AS risk identification and prediction method based on AI ultrasonic image and fluid mechanics

Abstract

The invention provides an AS risk identification and prediction method based on AI ultrasonic images and hydromechanics, which quantifies the influence of plaques on blood flow through deep learning and hydromechanics simulation, and adopts ultrasonic images and Doppler data to construct a plaque AI identification model, so as to identify the AI risk of the plaques. Identifying plaque morphological characteristics in the artery cross section ultrasonic image of the patient and blood flow characteristics in the Doppler blood flow data, and outputting a morphological score corresponding to the plaque morphological characteristics and a hemodynamic score corresponding to the blood flow characteristics; and comprehensively calculating and outputting an AS risk score of the AS patient. The plaque detection accuracy based on the ultrasonic image is improved, and quantitative evaluation of the acute thrombus risk is realized. Multi-level analysis can be carried out on the form and blood flow characteristics of plaques, and more accurate risk grading support is provided for doctors. The non-invasive ultrasonic image is combined with AI and fluid mechanics simulation to realize double analysis of morphological characteristics and hemodynamic parameters of the atherosclerotic plaque.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medical technology, and in particular to an AS risk identification and prediction method based on AI ultrasonic imaging and fluid mechanics, an AS risk identification and prediction method device based on AI ultrasonic imaging and fluid mechanics, an electronic device and a computer-readable storage medium. Background Art

[0002] Cardiovascular and cerebrovascular diseases are one of the leading causes of death worldwide. The formation and rupture of atherosclerotic plaques (atherosclerosis, referred to as "AS") often lead to acute thrombosis, which in turn causes serious events such as stroke or myocardial infarction in AS patients.

[0003] The existing clinical detection technologies for acute thrombosis caused by atherosclerotic plaques mainly include: Ultrasound imaging and quantitative risk assessment technology: It is the mainstream traditional clinical detection method, but this conventional test has limitations. Its detection method relies on the doctor's experience, lacks quantitative assessment of plaque rupture risk, and fails to deeply analyze the impact of plaques on hemodynamics; Quantification of blood flow: For example, technologies from companies such as HeartFlow use CT images combined with computational fluid dynamics (CFD) to quantify blood flow, but their applicability is mostly concentrated in coronary arteries and lacks real-time performance. Intelligent identification: AI is used to identify atherosclerotic plaque images, but its identification dimension is single and it is unable to conduct multi-level analysis of the plaque morphology and blood flow characteristics to provide doctors with more accurate risk grading support. Summary of the invention

[0004] In order to solve the technical problems existing in the prior art, the present invention provides the following technical solutions: On the one hand, a method for AS risk identification and prediction based on AI ultrasound imaging and fluid mechanics is provided, which is implemented by an electronic device and includes: S1. Collect clinical examination data of AS patients: arterial cross-sectional ultrasound images and Doppler blood flow data; S2. Identify the plaque morphological features in the arterial cross-sectional ultrasound image and the blood flow features in the Doppler blood flow data through a preset plaque AI recognition model, and output a morphological score corresponding to the plaque morphological features and a hemodynamic score corresponding to the blood flow features; S3. Comprehensively calculate and output the AS risk score of AS patients: AS risk score = a × morphological score + b × hemodynamic score, Among them, a and b are preset weight coefficients.

[0005] Preferably, the method for generating the plaque AI recognition model comprises: (1) Collect clinical history examination data of several AS patients, including arterial cross-sectional ultrasound images and Doppler blood flow data; (2) preprocessing the clinical history examination data; (3) performing feature engineering on the preprocessed clinical history examination data, including: Using a convolutional neural network model to identify the plaque area in the cross-sectional ultrasound image of the artery, and extracting the plaque morphological features in the plaque area; Performing blood flow characteristic analysis on the Doppler blood flow data through a CFD model, and calculating the blood flow characteristics in the plaque area; Performing feature annotation on the plaque area, wherein the annotation information includes: a morphological score of the plaque morphological feature and the blood flow feature and its hemodynamic score; After labeling, generate feature set; According to a preset ratio, the feature set is divided into a training set, a test set and a validation set, and the training set is input into a preset CNN model to perform feature training and learning to generate the plaque AI recognition model; The test set is used to test whether the model performance of the plaque AI recognition model is qualified: If yes, go to the next step; Otherwise repeat the above steps; (6) Use the validation set to verify whether the recognition accuracy of the plaque AI recognition model is qualified: If yes, deploy and apply the plaque AI recognition model; Otherwise repeat the above steps.

