Real-time imaging method and system for three-dimensional ultrasound imaging catheter

Abstract

The application discloses a real-time imaging method and system for a three-dimensional ultrasonic imaging catheter, relates to the technical field of three-dimensional imaging, and comprises the following steps: acquiring preset imaging targets and tissue type data, and establishing a three-dimensional imaging target model; performing high-frame-frequency electronic scanning on a target region, performing imaging quality analysis on returned scanning data, and acquiring a real-time imaging quality parameter set; performing deviation calculation to obtain a quality parameter deviation set; performing probe control and scanning path optimization analysis based on the quality parameter deviation set, and outputting an optimal probe control scheme and an optimal scanning path for subsequent three-dimensional imaging operations. The application solves the technical problem that three-dimensional imaging equipment in the prior art lacks effective optimization means for probe control and scanning paths, which leads to poor imaging quality and low scanning efficiency, achieves the optimization of probe control and scanning paths of the three-dimensional imaging equipment, and improves the imaging quality and scanning efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of three-dimensional imaging, in particular to a real-time imaging method and system for a three-dimensional ultrasonic imaging catheter. BACKGROUND

[0002] Traditional three-dimensional imaging technology faces many challenges. In terms of probe control, existing methods are difficult to adjust in real time according to different imaging targets and tissue types, resulting in deviations in the data collected by the probe. At the same time, the scanning path planning lacks flexibility and pertinence, often using fixed modes and failing to fully consider the complex characteristics of the target region. This seriously affects the imaging quality, and the image may appear blurred, details lost, etc., which not only increases the difficulty of diagnosis, but also may lead to misdiagnosis. In addition, the inefficient scanning process consumes a lot of time, reduces the examination efficiency, and is extremely unfavorable for time-sensitive scenes such as emergency.

[0003] The prior art has the technical problem of three-dimensional imaging equipment lacking effective optimization means for probe control and scanning path, resulting in poor imaging quality and low scanning efficiency. SUMMARY

[0004] The present application provides a real-time imaging method and system for a three-dimensional ultrasonic imaging catheter, which is used to solve the technical problem of the prior art that three-dimensional imaging equipment lacks effective optimization means for probe control and scanning path, resulting in poor imaging quality and low scanning efficiency.

[0005] In view of the above problems, the present application provides a real-time imaging method and system for a three-dimensional ultrasonic imaging catheter.

[0006] In a first aspect of the present application, a real-time imaging method for a three-dimensional ultrasonic imaging catheter is provided, the method comprising:

[0007] acquiring pre-set imaging target and tissue type data, and establishing a three-dimensional imaging target model, wherein the three-dimensional imaging target model is embedded with a standard imaging quality parameter set; performing high-frame-rate electronic scanning on the target region according to an initial probe control scheme and an initial scanning path through a predetermined three-dimensional imaging device, and performing imaging quality analysis on the returned scanning data to obtain a real-time imaging quality parameter set; performing deviation calculation according to the standard imaging quality parameter set and the real-time imaging quality parameter set to obtain a quality parameter deviation set; performing probe control and scanning path optimization analysis based on the quality parameter deviation set, and outputting an optimal probe control scheme and an optimal scanning path for subsequent three-dimensional imaging operation.

[0008] In a second aspect of the present application, a real-time imaging system for a three-dimensional ultrasonic imaging catheter is provided, the system comprising:

[0009] The three-dimensional imaging target model establishing module is configured to acquire preset imaging target and tissue type data, and establish a three-dimensional imaging target model, wherein the three-dimensional imaging target model is embedded with a standard imaging quality parameter set; the real-time imaging quality parameter set acquiring module is configured to perform high-frame-frequency electronic scanning on a target region according to an initial probe control scheme and an initial scanning path by using a predetermined three-dimensional imaging device, perform imaging quality analysis on the returned scanning data, and acquire a real-time imaging quality parameter set; the quality parameter deviation set acquiring module is configured to perform deviation calculation based on the standard imaging quality parameter set and the real-time imaging quality parameter set, and obtain a quality parameter deviation set; and the optimal probe control scheme output unit is configured to perform probe control and scanning path optimization analysis based on the quality parameter deviation set, and output an optimal probe control scheme and an optimal scanning path for subsequent three-dimensional imaging operation.

[0010] The one or more technical solutions provided in the present application have at least the following technical effects or advantages:

[0011] The preset imaging target and tissue type data are acquired, and a three-dimensional imaging target model is established; high-frame-frequency electronic scanning is performed on a target region according to an initial probe control scheme and an initial scanning path by using a predetermined three-dimensional imaging device, imaging quality analysis is performed on the returned scanning data, and a real-time imaging quality parameter set is acquired; deviation calculation is performed to obtain a quality parameter deviation set; and probe control and scanning path optimization analysis are performed based on the quality parameter deviation set, and an optimal probe control scheme and an optimal scanning path are output for subsequent three-dimensional imaging operation. The technical effect of realizing optimization of probe control and scanning path of a three-dimensional imaging device is achieved, and the imaging quality and scanning efficiency are improved. BRIEF DESCRIPTION OF DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed in the embodiment description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative effort.

[0013] Figure 1 A real-time imaging method flowchart for a three-dimensional ultrasonic imaging catheter provided by the embodiment of the present application;

[0014] Figure 2 A real-time imaging system structure diagram for a three-dimensional ultrasonic imaging catheter provided by the embodiment of the present application.

[0015] Legend of the drawings: three-dimensional imaging target model establishing module 10, real-time imaging quality parameter set acquiring module 20, quality parameter deviation set acquiring module 30, and optimal probe control scheme output module 40. DETAILED DESCRIPTION

[0016] The present application provides a real-time imaging method and system for a three-dimensional ultrasound imaging catheter to solve the technical problem that the probe control and scan path of the three-dimensional imaging device in the prior art lack effective optimization means, resulting in poor imaging quality and low scanning efficiency.

[0017] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative labor fall within the scope of protection of the present application.

