts to the real parts and the imaginary parts to the imaginary parts to obtain the calculated amplitude of the complex signal.

[0079] In this optional embodiment, the superimposed beamforming line can be such that the calculated amplitude of each complex signal is the result of the calculation of one spatial point on the beamforming line, and the calculation of the complex signal amplitude of all spatial points is the superimposed beamforming line.

[0080] For example, based on the effective aperture element number H, the focusing point is moved along the array direction in steps of H / 2 element distances to make adjacent focusing areas spatially overlap; for the same beamforming line in space, beamforming is performed using multiple different focal positions to obtain multiple complex composite signals corresponding to the same spatial position; the multiple complex composite signals are coherently superimposed in the complex domain, with the real parts added to the real parts and the imaginary parts added to the imaginary parts; the amplitude is calculated based on the superimposed complex signal as the final imaging data of the beamforming line, that is, the amplitude of the complex signal is calculated for all spatial points on the beamforming line, and finally the superimposed beamforming line is obtained.

[0081] As can be seen, by moving the focus by aH elements each time, the same beamforming line can be calculated multiple times. This coherent superposition reduces random noise and improves imaging resolution and image signal-to-noise ratio.

[0082] In another optional embodiment, during the training process, the initial cardiac center prediction model is constrained by a morphological constraint function to obtain a trained cardiac center prediction model, including: Based on the cardiac ultrasound image dataset, determine the actual coordinates of the heart center for each cardiac ultrasound image in the dataset. The cardiac ultrasound image dataset is input into the initial cardiac center prediction model to obtain the initial cardiac center coordinates; the morphological constraint function is used to calculate the morphological deviation between the initial cardiac center coordinates and the actual cardiac center coordinates to obtain the morphological deviation results. When the morphological deviation result is greater than or equal to the preset training threshold, the center correction coefficient is obtained based on the initial cardiac center coordinates; and the comprehensive correction coefficient is obtained based on all the center correction coefficients. When the morphological deviation result is less than the preset training threshold, the initial cardiac center coordinate point is output as the target result. The comprehensive correction coefficients are fed back into the initial cardiac center prediction model to obtain the trained cardiac center prediction model.

[0083] In this optional embodiment, the cardiac ultrasound image dataset can be a dataset of cardiac ultrasound images with a known cardiac center location; the initial cardiac center prediction model can be obtained by modifying a 3D U-Net model, retaining the encoder while removing the decoder; converting the encoder's 3D features into a one-dimensional vector for easier coordinate calculation; adding a fully connected layer to output three values ​​x, y, and z, which are the 3D coordinates of the cardiac center; the morphological constraint function can be...

[0084] in, For the morphological deviation results, ( , , ( ) represents the actual coordinates of the heart center. , , () represents the coordinates of the center point output by the model. (Judgment) Whether the training threshold is met, the training threshold can be 1, 2, 3, etc.

[0085] When the threshold is determined to be greater than or equal to the training threshold, a center correction coefficient is inserted. The function is modified to obtain:

[0086] Continue to assess the results of correcting morphological deviations. The center coordinates are output only when the center coordinates are less than the training threshold; at the same time, the center correction coefficient is fed back into the model to obtain a trained heart center prediction model.

[0087] As can be seen, in this optional embodiment, an initial heart center prediction model is obtained by improving the 3D Uet model, and the initial heart center prediction model is trained to obtain a trained heart center prediction model. In this process, a morphological constraint function is introduced to constrain the model training, making the results more accurate. By inputting the obtained two-dimensional ultrasound image dataset into the heart center prediction model of this application, accurate heart center coordinates can be obtained. Based on this, 4D heart modeling can improve the clarity of the 4D heart model.

[0088] Example 2 Please see Figure 2 , Figure 2 This is a schematic diagram of the structure of a 4D intracardiac ultrasound coherent composite imaging device disclosed in an embodiment of the present invention. The device may include: Data acquisition module 201 is used to acquire echo dataset; wherein, the echo dataset is obtained by transmitting a target waveform to the target object; The data processing module 202 is used to obtain subarray beam signals based on the echo dataset; obtain beamforming lines based on the subarray beam signals; calculate the beamforming lines using a coherent superposition algorithm to obtain superimposed beamforming lines; and obtain a two-dimensional ultrasound image dataset based on the superimposed beamforming lines. The ultrasound imaging module 203 is used to feed the two-dimensional ultrasound image dataset into the pre-trained cardiac center prediction model to obtain the center coordinates of the target object; and to obtain a 4D ultrasound image based on the two-dimensional ultrasound image dataset and the center coordinates of the target object. The well-trained cardiac center prediction model was obtained through the following methods: Obtain a dataset of cardiac ultrasound images; use the cardiac ultrasound image dataset as a training dataset to train an initial cardiac center prediction model. During the training process, constrain the training of the initial cardiac center prediction model through a morphological constraint function to obtain a trained cardiac center prediction model.