[0006] Preferably, the plaque morphological characteristics include at least the following characteristics: thickness, size, shape and surface state Preferably, the blood flow characteristics include at least the following characteristics: shear stress, flow velocity variation, pressure distribution and vortex.

[0007] Preferably, the method further comprises: According to the AS risk score of the AS patient, construct a corresponding SQL search instruction; Executing the SQL search instruction to retrieve an AS intervention strategy matching the AS risk score from a MySQL database; Generate and save an AS clinical assessment report for an AS patient, wherein the AS clinical assessment report includes clinical examination data of the AS patient, an AS risk score, and an AS intervention strategy.

[0008] On the other hand, an AS risk identification and prediction device based on AI ultrasound imaging and fluid mechanics is provided, and the AS risk identification and prediction device based on AI ultrasound imaging and fluid mechanics is used to implement the above-mentioned AS risk identification and prediction method based on AI ultrasound imaging and fluid mechanics, and the device includes: Ultrasound imaging equipment is used to collect clinical examination data of AS patients: arterial cross-sectional ultrasound images and Doppler blood flow data, and upload them to the backend server; Backend server, including: A data transmission module, used for receiving the clinical examination data and sending it to the plaque AI recognition module; A plaque AI recognition module, used to identify the plaque morphological characteristics in the arterial cross-sectional ultrasound image and the blood flow characteristics in the Doppler blood flow data through a preset plaque AI recognition model, and output a morphological score corresponding to the plaque morphological characteristics and a hemodynamic score corresponding to the blood flow characteristics; The prediction module is used to comprehensively calculate and output the AS risk score of AS patients: AS risk score = a × morphological score + b × hemodynamic score, where a and b are preset weight coefficients; The data transmission module, plaque AI recognition module and prediction module are all deployed on the background server; The ultrasonic imaging device is communicatively connected to the background server.

[0009] Preferably, the device further comprises: An intervention module, for constructing a corresponding SQL search instruction according to the AS risk score of the AS patient and sending the instruction to the MySQL database; MySQL database, used to execute the SQL search instruction, retrieve and output the AS intervention strategy matching the AS risk score; An assessment report generation module is used to generate and save an AS clinical assessment report for an AS patient, wherein the AS clinical assessment report includes clinical examination data of the AS patient, an AS risk score, and an AS intervention strategy; API port, used to send the AS clinical assessment report to the patient-side APP of the AS patient for the AS patient to view; The intervention module, MySQL database, assessment report generation module and API port are all deployed on the backend server; The patient-side APP is communicatively connected with the background server.

[0010] On the other hand, an electronic device is provided, comprising: a processor; a memory, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, any one of the above-mentioned AS risk identification and prediction methods based on AI ultrasound imaging and fluid mechanics is implemented.

[0011] On the other hand, a computer-readable storage medium is provided, in which at least one instruction is stored, and the at least one instruction is loaded and executed by a processor to implement any one of the above-mentioned AS risk identification and prediction methods based on AI ultrasound imaging and fluid mechanics. The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least: The present invention is based on ultrasound imaging data, and quantifies the effect of plaques on blood flow through AI model recognition and fluid dynamics simulation. Through deep learning and fluid dynamics simulation, ultrasound images and Doppler data are used to build a plaque AI recognition model to identify the plaque morphological characteristics in the patient's arterial cross-sectional ultrasound image and the blood flow characteristics in the Doppler blood flow data, and output the morphological score corresponding to the plaque morphological characteristics and the hemodynamic score corresponding to the blood flow characteristics; comprehensively calculate and output the AS risk score of AS patients. Improve the accuracy of plaque detection based on ultrasound imaging, and realize the quantitative assessment of acute thrombosis risk. The morphology and blood flow characteristics of the plaque can be analyzed at multiple levels to provide doctors with more accurate risk grading support.