[0018] Embodiment one, as shown in the present application provides a real-time imaging method for a three-dimensional ultrasound imaging catheter, which comprises: Figure 1

[0019] Step S100: acquiring pre-set imaging target and tissue type data, and establishing a three-dimensional imaging target model, wherein the three-dimensional imaging target model is embedded with a standard imaging quality parameter set.

[0020] Specifically, the acquisition and processing of imaging target and tissue type data are crucial. Imaging technology is applied in the medical field, such as heart examination, etc. The operator first needs to determine a specific imaging target, for example, to image different parts of the heart (atrium, ventricle, etc.), while clearly defining the relevant tissue type data, including the acoustic properties of myocardium, valve tissue, etc. These data are the basis for establishing a three-dimensional imaging target model. The model construction process simulates the three-dimensional structure of the target area with the help of professional software and algorithms (finite element analysis software and U-Net algorithm). Moreover, the model is embedded with a standard imaging quality parameter set, which is set according to medical imaging standards and clinical experience, covering key indicators such as resolution, contrast, signal-to-noise ratio, etc. For example, the 4D ICE technology pursues real-time generation of high-quality 3D ultrasound images. These standard parameters set an ideal benchmark for imaging quality, which is used for subsequent comparison and optimization of actual imaging effect, to ensure that the final presented ultrasound images can meet the needs of clinical diagnosis and treatment, and help doctors accurately observe the morphology and structure of tissues and organs.

[0021] Step S200: performing high-frame-rate electronic scanning on the target area according to the initial probe control scheme and the initial scan path through the predetermined three-dimensional imaging device, and performing imaging quality analysis on the returned scan data to obtain a real-time imaging quality parameter set.

[0022] ​Specifically, the predetermined three-dimensional imaging device is built based on a predetermined two-dimensional probe array and an ASIC architecture. A plurality of ultrasonic probe elements in the two-dimensional probe array are arranged in a predetermined geometric structure. At the beginning of imaging, the device accurately regulates the signal emission parameters of each ultrasonic probe element, such as signal time delay, amplitude, excitation sequence, beam width, and frequency, according to an initial probe control scheme, and simultaneously performs electronic scanning on the target region at a high frame rate according to an initial scanning path. In the scanning process, the ultrasonic probe elements emit ultrasonic waves to the target region, and produce echo signals when encountering different tissue interfaces. These echo signals are received by the probe and returned, forming returned scanning data. The returned scanning data is processed using an image quality analysis algorithm based on deep learning to obtain a set of real-time imaging quality parameters. Taking an algorithm based on a convolutional neural network (CNN) as an example, first, the returned scanning data is preprocessed to adjust it to a format suitable for CNN input, such as adjusting the data dimensions, normalization processing, and other operations, so that the data features are more easily learned by the network. Next, the preprocessed returned scanning data is input into a pre-trained CNN model, which includes multiple convolutional layers, pooling layers, and fully connected layers. The convolutional layers extract local features in the data, such as the boundaries and textures of different tissues, by sliding convolution kernels over the data. The pooling layers downsample the feature maps output by the convolutional layers, reducing the data volume while preserving the main features, improving the computational efficiency and generalization ability of the model. The fully connected layers integrate the features processed by multiple convolution and pooling layers and output the final prediction results. In this process, the model is trained on a large amount of ultrasonic image data with labeled imaging quality parameters to learn the complex mapping relationship between the returned scanning data and the imaging quality parameters. After training, the model can output a corresponding set of real-time imaging quality parameters, including resolution, contrast, signal-to-noise ratio, and other key parameters, providing data basis for subsequent imaging quality optimization, when the current returned scanning data is input.

[0023] Step S300: Calculate the quality parameter deviation set according to the deviation between the standard imaging quality parameter set and the real-time imaging quality parameter set.

[0024] Specifically, first, the standard imaging quality parameter set embedded in the three-dimensional imaging target model and the real-time imaging quality parameter set obtained after analyzing the returned scanning data are extracted respectively. Both of the two parameter sets cover key imaging quality indicators such as resolution, contrast, and signal-to-noise ratio. Then, the deviation of each quality indicator is calculated. Taking resolution as an example, the resolution standard value in the standard imaging quality parameter set is subtracted from the actual resolution value in the real-time imaging quality parameter set to obtain the resolution deviation. Similarly, the same method is used to calculate the deviations of contrast and signal-to-noise ratio, that is, the standard value is subtracted from the real-time value. The deviations calculated for different quality indicators are summarized to form the quality parameter deviation set. This deviation set can clearly reflect the gap between the current real-time imaging effect and the ideal standard in each quality dimension, providing a clear direction and quantitative basis for subsequent optimization of probe control and scanning path based on these deviations.

[0025] Step S400: Based on the quality parameter deviation set, the probe control and scanning path optimization analysis is carried out, and the optimal probe control scheme and optimal scanning path are output for subsequent three-dimensional imaging operation.

[0026] Specifically, based on the quality parameter deviation set, comprehensive and key optimization analysis is carried out to improve imaging quality, and the probe control and scanning path are optimized respectively. In terms of probe control optimization, the resolution deviation, contrast deviation, and signal-to-noise ratio deviation in the quality parameter deviation set are used as the guide, and the probe control parameter threshold value is used as the optimization space. For example, the probe control parameters include signal delay, amplitude, excitation sequence, beam width, and frequency, etc. An initial probe control scheme is randomly selected within the probe control parameter threshold value range, and the imaging quality compensation plug-in based on machine learning is used to predict the imaging quality compensation parameters such as resolution compensation data, contrast compensation data, and signal-to-noise ratio compensation data under this scheme. Then, the resolution deviation, contrast deviation, and signal-to-noise ratio deviation are used to construct the scheme fitness evaluation function (wherein, is the scheme fitness, , , resolution weight, contrast weight, and signal-to-noise ratio weight, respectively, K is the resolution deviation, k is the resolution compensation data, P is the contrast deviation, p is the contrast compensation data, Y is the signal-to-noise ratio deviation, and y is the signal-to-noise ratio compensation data.), the imaging quality compensation parameters are evaluated for fitness, and the scheme fitness is output. By continuously randomly selecting probe control schemes within the probe control parameter threshold value range for iterative compensation prediction and evaluation, the optimal probe control scheme corresponding to the maximum scheme fitness is determined as the optimal probe control scheme until a predetermined number of iterations is reached.