[0089] As can be seen, the embodiments of the present invention can acquire echo datasets; obtain subarray beam signals based on the echo datasets; obtain beamforming lines based on the subarray beam signals; calculate the beamforming lines using a coherent superposition algorithm to obtain superimposed beamforming lines; obtain a two-dimensional ultrasound image dataset based on the superimposed beamforming lines; improve the clarity of two-dimensional ultrasound imaging by acquiring and calculating the echo datasets; improve the beamforming rate and accuracy by dividing beamforming into two stages; and improve the clarity of 4D ultrasound image imaging by training a cardiac center prediction model, which can feed the two-dimensional ultrasound image datasets back to the trained cardiac center prediction model to obtain the center coordinates of the target heart, thereby improving the clarity of 4D ultrasound image imaging.

[0090] In an optional embodiment, the echo dataset is obtained as follows: Determine the target transmission direction; the target transmission direction describes the direction in which the target waveform is transmitted toward the target object. Based on the target's launch direction, the target receiving point is obtained; Based on the target receiving point, the target transmitted waveform is transmitted several times to obtain an enhanced transmission beam; wherein the target transmission direction of the target transmitted waveform in the several transmissions is the same; The enhanced echo is obtained by enhancing the transmitted beam; the echo dataset is obtained by enhancing the echo.

[0091] As can be seen, in this optional embodiment, for a fixed transmission direction and focusing depth, sinusoidal ultrasonic pulse waves of the same frequency, phase and waveform are continuously transmitted multiple times. The propagation direction, focusing position and timing phase of each pulse wave are kept consistent, so that the ultrasonic waves can achieve spatial coherent superposition in the imaging target area. Thus, without changing the single pulse waveform, the effective radiation energy and penetration depth of the transmitted sound waves can be improved.

[0092] In another alternative embodiment, the apparatus may further include: The phased array improvement module is used to divide a preset ultrasonic phased array into several N-order phased arrays. N subarrays; where N represents a preset positive integer; Bind the array elements in all subarrays to obtain binding units; bind the binding units on the preset control unit to obtain delay register units; wherein, the delay register units are used to enable the transmit or receive channels of all array elements in the subarrays to read the delay value and weighting coefficient from the delay register units. The subarray delay value is obtained based on the geometric center coordinates within the subarray and the target focal length; the subarray delay value and the preset weighting coefficients are fed back into the delay register unit to obtain the improved phased array.

[0093] As can be seen, in this optional embodiment, by binding the array elements in all subarrays, a binding unit is obtained, which in turn obtains a delay register unit; based on the geometric center coordinates and target focal length in the subarray, the subarray delay value is obtained, and the calculated subarray delay value and preset weighting coefficients are fed back to the delay register unit; in this way, each subarray is allocated a set of transmit delay registers, receive delay registers, and weighting coefficient registers; the transmit / receive channels of all array elements in the subarray read the delay value and weighting coefficients from the same set of registers, realizing shared control and improving the clarity of imaging.

[0094] In another optional embodiment, the data processing module 202 obtains the specific operation mode of the subarray beam signal based on the echo dataset, which may include: Obtain the target distance from the target element to the array center in the improved phased array; based on the target distance, obtain the total simulation delay corresponding to the target element; Obtain the minimum delay step size in the improved phased array; based on the total simulation delay and the minimum delay step size, obtain the total simulation delay series; based on the total simulation delay series and the minimum delay step size, obtain the total computational delay. Based on the preset clock frequency, the first analog delay step size is obtained; based on the first analog delay step size and the total calculation delay, the first analog delay level is obtained. The first simulation delay is obtained based on the first simulation delay series and the first simulation delay step size; the second simulation delay is obtained based on the first simulation delay and the total computational delay. Based on the first and second analog delays, the echo dataset is processed using a delay alignment algorithm to obtain the radio frequency signal; the radio frequency signal is then processed using a first analog weighting algorithm to obtain the subarray beam signal.