[0012] The present invention realizes the dual analysis of the morphological characteristics and hemodynamic parameters of atherosclerotic plaques by combining non-invasive ultrasound imaging with AI and fluid dynamics simulation, which has the following clinical application values: Efficient and non-invasive: The data collection method based on ultrasound imaging reduces the invasiveness of the test and facilitates follow-up and multiple evaluations; Real-time and accurate: Fluid mechanics simulation supported by supercomputers makes data processing more efficient and provides clinicians with real-time risk assessment results; Multi-site detection: supports simultaneous detection of multiple sites of arteries such as the carotid artery, femoral artery, and aorta, achieving comprehensive risk management of cardiovascular and cerebrovascular diseases; Quantitative assessment: Through risk scores and risk grading systems, a standardized basis for risk assessment is provided, which is suitable for individualized intervention guidance. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0014] Figure 1 This is a flow chart of an AS risk identification and prediction method based on AI ultrasound imaging and fluid mechanics provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of performing plaque recognition and labeling on an arterial image provided by an embodiment of the present invention; Figure 3 It is a block diagram of an AS risk identification and prediction device based on AI ultrasound imaging and fluid mechanics provided by an embodiment of the present invention; Figure 4 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0015] The technical solution of the present invention is described below in conjunction with the accompanying drawings.

[0016] In the embodiments of the present invention, words such as "exemplarily" and "for example" are used to indicate examples, illustrations or explanations. Any embodiment or design described as "example" in the present invention should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of the word "example" is intended to present the concept in a specific way. In addition, in the embodiments of the present invention, the meaning expressed by "and / or" can be both, or it can be either of the two.

[0017] In the embodiments of the present invention, "image" and "picture" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same. "of", "corresponding, relevant" and "corresponding" can sometimes be used interchangeably. It should be noted that when the difference between them is not emphasized, the meanings they intend to express are the same.

[0018] In the embodiments of the present invention, sometimes a subscript such as W1 may be mistakenly written as a non-subscript such as W1. When the difference is not emphasized, the meanings to be expressed are consistent.

[0019] In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0020] The present invention mainly solves the following technical problems: How to automatically identify the area of ​​atherosclerotic plaques and quantify their morphological characteristics through AI models based on ultrasound images? How to use fluid dynamics simulation to conduct real-time quantitative analysis of the hemodynamic impact of plaques? How to combine plaque morphological scores and hemodynamic scores to generate quantitative acute thrombosis risk scores and grades? In order to solve the above problems, the present invention proposes an AS risk identification and prediction method based on AI ultrasound imaging and fluid mechanics, which aims to improve the accuracy of plaque detection based on ultrasound imaging and achieve quantitative assessment of acute thrombosis risk through deep learning and fluid mechanics simulation. The system can perform multi-level analysis of plaque morphology and blood flow characteristics, providing doctors with more accurate risk grading support.

[0021] The implementation principle of the present invention will be described in detail below.

[0022] The embodiment of the present invention provides a method for identifying and predicting AS risk based on AI ultrasound imaging and fluid mechanics. The method can be implemented by an electronic device, which can be a terminal or a server. Figure 1 The flowchart of the AS risk identification and prediction method based on AI ultrasound imaging and fluid mechanics is shown. The processing flow of the method may include the following steps: S1. Collect clinical examination data of AS patients: arterial cross-sectional ultrasound images and Doppler blood flow data; S2. Identify the plaque morphological features in the arterial cross-sectional ultrasound image and the blood flow features in the Doppler blood flow data through a preset plaque AI recognition model, and output a morphological score corresponding to the plaque morphological features and a hemodynamic score corresponding to the blood flow features; S3. Comprehensively calculate and output the AS risk score of AS patients: AS risk score = a × morphological score + b × hemodynamic score, Among them, a and b are preset weight coefficients.

[0023] AS patients are patients with cardiovascular and cerebrovascular diseases, especially patients with atherosclerosis. Patients can log in to the backend server through APP and other ports to register for medical treatment, bind the examination data to the hospital management system, etc. The medical imaging department can upload the clinical data of the examined patients to the backend and bind it to the patient's medical ID / identity information, etc., and can also record it in the HIS system. The specific understanding can be combined with the data management services of existing hospitals.

[0024] The present invention is mainly based on ultrasound imaging data, through AI model recognition and fluid dynamics simulation, to quantify the effect of plaque on blood flow, specifically including the following steps: Data acquisition and preprocessing: Collect plaque structure and blood flow characteristics based on ultrasound images and Doppler data; AI model plaque recognition and morphological scoring: Use deep learning models (such as convolutional neural networks (CNN)) to automatically identify plaque areas, extract morphological features and generate morphological scores; Fluid mechanics simulation and hemodynamic scoring: Analyze blood flow characteristics based on CFD models, and calculate shear stress, flow velocity distribution and pressure changes; Comprehensive score and classification: The morphological score and hemodynamic score are combined into a risk score by weighting, and the risk classification is output. The morphological score and hemodynamic score are combined and the risk score is calculated using weights: Risk score = a × morphological score + b × hemodynamic score.