[0027] In the aspect of scan path optimization, a first quality parameter deviation set of an initial scan path is obtained, wherein the initial scan path includes an initial scan angle. A first comprehensive quality coefficient is determined according to the first quality parameter deviation set. If the coefficient is greater than or equal to a predetermined quality threshold, the initial scan angle is set as a secondary scan angle. If the coefficient is less than the predetermined quality threshold, any angle within a scan angle threshold except the initial scan angle is randomly selected as the secondary scan angle. In addition, a historical scan angle sequence and a historical comprehensive quality coefficient sequence are recorded. The historical scan angle sequence is clustered according to a predetermined angle interval to determine a plurality of historical scan angle intervals. The comprehensive quality coefficient mean of each interval is calculated according to the historical comprehensive quality coefficient sequence. If the comprehensive quality coefficient mean greater than or equal to the predetermined quality threshold is not 0, the historical scan angle interval corresponding to the maximum comprehensive quality coefficient mean is set as an optimal historical scan angle interval. Any angle within the interval is randomly selected as an optimal scan path. If the comprehensive quality coefficient mean is 0, an angle within a scan angle threshold except the plurality of historical scan angle intervals is randomly selected as the optimal scan path. Through the above optimization analysis of the probe control and the scan path, an optimal probe control scheme and an optimal scan path are finally outputted and applied to subsequent three-dimensional imaging operations, thereby improving the imaging quality.

[0028] In one possible implementation manner, step S200 further includes:

[0029] Step S210: The predetermined three-dimensional imaging device is built based on a predetermined two-dimensional probe array and an ASIC architecture. The predetermined two-dimensional probe array includes a plurality of ultrasonic probe elements, wherein the plurality of ultrasonic probe elements are arranged according to a predetermined geometric structure, and the predetermined geometric structure includes a predetermined shape, a predetermined size, and a predetermined arrangement manner.

[0030] Specifically, in the construction of a predetermined three-dimensional imaging device for three-dimensional ultrasound imaging, its core architecture is built based on a predetermined two-dimensional probe array and an ASIC (Application Specific Integrated Circuit). The predetermined two-dimensional probe array is a key component of the device, which is composed of a large number of ultrasonic probe elements. These ultrasonic probe elements are not randomly distributed, but are strictly arranged according to a predetermined geometry, which covers multiple aspects such as predetermined shape, predetermined size and predetermined arrangement. In terms of predetermined shape, the two-dimensional probe array can be designed into a rectangle, a circle or other specific shapes. Different shapes will affect the transmission and reception range of ultrasonic signals. For example, a rectangular probe array is suitable for large-area scanning of a specific area because it has regular signal coverage in certain directions; a circular probe array may have certain advantages in omnidirectional signal acquisition. The predetermined size determines the physical size of each ultrasonic probe element, which directly relates to its ability to transmit and receive ultrasonic signals. Larger probe elements may have an advantage in signal strength, while smaller probe elements may perform better in resolution. In terms of predetermined arrangement, ultrasonic probe elements are arranged in equal-interval straight lines or matrix form. Equal-interval straight-line arrangement is conducive to intensive scanning in a certain direction to obtain detailed information; matrix arrangement can achieve more flexible scanning modes and improve the imaging capability of complex-shaped target areas. Through such a carefully designed predetermined two-dimensional probe array, combined with the powerful signal processing capability of the ASIC architecture, the predetermined three-dimensional imaging device can operate efficiently and lay a solid foundation for subsequent three-dimensional ultrasound imaging work.

[0031] In one possible implementation, step S200 further includes:

[0032] Step S220: acquiring back scan data, wherein the back scan data includes a plurality of echo signals back from a plurality of ultrasonic probe elements.

[0033] Step S230: performing denoising processing and time domain alignment on the plurality of echo signals to obtain standard back scan data.

[0034] Step S240: performing imaging quality analysis on the standard back scan data according to a predetermined evaluation index to obtain a real-time imaging quality parameter set, wherein the imaging quality parameters at least include resolution, contrast and signal-to-noise ratio.

[0035] Specifically, the process of back data processing begins. After completing the scanning of the target area, the ultrasonic probe elements in the two-dimensional probe array of the predetermined three-dimensional imaging device will generate back scan data, which contains a plurality of echo signals back from a plurality of ultrasonic probe elements. Each echo signal carries acoustic information of different positions in the target area, and they are the basis of raw data for subsequent imaging analysis.

[0036] The echo signals are preprocessed to improve data quality. De-noising is one of the key steps. Various noises, such as electronic noise and environmental noise, are inevitably mixed in the signal transmission process, which will interfere with the extraction of the real information of the target area. By using appropriate filtering algorithms, such as Gaussian filtering and wavelet filtering, the noise components in the echo signals are removed, and the useful signal characteristics are retained. At the same time, time domain alignment operation is performed, because there may be slight differences in the time of receiving echo signals by different ultrasonic probe elements, which will affect the accurate judgment of the structure of the target area. By calibrating and adjusting the time information of the echo signals, each echo signal has consistency in the time dimension, and finally the standard back transmission scanning data is obtained, providing reliable data support for accurate imaging quality analysis.

[0037] After obtaining the standard back transmission scanning data, a comprehensive imaging quality analysis is performed according to the pre-set evaluation indicators to obtain a real-time imaging quality parameter set. These evaluation indicators revolve around the key aspects of imaging quality, among which resolution, contrast and signal-to-noise ratio are important measurement parameters. For resolution, the smallest details that can be distinguished in the standard back transmission scanning data are analyzed to evaluate, such as whether the fine structure of the myocardium can be clearly distinguished when imaging the heart; contrast focuses on the gray difference between different tissues, which is determined by calculating the gray variation of different regions. If the gray difference between different tissues is obvious, the contrast is high, and the imaging effect is more conducive to distinguishing different tissues; signal-to-noise ratio is used to measure the ratio of signal intensity to noise intensity, which is obtained by calculating the ratio of signal power to noise power. A higher signal-to-noise ratio means that the signal is clearer and less disturbed by noise. Through comprehensive analysis and calculation of these parameters, a real-time imaging quality parameter set that can accurately reflect the actual quality of the current imaging is obtained, providing a quantitative basis for subsequent evaluation of imaging effect and optimization of imaging process.