[0095] As can be seen, in this optional embodiment, by calculating the first and second simulated delays and assigning a dedicated delay to each subarray, the local focusing signals of all subarrays arrive at the global focus at the same time, achieving coherent superposition and improving the global imaging resolution. By calculating the radio frequency signal, the echoes from all array elements originating from the same spatial focus are aligned to the same moment in time. The radio frequency signal is processed using the first simulated weighting algorithm to obtain the subarray beam signal. Due to the small size of the subarray, the delay calculation is simple and has high real-time performance, and the local focusing accuracy is high, which can avoid phase errors under large aperture, thereby improving the clarity of the image.

[0096] In another optional embodiment, the data processing module 202 obtains the specific operation mode of the beamforming line based on the subarray beam signal, which may include: The second analog weighting algorithm is used to process all subarray beam signals to obtain beamforming lines; The calculation method of the second simulated weighting algorithm is as follows:

[0097]

[0098] In the above formula, The delay applied to the signal of the m-th subarray beam. Let F be the distance from the center of the m-th subarray to the center of the array, and let F be the distance from the center of the m-th subarray to the center of the array. The distance from each array element to the target object, where c is the speed of sound in a specific medium. These are the weighting coefficients at the subarray level. Here, M represents the subarray beam signal, and M is the number of subarrays participating in the global synthesis. This is the total time scale. For beamforming lines.

[0099] As can be seen, in this optional embodiment, beamforming lines are obtained through global beamforming, which retains the high precision of the subarray level while covering a larger imaging range; it reduces the computational pressure of directly performing beamforming on all array elements and improves the computational accuracy of the entire beamforming.

[0100] In another optional embodiment, the data processing module 202 uses a coherent superposition algorithm to calculate the beamforming line, and the specific operation method for obtaining the superimposed beamforming line may include: Determine several spatial points on the beamforming line; perform the following operation for each spatial point: Obtain the effective aperture array element number H corresponding to the ultrasonic beam emitted in the improved phased array; move the target receiving point along the array direction by aH array element distances to obtain multiple target moving receiving points corresponding to the spatial point; where a represents a specific value and H represents a preset positive integer. Beamforming is performed on each target mobile receiving point to obtain the complex synthesized signal corresponding to each target mobile receiving point; The complex signal is calculated using a coherent superposition algorithm to obtain the calculated amplitude of the complex signal; The amplitude is calculated based on the complex signals corresponding to all spatial points to obtain the superimposed beamforming line.

[0101] As can be seen, by moving the focus by aH elements each time, the same beamforming line can be calculated multiple times. This coherent superposition reduces random noise and improves imaging resolution and image signal-to-noise ratio.

[0102] In another optional embodiment, the ultrasound imaging module 202 constrains the training of the initial cardiac center prediction model through a morphological constraint function during the training process, and obtains the specific operation method of the trained cardiac center prediction model, which may include: Based on the cardiac ultrasound image dataset, determine the actual coordinates of the heart center for each cardiac ultrasound image in the dataset. The cardiac ultrasound image dataset is input into the initial cardiac center prediction model to obtain the initial cardiac center coordinates; the morphological constraint function is used to calculate the morphological deviation between the initial cardiac center coordinates and the actual cardiac center coordinates to obtain the morphological deviation results. When the morphological deviation result is greater than or equal to the preset training threshold, the center correction coefficient is obtained based on the initial cardiac center coordinates. When the morphological deviation result is less than the preset training threshold, the initial cardiac center coordinate point is output as the target result. The center correction coefficient is fed back into the initial cardiac center prediction model to obtain the trained cardiac center prediction model.

[0103] As can be seen, in this optional embodiment, an initial heart center prediction model is obtained by improving the 3D Uet model, and the initial heart center prediction model is trained to obtain a trained heart center prediction model. In this process, a morphological constraint function is introduced to constrain the model training, making the results more accurate. By inputting the obtained two-dimensional ultrasound image dataset into the heart center prediction model of this application, accurate heart center coordinates can be obtained. Based on this, 4D heart modeling can improve the clarity of the 4D heart model.

[0104] Example 3 Please see Figure 3 , Figure 3 This is a schematic diagram of the structure of a device including a memory and a processor, as disclosed in an embodiment of the present invention. Figure 3 As shown, the device including memory and processor may include: Memory 301 storing executable program code; Processor 302 coupled to memory 301; The processor 302 calls the executable program code stored in the memory 301 to execute some or all of the steps in any of the 4D intracardiac ultrasound coherent composite imaging methods in Embodiment 1 of the present invention.

[0105] Example 4 This invention discloses a computer storage medium storing computer instructions. When these computer instructions are invoked, they are used to execute some or all of the steps in any of the 4D intracardiac ultrasound coherent composite imaging methods disclosed in Embodiment 1 of this invention.