[0025] The following describes the training and working mechanism of the plaque AI recognition model (referred to as “AI model”) of the present invention.

[0026] Preferably, the method for generating the plaque AI recognition model comprises: (1) Collect clinical history examination data of several AS patients, including arterial cross-sectional ultrasound images and Doppler blood flow data; (2) preprocessing the clinical history examination data; (3) performing feature engineering on the preprocessed clinical history examination data, including: Using a convolutional neural network model to identify the plaque area in the cross-sectional ultrasound image of the artery, and extracting the plaque morphological features in the plaque area; Performing blood flow characteristic analysis on the Doppler blood flow data through a CFD model, and calculating the blood flow characteristics in the plaque area; Performing feature annotation on the plaque area, wherein the annotation information includes: a morphological score of the plaque morphological feature and the blood flow feature and its hemodynamic score; After labeling, generate feature set; According to a preset ratio, the feature set is divided into a training set, a test set and a validation set, and the training set is input into a preset CNN model to perform feature training and learning to generate the plaque AI recognition model; The test set is used to test whether the model performance of the plaque AI recognition model is qualified: If yes, go to the next step; Otherwise repeat the above steps; (6) Use the validation set to verify whether the recognition accuracy of the plaque AI recognition model is qualified: If yes, deploy and apply the plaque AI recognition model; Otherwise repeat the above steps.

[0027] Preferably, the plaque morphological characteristics include at least the following characteristics: thickness, size, shape and surface state Preferably, the blood flow characteristics include at least the following characteristics: shear stress, flow velocity variation, pressure distribution and vortex.

[0028] The present invention trains the plaque AI recognition model based on the CNN model. The main implementation is as follows: 1. Dataset Construction Data source: Ultrasound images and Doppler data provided by multiple medical institutions, covering a variety of plaque types and different blood flow states.

[0029] 2. Data preprocessing: denoise and standardize the collected data to ensure data quality.

[0030] 3. Data annotation: Experts annotate the morphology and blood flow characteristics of plaques to form a labeled data set for model training.

[0031] Labeling scheme Plaque area annotation: Manual or semi-automatic annotation of plaque shape, thickness, surface state and other features.

[0032] Blood flow feature annotation: Mark high-risk areas such as shear stress peak position and eddy current distribution to ensure accurate annotation of blood flow features in the data set.

[0033] like Figure 2 The figure shows a schematic diagram of data annotation on an ultrasound image. A schematic diagram of identifying and marking the plaque location, plaque area contour, and plaque annotation on an image (the annotation information is indicated by "xx" in the figure). In specific operations, the administrator can annotate images of different angles and different patients to construct corresponding annotation training sets.

[0034] 4. Model Training Model structure: The convolutional neural network (CNN) structure is used to input ultrasound images and output plaque recognition results and feature quantification information.

[0035] Training process: Use the labeled dataset to train the model and optimize the model through the cross entropy loss function.

[0036] Hyperparameter Optimization: Optimize hyperparameters via grid search.

[0037] Algorithm performance requirements: morphological recognition accuracy on the test set ≥ 95%, plaque feature extraction accuracy ≥ 90%.

[0038] 5. Algorithm performance verification method Model Validation: Use the validation set to evaluate recognition accuracy, recall, and F1 score.

[0039] Fluid dynamics simulation verification: Compare the simulation data with the actual clinical data to verify the accuracy of the CFD model.

[0040] Clinical validation: Validate the predictive accuracy of the risk score on actual patient data.

[0041] The specific model verification can be completed by the administrator.

[0042] After the model is deployed, the model parameters can be configured and deployed on the backend server to identify the patient clinical data collected by the ultrasound imaging device. After the data is uploaded to the backend, it is recorded in the medical records of the HIS and input into the AI ​​model to identify the plaque morphology and blood flow characteristics, and output the scores marked on the corresponding characteristics.

[0043] The analysis of blood flow characteristics based on CFD (computational fluid dynamics) models, including calculation of shear stress, flow velocity distribution, and pressure changes, usually follows the following steps: 1). Problem definition and geometric modeling ‌Problem Definition‌: Define the analysis target, such as blood flow characteristics in a specific blood vessel or part of the heart.

[0044] Determine the desired physical quantities and accuracy requirements.

[0045] ‌Geometry Modeling‌: Use the geometric modeling tools provided by the CAD software or CFD software to construct a geometric model of the blood flow channel.