[0038] In one possible implementation manner, step S400 further includes:

[0039] Step S410: obtaining the quality parameter deviation set, wherein the quality parameter deviation set includes resolution deviation, contrast deviation and signal-to-noise ratio deviation.

[0040] Step S420: performing probe control optimization analysis with the purpose of satisfying the resolution deviation, contrast deviation and signal-to-noise ratio deviation, and taking the probe control parameter threshold as the optimization space, and outputting an optimal probe control scheme, wherein the probe control scheme includes a plurality of probe control parameters of the probe, and the probe control parameters include signal time delay, amplitude, excitation sequence, beam width and frequency.

[0041] Specifically, first, a set of quality parameter deviations is extracted, which is the key basis for judging the gap between the current imaging quality and the ideal standard. Among them, the resolution deviation reflects the difference between the actual imaging resolution ability of fine structures and the standard requirement. For example, when imaging human organs, if the actual imaging cannot clearly present the branching of the micro blood vessels in the organs, while the standard imaging can clearly distinguish, the difference between them is the resolution deviation; the contrast deviation reflects the degree of deviation of the gray difference of different tissues in the image from the ideal state. Under normal circumstances, different tissues should have obvious gray difference in the ultrasound image to facilitate observation. If the gray of the tissues in the actual imaging is similar and difficult to distinguish, the contrast deviation is generated; the signal-to-noise ratio deviation measures the proportion of signal intensity and noise intensity deviating from the standard. If too much noise is mixed in the imaging process, the signal clarity is reduced, and the proportion of signal and noise is unbalanced, the signal-to-noise ratio deviation will appear. By obtaining a series of deviation data, the deficiencies of the current imaging quality in various key indicators can be comprehensively understood, and accurate direction guidance is provided for subsequent targeted optimization and adjustment, so as to realize the improvement of the imaging quality.

[0042] With the goal of meeting the resolution deviation, the contrast deviation and the signal-to-noise ratio deviation, the probe control parameter threshold is taken as the optimization space. The probe control scheme covers multiple probe control parameters of the probe, including signal time delay, amplitude, excitation sequence, beam width and frequency, which have a direct impact on the quality of ultrasound imaging. For example, the signal time delay determines the time difference between the transmission and reception of the ultrasound signal. Reasonable adjustment of the time delay can optimize the signal focusing effect and thus improve the resolution; the size of the amplitude affects the intensity of the ultrasound signal. Properly changing the amplitude can enhance the distinction between the signal and the noise and improve the signal-to-noise ratio; the adjustment of the excitation sequence can change the working mode of the probe array, affect the formation of the beam and the scanning range, and affect the contrast; the beam width determines the coverage area of the ultrasound beam. Suitable beam width helps to improve the resolution and contrast; the selection of the frequency is closely related to the penetration depth and resolution of the imaging. Higher frequency is suitable for high-resolution imaging of shallow structures, and lower frequency can penetrate deeper tissues but has relatively lower resolution. In the optimization process, different parameter combinations are constantly tried within the range defined by the probe control parameter threshold. Through simulation calculation or actual test, the improvement effect of each combination on the imaging quality deviation is evaluated, and finally the optimal probe control scheme that can minimize these deviations is output, providing strong support for improving the quality of three-dimensional ultrasound imaging.

[0043] In one possible implementation manner, step S420 further includes:

[0044] Step S421: randomly selecting a first probe control scheme within the probe control parameter threshold, performing imaging quality compensation prediction on the first probe control scheme by using an imaging quality compensation plug-in, and outputting first imaging quality compensation parameters, wherein the imaging quality compensation plug-in is constructed based on machine learning, and the imaging quality compensation parameters include resolution compensation data, contrast compensation data, and signal-to-noise ratio compensation data.

[0045] Step S422: constructing a scheme fitness evaluation function based on the resolution deviation, contrast deviation, and signal-to-noise ratio deviation, performing fitness evaluation on the first imaging quality compensation parameters, and outputting a first scheme fitness.

[0046] Step S423: continuing to randomly select probe control schemes within the probe control parameter threshold for iterative compensation prediction and evaluation until a predetermined number of iterations is reached, and outputting a probe control scheme with the maximum scheme fitness as the optimal probe control scheme.

[0047] Step S424: the expression of the scheme fitness evaluation function is:

[0048] ; wherein, is the scheme fitness, , , the resolution weight, contrast weight, and signal-to-noise ratio weight, respectively, K is the resolution deviation, k is the resolution compensation data, P is the contrast deviation, p is the contrast compensation data, Y is the signal-to-noise ratio deviation, and y is the signal-to-noise ratio compensation data.

[0049] Specifically, within the probe control parameter threshold, signal time delay, amplitude, excitation order, beam width, and frequency parameters are randomly combined to generate a first probe control scheme. The imaging quality compensation plug-in is constructed based on a neural network, which uses a large amount of historical data in the training stage. These data contain different probe control parameter combinations and corresponding imaging quality index changes. During training, the probe control parameters are used as the node data of the input layer, and the resolution compensation data, contrast compensation data, and signal-to-noise ratio compensation data are used as the target values of the output layer. Through the back propagation algorithm, the weights and biases between neurons in the network are constantly adjusted, so that the network can learn the complex mapping relationship between the probe control parameters and the imaging quality compensation. After obtaining the first probe control scheme, it is input into the trained neural network. The network performs calculation and reasoning according to the learned pattern, predicts the imaging quality compensation under this scheme, and finally outputs the first imaging quality compensation parameters including resolution compensation data, contrast compensation data, and signal-to-noise ratio compensation data, providing a quantitative basis for evaluating the improvement effect of the scheme on imaging quality.