[0106] The device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0107] Through the detailed description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically-Erasable Programmable Read-Only Memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0108] Finally, it should be noted that the above embodiments are merely preferred embodiments of the present invention and are only used to illustrate the technical solutions of the present invention, not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A 4D intracardiac ultrasound coherent composite imaging method, characterized in that, The method includes: Determine the target transmission direction; the target transmission direction is used to describe the direction in which the target waveform is transmitted toward the target object; Based on the target transmission direction, the target receiving point is obtained; based on the target receiving point, the target transmission waveform is transmitted several times to obtain several transmission beams; wherein, the target receiving direction corresponding to the target transmission waveforms transmitted several times is the same; Based on all the transmitted beams, the enhanced echo is obtained; based on the enhanced echo, the echo dataset is obtained; Based on the echo dataset, the subarray beam signal is obtained; based on the subarray beam signal, the beamforming line is obtained; several spatial points on the beamforming line are determined; for each spatial point, the following operations are performed: Obtain the effective aperture array element number H corresponding to the emission of ultrasonic beams in the improved phased array; move the target receiving point along the array direction by aH array element distances to obtain multiple target moving receiving points corresponding to the spatial point; wherein, a represents a specific numerical value, and H represents a preset positive integer; Beamforming is performed on each of the target mobile receiving points to obtain a complex synthesized signal corresponding to each target mobile receiving point; The complex composite signal is calculated using a coherent superposition algorithm to obtain the calculated amplitude of the complex signal; The amplitude is calculated based on the complex signal corresponding to all the spatial points to obtain the superimposed beamforming line; a two-dimensional ultrasound image dataset is obtained based on the superimposed beamforming line. The two-dimensional ultrasound image dataset is fed back into a pre-trained cardiac center prediction model to obtain the center coordinates of the target object; a three-dimensional coordinate system is constructed with the center coordinates of the target object as the origin; based on the three-dimensional coordinate system, the two-dimensional ultrasound image dataset is repositioned according to its spatial location to obtain 3D cardiac body data; based on the 3D cardiac body data, a 4D ultrasound image is obtained; the cardiac center prediction model is obtained by modifying and training a 3D U-Net model. The trained cardiac center prediction model was obtained through the following method: Obtain a dataset of cardiac ultrasound images; based on the cardiac ultrasound image dataset, determine the actual cardiac center coordinates corresponding to each cardiac ultrasound image in the dataset; The cardiac ultrasound image dataset is input into the initial cardiac center prediction model to obtain the initial cardiac center coordinates; the morphological constraint function is used to calculate the morphological deviation between the initial cardiac center coordinates and the actual cardiac center coordinates to obtain the morphological deviation result. The specific calculation method for the morphological constraint function is as follows: In the above formula, For the morphological deviation results, ( , , ( ) represents the actual coordinates of the heart center. , , () represents the coordinates of the center point output by the model; When the morphological deviation result is greater than or equal to the preset training threshold, the center correction coefficient is obtained based on the initial heart center coordinates; the center correction coefficient is fed back into the initial heart center prediction model to obtain the trained heart center prediction model.

2. The 4D intracardiac ultrasound coherent composite imaging method according to claim 1, characterized in that, The method further includes: The preset ultrasonic phased array is divided into several N-order phased arrays. N subarrays; where N represents a preset positive integer; Binding is performed on all the array elements in the subarray to obtain a binding unit; the binding unit is then bound on a preset control unit to obtain a delay register unit; wherein, the delay register unit is used to enable the transmit or receive channels of all the array elements in the subarray to read the delay value and weighting coefficient from the delay register unit; The subarray delay value is obtained based on the geometric center coordinates and target focal length within the subarray; the subarray delay value and a preset weighting coefficient are fed back into the delay register unit to obtain an improved phased array.

3. The 4D intracardiac ultrasound coherent composite imaging method according to claim 2, characterized in that, The step of obtaining the subarray beam signal based on the echo dataset includes: Obtain the target distance from the target element to the array center in the improved phased array; based on the target distance, obtain the total analog delay corresponding to the target element; Obtain the minimum delay step size in the improved phased array; obtain the total simulation delay series based on the total simulation delay and the minimum delay step size; obtain the total computational delay based on the total simulation delay series and the minimum delay step size. Based on a preset clock frequency, a first analog delay step size is obtained; based on the first analog delay step size and the total computational delay, a first analog delay level is obtained. The first simulation delay is obtained based on the first simulation delay level and the first simulation delay step size; the second simulation delay is obtained based on the first simulation delay and the total computation delay. Based on the first simulated delay and the second simulated delay, the echo dataset is processed using a delay alignment algorithm to obtain an RF signal; the RF signal is then processed using a first simulated weighting algorithm to obtain a subarray beam signal.