[0046] For complex vascular networks, simplification or segmentation may be necessary.

[0047] 2). Grid division ‌Grid type selection‌: Select the appropriate mesh type (such as structured mesh, unstructured mesh) based on the complexity of the geometric model and computational requirements.

[0048] ‌Mesh Generation‌: Use the mesh generation tool in the CFD software to mesh the geometric model.

[0049] Ensure that the mesh density is sufficient to capture detailed features of blood flow while taking into account the limitations of computational resources.

[0050] ‌Grid Optimization‌: The mesh is refined or coarsened to balance computational accuracy and efficiency.

[0051] 3). Mathematical model and boundary conditions ‌Mathematical model selection‌: Select an appropriate CFD mathematical model, such as the Navier-Stokes equations, to describe the flow behavior of blood flow.

[0052] ‌Boundary condition settings‌: Set the inlet boundary conditions (such as flow rate, pressure, etc.).

[0053] Set outlet boundary conditions (such as pressure outlet, free outflow, etc.).

[0054] Set wall boundary conditions (such as no-slip wall, slip wall, etc.).

[0055] 4). Solution settings ‌Solver Selection‌: Select an appropriate solver (e.g., implicit solver, explicit solver) based on the complexity of the problem and computational requirements.

[0056] ‌Solving parameter settings‌: Set solution parameters such as time step and number of iterations.

[0057] Select an appropriate convergence criterion to ensure the accuracy of the calculation results.

[0058] 5). Calculation and post-processing calculate: Run CFD software to perform numerical simulation calculations.

[0059] Monitor the convergence and residual changes during the calculation process to ensure the stability of the calculation.

[0060] Post-processing: Extract calculation results, including velocity distribution, pressure distribution, shear stress, etc.

[0061] Use post-processing tools in CFD software or third-party visualization software to visualize and analyze calculation results.

[0062] 6). Result analysis and verification ‌Result Analysis‌: Analyze blood flow characteristics such as velocity distribution, pressure changes and shear stress to understand the flow behavior and dynamic characteristics of blood flow.

[0063] Compare the calculation results under different working conditions and evaluate the changes and impacts of blood flow characteristics.

[0064] ‌Result Verification‌: Compare the calculation results with experimental data or clinical data to verify the accuracy and reliability of the CFD model.

[0065] If necessary, the model is modified and optimized to improve the accuracy of the calculation results.

[0066] In summary, the process of analyzing blood flow characteristics and calculating shear stress, velocity distribution and pressure changes based on CFD models involves multiple steps, including problem definition and geometric modeling, meshing, mathematical model and boundary condition setting, solution setting, calculation and post-processing, and result analysis and verification. These steps are interrelated and together constitute a complete CFD analysis process.

[0067] The specific applications of the constructed model include: data input, processing and output logic: 1. Data Input Input sources: arterial cross-sectional images and Doppler blood flow data collected by ultrasound imaging equipment.

[0068] Input data types: including plaque morphological characteristics (such as thickness, size, surface characteristics) and hemodynamic characteristics (i.e. blood flow characteristics, such as blood flow velocity, flow, pressure, etc.).

[0069] 2. Data processing Plaque identification and morphological scoring: Identification and feature extraction: Use AI models to identify plaque areas and extract their morphological features, including thickness, size, shape, and surface state.

[0070] 3. Scoring logic: Weights are assigned according to feature importance to form a morphological score. The total score of the morphological score is generally set to 100 points, including the following specific scores: Thickness: Thicker plaques are scored higher, and a weight of 30 points is generally given.

[0071] Shape: Irregular plaques are scored higher, and are generally given a weight of 40 points.

[0072] Surface condition: Rough and ulcerated plaques have high scores, and the weight is generally set at 30 points.

[0073] The morphological features are combined to calculate the morphological score, which is used to characterize the stability of the plaque. (If there is also size, the corresponding weight score can be set accordingly).

[0074] Hemodynamic score: Feature extraction: CFD simulation is used to calculate blood flow characteristics, including shear stress, flow velocity changes, pressure distribution, and eddy currents.

[0075] Scoring logic: Shear stress: The surface and adjacent areas with high shear stress have a high score, set at 40 points.

[0076] Flow velocity distribution: The score of sudden change and acceleration of flow velocity in narrow areas is high, set at 40 points.

[0077] Pressure change: The high pressure difference score is higher, especially the alternating distribution of low pressure and high pressure, which is set at 10 points.