[0050] After obtaining the resolution deviation, contrast deviation, and signal-to-noise ratio deviation, and the first imaging quality compensation parameter (including resolution compensation data, contrast compensation data, and signal-to-noise ratio compensation data), the scheme fitness evaluation function is constructed based on these data. The function comprehensively considers the importance of different imaging quality indicators, and reflects the proportion of each indicator in the overall evaluation through the resolution weight, contrast weight, and signal-to-noise ratio weight. The function expression clearly shows the relationship between the parameters. The first imaging quality compensation parameter is substituted into the evaluation function for calculation, and the first scheme fitness is finally output. This value can intuitively and quantitatively reflect the comprehensive effect of the first probe control scheme in improving the imaging quality deviation. The higher the value, the more effective the scheme is in narrowing the gap between the current imaging quality and the ideal standard, providing an important reference for subsequent selection of the optimal probe control scheme.

[0051] After the imaging quality compensation prediction and fitness evaluation of the first probe control scheme have been completed, a better scheme is started to be explored. Within the threshold range of the probe control parameter, new probe control schemes are randomly selected. For each new scheme, the imaging quality compensation prediction based on machine learning is used to obtain the corresponding imaging quality compensation parameter. Then, the scheme fitness evaluation function based on the resolution deviation, contrast deviation, and signal-to-noise ratio deviation is used to evaluate the fitness of the new imaging quality compensation parameter, and the fitness of the scheme is output. This series of operations is repeated continuously, and each iteration may produce a new scheme fitness value. The iterative compensation prediction and evaluation are continuously performed until the pre-set number of iterations is reached. During the entire iteration process, the scheme fitness produced by each iteration is recorded, and the maximum value among all recorded fitness values is selected. Finally, the probe control scheme corresponding to the maximum scheme fitness is determined as the optimal probe control scheme. This optimal scheme can theoretically minimize the imaging quality deviation and provide the most ideal probe control parameter setting for subsequent three-dimensional imaging, thereby effectively improving the quality of ultrasonic imaging.

[0052] The function expression Comprehensively considers the key indicators of imaging quality and provides a quantitative basis for evaluating the pros and cons of the probe control scheme. is an important indicator for measuring the comprehensive effect of a probe control scheme in improving imaging quality. In the formula, , , respectively represent resolution weight, contrast weight and signal-to-noise ratio weight, which reflect the relative importance of different imaging quality indicators in overall evaluation. In practical applications, these weight values can be flexibly adjusted according to specific imaging needs and scenarios. For example, in the case of high requirements for imaging of fine structures, the resolution weight can be appropriately increased to highlight the role of the resolution indicator in evaluation.

[0053] K, P, Y respectively represent resolution deviation, contrast deviation and signal-to-noise ratio deviation, which reflect the gap between current imaging quality and ideal imaging quality in each indicator. k, p, y are the corresponding resolution compensation data, contrast compensation data and signal-to-noise ratio compensation data, representing the improvement degree that a certain probe control scheme may bring in the corresponding indicator. By calculating the reciprocal of the absolute value of the difference between each indicator deviation and compensation data, multiplying by the corresponding weight, and adding the three items, the scheme fitness can comprehensively reflect the ability of the probe control scheme to narrow the imaging quality deviation. The larger the scheme fitness value, the more effectively the scheme can reduce the imaging quality deviation, which means that the scheme performs better in improving the imaging quality. In summary, this scheme fitness evaluation function provides a scientific and accurate evaluation standard for screening the optimal probe control scheme by comprehensively considering the resolution, contrast and signal-to-noise ratio three key imaging quality indicators, which helps to improve the quality and effect of three-dimensional ultrasound imaging.

[0054] In one possible implementation manner, step S400 further includes:

[0055] Step S430: obtaining a first set of quality parameter deviations of an initial scanning path, wherein the initial scanning path includes an initial scanning angle.

[0056] Step S440: determining a first comprehensive quality coefficient according to the first set of quality parameter deviations, and setting the initial scanning angle as a secondary scanning angle if the first comprehensive quality coefficient is greater than or equal to a predetermined quality threshold.

[0057] Step S450: setting any angle within a scanning angle threshold except the initial scanning angle as a secondary scanning angle if the first comprehensive quality coefficient is less than the predetermined quality threshold, and setting the secondary scanning angle as an optimal scanning path.

[0058] ​Specifically, first focus on the initial scan path, which contains the initial scan angle. Based on the previously obtained standard imaging quality parameter set and real-time imaging quality parameter set, a deep analysis is conducted on the imaging data collected under the initial scan path. By carefully comparing the key quality parameters such as resolution, contrast, signal-to-noise ratio, etc. in the imaging data with the standard values, the deviation values of each parameter are calculated, and these deviation values are aggregated to form the first quality parameter deviation set. For example, when calculating the resolution deviation, the smallest resolvable detail specified in the standard imaging quality parameter set is compared with the smallest detail that can be actually imaged under the current initial scan path, and the resolution deviation is obtained. The first quality parameter deviation set comprehensively reflects the gap between the current imaging quality and the ideal imaging quality in various key quality dimensions under the initial scan path, providing key data support for subsequent evaluation of the imaging effect of the initial scan path, judgment of whether the scan angle needs to be adjusted, and how to adjust.

[0059] The resolution deviation, contrast deviation, and signal-to-noise ratio deviation in the first quality parameter deviation set are used as input data to construct a neural network model. The neural network can include an input layer, several hidden layers, and an output layer, and the neurons in the hidden layers are connected through weights and biases. In the training phase, a large number of existing quality parameter deviation sets and corresponding known comprehensive quality coefficient data are used as training samples, and the weights and biases in the network are continuously adjusted through the backpropagation algorithm, so that the network learns the mapping relationship between the quality parameter deviation and the comprehensive quality coefficient. When the model training is completed, the current first quality parameter deviation set is input into the trained neural network, and the network will output the predicted first comprehensive quality coefficient. Then, this coefficient is compared with the predetermined quality threshold value, if greater than or equal to the predetermined quality threshold value, it indicates that the imaging quality under the current initial scan path meets the requirements, and the initial scan angle is set as the secondary scan angle, so that the subsequent imaging work can continue at this angle, thereby ensuring the stability and reliability of the imaging quality.