4. The 4D intracardiac ultrasound coherent composite imaging method according to claim 3, characterized in that, The step of obtaining the beamforming line based on the subarray beam signal includes: The second analog weighting algorithm is used to process all the subarray beam signals to obtain the beamforming line; The calculation method of the second simulated weighting algorithm is as follows: In the above formula, The delay applied to the signal of the m-th subarray beam. Let F be the distance from the center of the m-th subarray to the center of the array, and let F be the distance from the center of the m-th subarray to the center of the array. The distance from each array element to the target object, where c is the speed of sound in a specific medium. These are the weighting coefficients at the subarray level. Here, M represents the subarray beam signal, and M is the number of subarrays participating in the global synthesis. This is the total time scale. For beamforming lines.

5. The 4D intracardiac ultrasound coherent composite imaging method according to claim 4, characterized in that, The method further includes: When the morphological deviation result is less than the preset training threshold, the initial cardiac center coordinate point is output as the target result.

6. A 4D intracardiac ultrasound coherent composite imaging device, characterized in that, The device includes: The data acquisition module determines the target transmission direction; the target transmission direction describes the direction in which the target waveform is transmitted toward the target object. Based on the target transmission direction, the target receiving point is obtained; based on the target receiving point, the target transmission waveform is transmitted several times to obtain several transmission beams; wherein, the target receiving direction corresponding to the target transmission waveforms transmitted several times is the same; Based on all the transmitted beams, the enhanced echo is obtained; based on the enhanced echo, the echo dataset is obtained; The data processing module is used to obtain subarray beam signals based on the echo dataset; obtain beamforming lines based on the subarray beam signals; determine several spatial points on the beamforming lines; and perform the following operations on each spatial point: Obtain the effective aperture array element number H corresponding to the emission of ultrasonic beams in the improved phased array; move the target receiving point along the array direction by aH array element distances to obtain multiple target moving receiving points corresponding to the spatial point; wherein, a represents a specific numerical value, and H represents a preset positive integer; Beamforming is performed on each of the target mobile receiving points to obtain a complex synthesized signal corresponding to each target mobile receiving point; The complex composite signal is calculated using a coherent superposition algorithm to obtain the calculated amplitude of the complex signal; The amplitude is calculated based on the complex signal corresponding to all the spatial points to obtain the superimposed beamforming line; a two-dimensional ultrasound image dataset is obtained based on the superimposed beamforming line. An ultrasound imaging module is used to feed the two-dimensional ultrasound image dataset into a pre-trained cardiac center prediction model to obtain the center coordinates of the target object; a three-dimensional coordinate system is constructed with the center coordinates of the target object as the origin; the two-dimensional ultrasound image dataset is repositioned according to its spatial location based on the three-dimensional coordinate system to obtain 3D cardiac body data; and a 4D ultrasound image is obtained based on the 3D cardiac body data; the cardiac center prediction model is trained by modifying the 3D U-Net model. The trained cardiac center prediction model was obtained through the following method: Obtain a dataset of cardiac ultrasound images; based on the cardiac ultrasound image dataset, determine the actual cardiac center coordinates corresponding to each cardiac ultrasound image in the dataset; The cardiac ultrasound image dataset is input into the initial cardiac center prediction model to obtain the initial cardiac center coordinates; the morphological constraint function is used to calculate the morphological deviation between the initial cardiac center coordinates and the actual cardiac center coordinates to obtain the morphological deviation result. The specific calculation method for the morphological constraint function is as follows: In the above formula, For the morphological deviation results, ( , , ( ) represents the actual coordinates of the heart center. , , () represents the coordinates of the center point output by the model; When the morphological deviation result is greater than or equal to the preset training threshold, the center correction coefficient is obtained based on the initial heart center coordinates; the center correction coefficient is fed back into the initial heart center prediction model to obtain the trained heart center prediction model.

7. An apparatus comprising a memory and a processor, characterized in that, The device includes: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the 4D intracardiac ultrasound coherent composite imaging method as described in any one of claims 1-5.

8. A computer storage medium, characterized in that, The computer storage medium stores computer instructions, which, when invoked by a processor, are used to execute the 4D intracardiac ultrasound coherent composite imaging method as described in any one of claims 1-5.

Images