[0078] Eddy current: The area with strong eddy current has a high score, set at 10 points.

[0079] The total score of the hemodynamic score is set to 100 points and is used to quantify the effect of plaque on blood flow.

[0080] Weighted calculation: The morphological score and hemodynamic score are combined and the risk score is calculated using weights: Risk score = a × morphological score + b × hemodynamic score, The weight parameters a and b were determined based on clinical data.

[0081] 4. Data Output Output: Generates a quantitative risk score and grade (low, intermediate, high) for acute thrombosis.

[0082] Output purpose: Provide doctors with clear risk assessment data for intervention guidance.

[0083] The backend system can pre-configure the assessment report template, so that the backend can write the patient's data into the template to generate the AS clinical assessment report for AS patients. The template can be configured by the administrator.

[0084] Preferably, the method further comprises: According to the AS risk score of the AS patient, construct a corresponding SQL search instruction; Executing the SQL search instruction to retrieve an AS intervention strategy matching the AS risk score from a MySQL database; Generate and save an AS clinical assessment report for an AS patient, wherein the AS clinical assessment report includes clinical examination data of the AS patient, an AS risk score, and an AS intervention strategy.

[0085] Here, the background can also automatically recommend AS intervention strategies corresponding to the risk level for AS patients. AS intervention strategies corresponding to different AS risk scores can be pre-configured in the background MySQL database, and then the AS intervention strategies corresponding to the AS risk scores can be retrieved. SQL instructions based on the AS risk scores can be used to retrieve data in the database, retrieve and output the corresponding AS intervention strategies. The AS intervention strategies that match the patient's AS risk score can be written into the patient's report, saved and sent to the patient APP for the patient to view.

[0086] Doctors and others can log in to the backend through a PC, view the patient's HIS files according to their permissions, and view the patient's AS clinical assessment report.

[0087] Therefore, the present invention provides a method for real-time detection and prediction of acute thrombosis risk through AI ultrasound image analysis and fluid dynamics simulation combined with supercomputer processing. The system can not only analyze the impact of plaques on blood flow, but also quantify the risks caused by fluid dynamics abnormalities, provide doctors with detailed risk assessment data, and support the early prevention and management of cardiovascular and cerebrovascular diseases.

[0088] Figure 3 1 is a block diagram of an AS risk identification and prediction device based on AI ultrasound imaging and fluid mechanics according to an exemplary embodiment, wherein the device is used for an AS risk identification and prediction method based on AI ultrasound imaging and fluid mechanics. Figure 3 The device includes an ultrasonic imaging device 310, a backend server 320, and a patient-side APP 330 (it may also include a doctor-side PC, which is not shown here). Among them: Ultrasound imaging equipment, used to collect clinical examination data of AS patients: arterial cross-sectional ultrasound images and Doppler blood flow data, and upload them to the backend server; Backend server, including: A data transmission module, used for receiving the clinical examination data and sending it to the plaque AI recognition module; A plaque AI recognition module, used to identify the plaque morphological characteristics in the arterial cross-sectional ultrasound image and the blood flow characteristics in the Doppler blood flow data through a preset plaque AI recognition model, and output a morphological score corresponding to the plaque morphological characteristics and a hemodynamic score corresponding to the blood flow characteristics; The prediction module is used to comprehensively calculate and output the AS risk score of AS patients: AS risk score = a × morphological score + b × hemodynamic score, where a and b are preset weight coefficients; The data transmission module, plaque AI recognition module and prediction module are all deployed on the background server; The ultrasonic imaging device is communicatively connected to the background server.

[0089] Preferably, the device further comprises: An intervention module, for constructing a corresponding SQL search instruction according to the AS risk score of the AS patient and sending the instruction to the MySQL database; MySQL database, used to execute the SQL search instruction, retrieve and output the AS intervention strategy matching the AS risk score; An assessment report generation module is used to generate and save an AS clinical assessment report for an AS patient, wherein the AS clinical assessment report includes clinical examination data of the AS patient, an AS risk score, and an AS intervention strategy; API port, used to send the AS clinical assessment report to the patient-side APP of the AS patient for the AS patient to view; The intervention module, MySQL database, assessment report generation module and API port are all deployed on the backend server; The patient-side APP is communicatively connected with the background server.

[0090] The functions and interactive relationships of each subject in the above-mentioned device should be understood in conjunction with the above-mentioned method steps, and will not be described in detail in this embodiment.