[0060] If the first comprehensive quality coefficient obtained in the previous step is less than the predetermined quality threshold, it means that the imaging quality under the initial scanning path does not meet the expected standard. At this time, the scanning angle needs to be adjusted to improve the imaging effect. The operation is carried out within a scanning angle threshold range other than the initial scanning angle, which is preset according to the performance of the imaging device, the characteristics of the imaging target, and other factors, and it limits the feasible variation interval of the scanning angle. Within this interval, a random angle is selected and determined as the secondary scanning angle. Since there is currently insufficient information to determine which angle can most effectively improve the imaging quality, the random selection method can widely explore other possible angles. Once the secondary scanning angle is determined, it is directly set as the optimal scanning path, and the subsequent imaging work will be carried out based on this new scanning path. It is expected that by changing the scanning angle, the resolution, contrast, and signal-to-noise ratio of the imaging quality parameters will be improved, and the overall imaging quality will be improved to meet or exceed the requirements of the predetermined quality threshold.

[0061] In one possible implementation manner, step S450 further includes:

[0062] Step S451: record the historical scanning angle sequence and the historical comprehensive quality coefficient sequence.

[0063] Step S452: cluster the multiple historical scanning angles in the historical scanning angle sequence according to a predetermined angle interval, determine multiple historical scanning angle intervals, and calculate multiple comprehensive quality coefficient means of the multiple historical scanning angle intervals according to the historical comprehensive quality coefficient sequence.

[0064] Step S453: if the comprehensive quality coefficient mean greater than or equal to the predetermined quality threshold is not 0, set the historical scanning angle interval corresponding to the maximum comprehensive quality coefficient mean as the optimal historical scanning angle interval, and randomly select any angle in the optimal historical scanning angle interval as the optimal scanning path.

[0065] Step S454: if the comprehensive quality coefficient mean greater than or equal to the predetermined quality threshold is 0, randomly select any angle within a scanning angle threshold other than the multiple historical scanning angle intervals as the optimal scanning path.

[0066] Specifically, after each ultrasound imaging operation is completed, the scan angle used in this imaging is accurately recorded, and these angle information is sequentially added to the historical scan angle sequence according to the imaging order. At the same time, according to the quality parameter deviation set obtained in the imaging process, the key parameters affecting the imaging quality, such as resolution deviation, contrast deviation, and signal-to-noise ratio deviation, are first found out by weighted summation, each parameter is assigned a corresponding weight, the sum of these weights is 1, then the parameter values are functionally converted (for example, take the reciprocal of the small deviation value), multiply the converted values by the weights and sum them up to obtain the comprehensive quality coefficient, and calculate the corresponding comprehensive quality coefficient. Record these coefficients in the historical comprehensive quality coefficient sequence according to the imaging order. In this way, the historical scan angle sequence and the historical comprehensive quality coefficient sequence completely record the scan angle and the corresponding quality performance of each imaging, providing rich data support for subsequent analysis.

[0067] The historical scan angles in the historical scan angle sequence are clustered according to the pre-set angle interval. For example, if the predetermined angle interval is every 10 degrees, the scan angles will be divided into intervals such as 0-10 degrees, 10-20 degrees, etc., thereby determining multiple historical scan angle intervals. Then, in combination with the historical comprehensive quality coefficient sequence, the corresponding comprehensive quality coefficients in each historical scan angle interval are summarized and the average value is calculated. This average value can reflect the average level of imaging quality in the angle interval, helping to select a better angle interval in the subsequent screening.

[0068] When the clustering of the historical scan angles is completed, and the average values of the comprehensive quality coefficients of the historical scan angle intervals are calculated, these average values are compared with the predetermined quality threshold. If the average value of the comprehensive quality coefficient that is greater than or equal to the predetermined quality threshold is not 0, that is, there is at least one historical scan angle interval whose average value of the comprehensive quality coefficient reaches or exceeds the predetermined quality threshold, which indicates that in the past scanning attempts, the imaging quality of some angle intervals has reached or even exceeded the expected standard. At this time, from these intervals that meet the conditions, the historical scan angle interval with the maximum average value of the comprehensive quality coefficient is selected. This is because the maximum average value of the comprehensive quality coefficient means that this interval has the most excellent performance when considering various quality indicators of imaging (such as resolution, contrast, signal-to-noise ratio, etc.), and has higher potential for imaging quality, so it is set as the optimal historical scan angle interval. In order to further explore the suitable scan angle in this high-quality angle interval and avoid falling into a local optimal solution, a random angle is selected in the optimal historical scan angle interval, and it is determined as the optimal scan path for subsequent ultrasound imaging operation, expecting to obtain high-quality imaging results.

[0069] If the average of the comprehensive quality coefficient greater than or equal to the predetermined quality threshold is 0, it means that the comprehensive quality of the imaging in all the historical scanning angle intervals that have been counted does not meet the expected standard, and the existing historical scanning angle data cannot find a path that meets the quality requirements. At this time, in order to obtain imaging results that meet the quality requirements, exploration is carried out in the area within the scanning angle threshold but outside the existing multiple historical scanning angle intervals. This scanning angle threshold is a reasonable angle range set in advance according to device performance, imaging requirements and other factors. In this remaining feasible angle range, a random angle is selected and determined as the optimal scanning path. In this way, the limitations of the past scanning angles are broken through, and imaging is performed at the randomly selected new angle, hoping to find a path that can bring high-quality imaging results, so that the imaging quality meets the predetermined standard.

[0070] Embodiment two, based on the same inventive concept as the real-time imaging method for three-dimensional ultrasonic imaging catheter in the foregoing embodiment, as shown in the figure, the present application provides a real-time imaging system for three-dimensional ultrasonic imaging catheter, and the system and method embodiments in the present application are based on the same inventive concept. Among them, the system includes: Figure 2

[0071] The three-dimensional imaging target model establishing module 10 is used to obtain the pre-set imaging target and tissue type data, and establish a three-dimensional imaging target model, wherein the three-dimensional imaging target model is embedded with a standard imaging quality parameter set.

[0072] The real-time imaging quality parameter set obtaining module 20 is used to perform high-frame-rate electronic scanning on the target area according to the initial probe control scheme and the initial scanning path through the predetermined three-dimensional imaging device, and perform imaging quality analysis on the returned scanning data to obtain a real-time imaging quality parameter set.

[0073] The quality parameter deviation set obtaining module 30 is used to obtain a quality parameter deviation set after deviation calculation based on the standard imaging quality parameter set and the real-time imaging quality parameter set.

[0074] The optimal probe control scheme output module 40 is used to perform probe control and scanning path optimization analysis based on the quality parameter deviation set, and output an optimal probe control scheme and an optimal scanning path for subsequent three-dimensional imaging operation.

[0075] Further, the system is also used to realize the following functions:

[0076] ​The predetermined three-dimensional imaging device is built based on a predetermined two-dimensional probe array and an ASIC architecture, the predetermined two-dimensional probe array includes a plurality of ultrasonic probe elements, wherein the plurality of ultrasonic probe elements are arranged according to a predetermined geometry, and the predetermined geometry includes a predetermined shape, a predetermined size, and a predetermined arrangement manner.

[0077] Further, the system is also used to realize the following functions:

[0078] Obtaining the backhaul scanning data, wherein the backhaul scanning data includes a plurality of echo signals backhauled by a plurality of ultrasonic probe elements; performing denoising processing and time domain alignment on the plurality of echo signals to obtain standard backhaul scanning data; performing imaging quality analysis on the standard backhaul scanning data according to a predetermined evaluation index to obtain a real-time imaging quality parameter set, wherein the imaging quality parameters at least include resolution, contrast, and signal-to-noise ratio.

[0079] Further, the system is also used to realize the following functions:

[0080] Obtaining the quality parameter deviation set, wherein the quality parameter deviation set includes resolution deviation, contrast deviation, and signal-to-noise ratio deviation; performing probe control optimization analysis with the probe control parameter threshold as the optimization space for the purpose of satisfying the resolution deviation, contrast deviation, and signal-to-noise ratio deviation, and outputting an optimal probe control scheme, wherein the probe control scheme includes probe control parameters of a plurality of probes, and the probe control parameters include signal time delay, amplitude, excitation sequence, beam width, and frequency.

[0081] Further, the system is also used to realize the following functions:

[0082] Randomly selecting a first probe control scheme within the probe control parameter threshold, using an imaging quality compensation plug-in to perform imaging quality compensation prediction on the first probe control scheme, and outputting first imaging quality compensation parameters, wherein the imaging quality compensation plug-in is built based on machine learning, and the imaging quality compensation parameters include resolution compensation data, contrast compensation data, and signal-to-noise ratio compensation data; constructing a scheme fitness evaluation function based on the resolution deviation, contrast deviation, and signal-to-noise ratio deviation to perform fitness evaluation on the first imaging quality compensation parameters, and outputting a first scheme fitness; continuing to randomly select probe control schemes within the probe control parameter threshold for iterative compensation prediction and evaluation until a predetermined iteration number is reached, and outputting the probe control scheme with the maximum scheme fitness as the optimal probe control scheme; the expression of the scheme fitness evaluation function is: ; wherein, is the scheme fitness, , , respectively, K is a resolution deviation, k is a resolution compensation data, P is a contrast deviation, p is a contrast compensation data, Y is a signal-to-noise ratio deviation, and y is a signal-to-noise ratio compensation data.

[0083] Further, the system is further configured to implement the following functions:

[0084] obtaining a first quality parameter deviation set of an initial scanning path, wherein the initial scanning path comprises an initial scanning angle; determining a first comprehensive quality coefficient according to the first quality parameter deviation set, if the first comprehensive quality coefficient is greater than or equal to a predetermined quality threshold, setting the initial scanning angle as a secondary scanning angle; if the first comprehensive quality coefficient is less than the predetermined quality threshold, randomly selecting any angle within a scanning angle threshold other than the initial scanning angle as the secondary scanning angle, and setting the secondary scanning angle as an optimal scanning path.

[0085] Further, the system is further configured to implement the following functions:

[0086] recording a history scanning angle sequence and a history comprehensive quality coefficient sequence; clustering a plurality of history scanning angles in the history scanning angle sequence according to a predetermined angle interval, determining a plurality of history scanning angle intervals, calculating a plurality of comprehensive quality coefficient means of the plurality of history scanning angle intervals according to the history comprehensive quality coefficient sequence; if the comprehensive quality coefficient means greater than or equal to the predetermined quality threshold are not 0, setting a history scanning angle interval corresponding to a maximum comprehensive quality coefficient mean as an optimal history scanning angle interval, and randomly selecting any angle within the optimal history scanning angle interval as an optimal scanning path; if the comprehensive quality coefficient means greater than or equal to the predetermined quality threshold are 0, randomly selecting any angle within a scanning angle threshold other than the plurality of history scanning angle intervals as the optimal scanning path.

[0087] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. And the above describes a specific embodiment of the present application. In addition, the processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-task processing and parallel processing are possible or can be advantageous.

[0088] The above only describes the preferred embodiments of the present application and does not limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

[0089] The specification and drawings are, of course, to be regarded in an illustrative rather than a restrictive sense. It is to be understood that any such modifications, variations, combinations or equivalents that fall within the scope of the application are intended to be embraced herein.