[0091] The communication connection methods between the device and the APP, etc., and the background, respectively, can refer to the methods of existing hospital management systems, and this embodiment does not limit it.

[0092] Figure 4 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention, such as Figure 4 As shown, the electronic device may include the above Figure 3 The AS risk identification and prediction device based on AI ultrasound imaging and fluid mechanics is shown. Optionally, the electronic device 410 may include a first processor 2001.

[0093] Optionally, the electronic device 410 may further include a memory 2002 and a transceiver 2003 .

[0094] The first processor 2001, the memory 2002 and the transceiver 2003 may be connected via a communication bus.

[0095] Combine the following Figure 4 The components of the electronic device 410 are described in detail: The first processor 2001 is the control center of the electronic device 410, and may be a processor or a general term for multiple processing elements. For example, the first processor 2001 is one or more central processing units (CPUs), or may be application specific integrated circuits (ASICs), or may be one or more integrated circuits configured to implement the embodiments of the present invention, such as one or more microprocessors (digital signal processors, DSPs), or one or more field programmable gate arrays (field programmable gate arrays, FPGAs).

[0096] Optionally, the first processor 2001 can execute various functions of the electronic device 410 by running or executing a software program stored in the memory 2002 and calling data stored in the memory 2002 .

[0097] In a specific implementation, as an embodiment, the first processor 2001 may include one or more CPUs, such as Figure 4 CPU0 and CPU1 are shown in FIG.

[0098] In a specific implementation, as an embodiment, the electronic device 410 may also include multiple processors, such as Figure 4 The first processor 2001 and the second processor 2004 are shown in FIG. Each of these processors may be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). The processor here may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0099] The memory 2002 is used to store the software program for executing the solution of the present invention, and is controlled to be executed by the first processor 2001. The specific implementation method can refer to the above method embodiment, which will not be repeated here.

[0100] Optionally, the memory 2002 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 2002 may be integrated with the first processor 2001, or may exist independently and access the first processor 2001 through the interface circuit ( Figure 4 (not shown) is coupled to the first processor 2001, which is not specifically limited in this embodiment of the present invention.

[0101] The transceiver 2003 is used to communicate with a network device or a terminal device.

[0102] Optionally, the transceiver 2003 may include a receiver and a transmitter ( Figure 4 The receiver is used to implement a receiving function, and the transmitter is used to implement a sending function.

[0103] Optionally, the transceiver 2003 may be integrated with the first processor 2001, or may exist independently and communicate with the first processor 2001 through the interface circuit ( Figure 4 (not shown) is coupled to the first processor 2001, which is not specifically limited in this embodiment of the present invention.

[0104] It should be noted that Figure 4 The structure of the electronic device 410 shown in the figure does not constitute a limitation on the router, and the actual knowledge structure recognition device may include more or fewer components than those shown in the figure, or combine certain components, or arrange the components differently.

[0105] In addition, the technical effects of the electronic device 410 can refer to the technical effects of the AS risk identification and prediction method based on AI ultrasonic imaging and fluid mechanics described in the above method embodiment, which will not be repeated here.

[0106] It should be understood that the first processor 2001 in the embodiment of the present invention may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0107] It should also be understood that the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic random access memory (DRAM), synchronous DRAM (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0108] The above embodiments can be implemented in whole or in part by software, hardware (such as circuits), firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present invention is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0109] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0110] In the present invention, "at least one" means one or more, and "plurality" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0111] It should be understood that in various embodiments of the present invention, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0112] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0113] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described equipment, devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0114] In the several embodiments provided by the present invention, it should be understood that the disclosed devices, apparatuses and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0115] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0116] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0117] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0118] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for AS risk identification and prediction based on AI ultrasound imaging and fluid mechanics, characterized in that: The method comprises: S1. Collect clinical examination data of AS patients: arterial cross-sectional ultrasound images and Doppler blood flow data; S2. Identify the plaque morphological features in the arterial cross-sectional ultrasound image and the blood flow features in the Doppler blood flow data through a preset plaque AI recognition model, and output a morphological score corresponding to the plaque morphological features and a hemodynamic score corresponding to the blood flow features; S3. Comprehensively calculate and output the AS risk score of AS patients: AS risk score = a × morphological score + b × hemodynamic score, Among them, a and b are preset weight coefficients.