Claims

1. A method for real-time imaging of a three-dimensional ultrasound imaging catheter, characterized in that, The method comprises: acquiring pre-set imaging target and tissue type data, and establishing a three-dimensional imaging target model, wherein the three-dimensional imaging target model is embedded with a standard imaging quality parameter set; performing high-frame-rate electronic scanning on a target region according to an initial probe control scheme and an initial scanning path by a predetermined three-dimensional imaging device, and performing imaging quality analysis on the returned scanning data to acquire a real-time imaging quality parameter set; performing deviation calculation according to the standard imaging quality parameter set and the real-time imaging quality parameter set to obtain a quality parameter deviation set; performing probe control and scanning path optimization analysis based on the quality parameter deviation set, and outputting an optimal probe control scheme and an optimal scanning path for subsequent three-dimensional imaging operation; wherein the probe control optimization analysis based on the quality parameter deviation set comprises: acquiring the quality parameter deviation set, wherein the quality parameter deviation set comprises resolution deviation, contrast deviation and signal-to-noise ratio deviation; performing probe control optimization analysis with the threshold value of the probe control parameter as the optimization space for the purpose of satisfying the resolution deviation, the contrast deviation and the signal-to-noise ratio deviation, and outputting an optimal probe control scheme, wherein the probe control scheme comprises probe control parameters of a plurality of probes, and the probe control parameters comprise signal time delay, amplitude, excitation sequence, beam width and frequency; performing probe control optimization analysis with the threshold value of the probe control parameter as the optimization space for the purpose of satisfying the resolution deviation, the contrast deviation and the signal-to-noise ratio deviation comprises: randomly selecting a first probe control scheme within the threshold value of the probe control parameter, performing imaging quality compensation prediction on the first probe control scheme by using an imaging quality compensation plug-in, and outputting first imaging quality compensation parameters, wherein the imaging quality compensation plug-in is constructed based on machine learning, and the imaging quality compensation parameters comprise resolution compensation data, contrast compensation data and signal-to-noise ratio compensation data; constructing a scheme fitness evaluation function based on the resolution deviation, the contrast deviation and the signal-to-noise ratio deviation, and performing fitness evaluation on the first imaging quality compensation parameters to output a first scheme fitness; continuing to randomly select probe control schemes within the threshold value of the probe control parameter for iterative compensation prediction and evaluation until a predetermined number of iterations is reached, and outputting the probe control scheme with the maximum scheme fitness as the optimal probe control scheme; the expression of the scheme fitness evaluation function is: ； wherein, is the fitness of the solution, , , are the resolution weight, the contrast weight and the signal-to-noise ratio weight, respectively, K is the resolution deviation, k is the resolution compensation data, P is the contrast deviation, p is the contrast compensation data, Y is the signal-to-noise ratio deviation, and y is the signal-to-noise ratio compensation data.

2. The method for real-time imaging of a three-dimensional ultrasonic imaging catheter of claim 1, wherein, the predetermined three-dimensional imaging device is constructed based on a predetermined two-dimensional probe array and an ASIC architecture, and the predetermined two-dimensional probe array comprises a plurality of ultrasonic probe elements, wherein the plurality of ultrasonic probe elements are arranged according to a predetermined geometric structure, and the predetermined geometric structure comprises a predetermined shape, a predetermined size and a predetermined arrangement manner.

3. The method for real-time imaging of a three-dimensional ultrasonic imaging catheter of claim 2, wherein, performing imaging quality analysis on the returned scanning data to acquire a real-time imaging quality parameter set comprises: acquiring returned scanning data, wherein the returned scanning data comprises a plurality of echo signals returned by a plurality of ultrasonic probe elements; performing denoising processing and time domain alignment on the plurality of echo signals to obtain standard returned scanning data; According to the predetermined evaluation index, the standard backhaul scanning data is subjected to imaging quality analysis to obtain a real-time imaging quality parameter set, wherein the imaging quality parameters at least include resolution, contrast and signal-to-noise ratio.

4. The method for real-time imaging of a three-dimensional ultrasonic imaging catheter of claim 1, wherein, Based on the quality parameter deviation set, scanning path optimization analysis is performed, including: obtaining a first quality parameter deviation set of an initial scanning path, wherein the initial scanning path includes an initial scanning angle; determining a first comprehensive quality coefficient according to the first quality parameter deviation set, and if the first comprehensive quality coefficient is greater than or equal to a predetermined quality threshold, setting the initial scanning angle as a secondary scanning angle; if the first comprehensive quality coefficient is less than the predetermined quality threshold, randomly selecting any angle within a scanning angle threshold other than the initial scanning angle as the secondary scanning angle, and setting the secondary scanning angle as the optimal scanning path.

5. The method for real-time imaging of a three-dimensional ultrasonic imaging catheter of claim 4, wherein, obtaining the optimal scanning path, further including: recording a historical scanning angle sequence and a historical comprehensive quality coefficient sequence; clustering a plurality of historical scanning angles in the historical scanning angle sequence according to a predetermined angle interval to determine a plurality of historical scanning angle intervals, and calculating a plurality of comprehensive quality coefficient means of the plurality of historical scanning angle intervals according to the historical comprehensive quality coefficient sequence; if the comprehensive quality coefficient mean greater than or equal to the predetermined quality threshold is not 0, setting the historical scanning angle interval corresponding to the maximum comprehensive quality coefficient mean as an optimal historical scanning angle interval, and randomly selecting any angle within the optimal historical scanning angle interval as the optimal scanning path; if the comprehensive quality coefficient mean greater than or equal to the predetermined quality threshold is 0, randomly selecting any angle within a scanning angle threshold other than the plurality of historical scanning angle intervals as the optimal scanning path.

6. A real-time imaging system for a three-dimensional ultrasound imaging catheter, characterized by, The system is used to implement the real-time imaging method for a three-dimensional ultrasonic imaging catheter according to any one of claims 1-5, and the system includes: a three-dimensional imaging target model establishing module for obtaining pre-set imaging target and tissue type data, and establishing a three-dimensional imaging target model, wherein the three-dimensional imaging target model is embedded with a standard imaging quality parameter set; a real-time imaging quality parameter set obtaining module for performing high-frame-rate electronic scanning on a target region according to an initial probe control scheme and an initial scanning path through a predetermined three-dimensional imaging device, and performing imaging quality analysis on backhaul scanning data to obtain a real-time imaging quality parameter set; a quality parameter deviation set obtaining module for obtaining a quality parameter deviation set after deviation calculation based on the standard imaging quality parameter set and the real-time imaging quality parameter set; an optimal probe control scheme output module for performing probe control and scanning path optimization analysis based on the quality parameter deviation set, and outputting an optimal probe control scheme and an optimal scanning path for subsequent three-dimensional imaging operation.

Images