2. The AS risk identification and prediction method based on AI ultrasound imaging and fluid mechanics according to claim 1 is characterized in that: The method for generating the plaque AI recognition model comprises: (1) Collect clinical history examination data of several AS patients, including arterial cross-sectional ultrasound images and Doppler blood flow data; (2) preprocessing the clinical history examination data; (3) performing feature engineering on the preprocessed clinical history examination data, including: Using a convolutional neural network model to identify the plaque area in the cross-sectional ultrasound image of the artery, and extracting the plaque morphological features in the plaque area; Performing blood flow characteristic analysis on the Doppler blood flow data through a CFD model, and calculating the blood flow characteristics in the plaque area; Performing feature annotation on the plaque area, wherein the annotation information includes: a morphological score of the plaque morphological feature and the blood flow feature and its hemodynamic score; After labeling, generate feature set; According to a preset ratio, the feature set is divided into a training set, a test set and a validation set, and the training set is input into a preset CNN model to perform feature training and learning to generate the plaque AI recognition model; The test set is used to test whether the model performance of the plaque AI recognition model is qualified: If yes, go to the next step; Otherwise repeat the above steps; (6) Use the validation set to verify whether the recognition accuracy of the plaque AI recognition model is qualified: If yes, deploy and apply the plaque AI recognition model; Otherwise repeat the above steps.

3. The AS risk identification and prediction method based on AI ultrasound imaging and fluid mechanics according to claim 2 is characterized in that: The plaque morphological characteristics include at least the following characteristics: thickness, size, shape and surface state.

4. The AS risk identification and prediction method based on AI ultrasound imaging and fluid mechanics according to claim 2 is characterized in that: The blood flow characteristics include at least the following characteristics: shear stress, flow velocity change, pressure distribution and vortex.

5. The AS risk identification and prediction method based on AI ultrasound imaging and fluid mechanics according to claim 1 is characterized in that: The method further comprises: According to the AS risk score of the AS patient, construct a corresponding SQL search instruction; Executing the SQL search instruction to retrieve an AS intervention strategy matching the AS risk score from a MySQL database; Generate and save an AS clinical assessment report for an AS patient, wherein the AS clinical assessment report includes clinical examination data of the AS patient, an AS risk score, and an AS intervention strategy.

6. An AS risk identification and prediction device based on AI ultrasound imaging and fluid mechanics, the AS risk identification and prediction device based on AI ultrasound imaging and fluid mechanics is used to implement the AS risk identification and prediction method based on AI ultrasound imaging and fluid mechanics as claimed in any one of claims 1 to 5, characterized in that: The device comprises: Ultrasound imaging equipment is used to collect clinical examination data of AS patients: arterial cross-sectional ultrasound images and Doppler blood flow data, and upload them to the backend server; Backend server, including: A data transmission module, used for receiving the clinical examination data and sending it to the plaque AI recognition module; A plaque AI recognition module, used to identify the plaque morphological characteristics in the arterial cross-sectional ultrasound image and the blood flow characteristics in the Doppler blood flow data through a preset plaque AI recognition model, and output a morphological score corresponding to the plaque morphological characteristics and a hemodynamic score corresponding to the blood flow characteristics; The prediction module is used to comprehensively calculate and output the AS risk score of AS patients: AS risk score = a × morphological score + b × hemodynamic score, where a and b are preset weight coefficients; The data transmission module, plaque AI recognition module and prediction module are all deployed on the background server; The ultrasonic imaging device is communicatively connected to the background server.

7. The AS risk identification and prediction device based on AI ultrasound imaging and fluid mechanics according to claim 6 is characterized in that: The device also includes: An intervention module, for constructing a corresponding SQL search instruction according to the AS risk score of the AS patient and sending the instruction to the MySQL database; MySQL database, used to execute the SQL search instruction, retrieve and output the AS intervention strategy matching the AS risk score; An assessment report generation module is used to generate and save an AS clinical assessment report for an AS patient, wherein the AS clinical assessment report includes clinical examination data of the AS patient, an AS risk score, and an AS intervention strategy; API port, used to send the AS clinical assessment report to the patient-side APP of the AS patient for the AS patient to view; The intervention module, MySQL database, assessment report generation module and API port are all deployed on the backend server; The patient-side APP is communicatively connected with the background server.

8. An electronic device, characterized in that: The electronic device comprises: processor; A memory having computer-readable instructions stored thereon, wherein when the computer-readable instructions are executed by the processor, the method according to any one of claims 1 to 5 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program codes, which can be called by a processor to execute the method according to any one of claims 1 to 5.

Images