Heart disease prediction method and device based on deep learning

Through a deep learning-based heart disease prediction method, feature extraction and model training are used for cardiac ultrasound video sets, and cardiac ultrasound images are automatically analyzed, which solves the problems of high cost and complex operation of existing heart disease diagnosis methods, and improves early diagnosis capabilities and diagnostic accuracy reduces the risk of side effects.

CN118983090APending Publication Date: 2024-11-19SHANDONG IND TECH RES INST (QINGDAO) +1
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202410998003.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-24
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

The existing diagnosis methods for heart disease are costly, complex in operation, and potential surgical risks, and radiation has side effects on subjects, limiting the application of early diagnosis.

Method used

Using a deep learning-based heart disease prediction method, we can obtain cardiac ultrasound video sets, mark and segment images, extract features, and input deep learning models for feature learning, automatically analyze cardiac ultrasound images, quantify myocardial movement characteristics, and calculate cardiovascular disease risks. Probability, provides predicted results for disease type and future probability of occurrence.

Benefits of technology

It reduces the cost of heart disease diagnosis, improves early diagnosis capabilities, provides more detailed auxiliary diagnostic information, reduces subjective errors of doctors, and improves the quality of treatment and management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118983090A_ABST
    Figure CN118983090A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of medical treatment, and discloses a heart disease prediction method and device based on deep learning, and the method comprises the steps: obtaining a to-be-detected cardiac ultrasound image of a patient, and enabling the to-be-detected cardiac ultrasound image to be collected by a cardiac color ultrasound instrument; the to-be-detected cardiac ultrasound image is input into a pre-established heart disease prediction model, a heart disease prediction result is output, and the heart disease prediction result comprises the disease type and the disease probability of the disease in the future preset time. The heart disease risk prediction is carried out by adopting the cardiac ultrasound image, the operation is simple, the examination cost can be effectively reduced, and the diagnosis experience of a patient is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and in particular to a method and device for predicting heart disease based on deep learning. Background Art

[0002] Heart disease is a common health problem, and currently most heart disease diagnoses are based on radiological methods, such as computed tomography angiography and coronary angiography, etc. These methods can directly image the coronary arteries and quantify the degree of arterial occlusion, and are therefore considered the gold standard for diagnosis.

[0003] However, due to the high requirements for the operating environment and technology, as well as the high equipment and maintenance costs, the cost of this method is much higher than that of ultrasound examination, which severely limits its application in the early diagnosis of heart disease. Most importantly, this method also has potential surgical risks, and its radiation can cause serious side effects on the subjects. Summary of the invention

[0004] In view of this, the present invention provides a method and device for predicting heart disease based on deep learning to solve the problem that current heart disease diagnosis methods have high requirements and costs.

[0005] In a first aspect, the present invention provides a method for predicting heart disease based on deep learning, the method comprising:

[0006] Obtain the patient's pending echocardiogram image, which is acquired by a cardiac color doppler ultrasound instrument;

[0007] The to-be-detected echocardiogram image is input into a pre-established heart disease prediction model, and a heart disease prediction result is output. The heart disease prediction result includes: disease type and the probability of the disease occurring within a preset time in the future.

[0008] By using non-invasive color Doppler echocardiogram to collect echocardiogram images as the data to be tested and inputting the echocardiogram images to be tested into a pre-established heart disease prediction model, the heart disease prediction results can be obtained. Using echocardiogram images to predict the risk of heart disease is not only simple to operate, but also can effectively reduce the inspection cost and improve the patient's diagnostic experience.

[0009] In an optional embodiment, the heart disease prediction model is established by the following steps:

[0010] Obtain the patient's cardiac ultrasound video set;

[0011] Mark each frame of the echocardiogram in the echocardiogram video set, and divide the echocardiogram into a training set, a test set, and a validation set;

[0012] identifying the cardiac chamber wall and the ventricular endocardium in each frame of the echocardiogram video set;

[0013] Dynamically track the position changes of the heart chamber wall and the ventricular endocardium, and extract features to obtain a feature set corresponding to each frame of echocardiogram;

[0014] The labeled echocardiogram images and the feature sets corresponding to the echocardiogram images are input into the deep learning model for feature learning to obtain a heart disease prediction model after feature learning.

[0015] Through the new computer vision and unsupervised machine learning model, the patient's cardiac ultrasound images are automatically analyzed, the patient's myocardial motion characteristics are quantified, and the risk probability of the patient suffering from cardiovascular disease is calculated using the machine learning model. The heart disease prediction model can effectively improve the ability to diagnose heart disease in the early stage and provide more auxiliary information for clinicians.

[0016] In an optional embodiment, after obtaining the heart disease prediction result, the method further includes:

[0017] Obtain the patient's clinical information and the imaging parameters when collecting the echocardiogram to be tested;

[0018] Generate analysis reports based on patient clinical information, imaging parameters, and heart disease prediction results.

[0019] The analysis reports generated through patient clinical data and predicted results can provide doctors with clearer and more detailed auxiliary diagnostic information, which can help doctors better understand the patient's information and condition, while improving the quality of patient treatment and management.

[0020] In an optional implementation, marking each frame of the echocardiogram in the echocardiogram video set includes:

[0021] In case of lesions on echocardiogram, the lesion area is marked;

[0022] In the case that there is no lesion in the echocardiogram, the echocardiogram is marked as normal;

[0023] Mark the locations of the heart chamber walls and ventricular lining.

[0024] Marking each frame of echocardiogram in the echocardiogram video set can effectively improve the accuracy of model building.

[0025] In an optional embodiment, the heart disease prediction result also includes: myocardial motion indicators corresponding to the disease type, wherein the myocardial motion indicators include: the movement of the heart chamber wall during heart contraction or relaxation, myocardial strain, and ejection fraction.

[0026] In terms of auxiliary diagnosis of heart disease, it can not only give a heart disease risk prediction based on myocardial motion characteristics, but also automatically give more accurate and detailed myocardial motion indicators, avoiding the limitations of a single decision and making the final result more stable and reliable, thereby further assisting doctors in diagnosis and avoiding diagnostic errors caused by individual subjective factors and human errors of doctors.

[0027] In an optional embodiment, the disease types include: myocardial ischemia, coronary heart disease, hypertrophic cardiomyopathy, and myocardial amyloidosis.

[0028] The provided heart disease prediction model can predict various types of diseases and has broad application prospects.

[0029] In a second aspect, the present invention provides a heart disease prediction device based on deep learning, the device comprising:

[0030] An acquisition module is used to acquire the patient's to-be-detected echocardiogram, where the to-be-detected echocardiogram is acquired by a cardiac color doppler ultrasound instrument;

[0031] The prediction module is used to input the to-be-detected echocardiogram image into a pre-established heart disease prediction model and output a heart disease prediction result, which includes: the disease type and the probability of the disease occurring within a preset time in the future.

[0032] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the deep learning-based heart disease prediction method of the above-mentioned first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0033] In a fourth aspect, the present invention provides a computer-readable storage medium storing computer instructions for enabling a computer to execute the deep learning-based heart disease prediction method of the first aspect or any corresponding embodiment thereof.

[0034] In a fifth aspect, the present invention provides a computer program product, comprising computer instructions, which are used to enable a computer to execute the deep learning-based heart disease prediction method of the above-mentioned first aspect or any corresponding embodiment thereof.

[0035] It should be noted that the heart disease prediction device based on deep learning, the computer device and the computer-readable storage medium provided by the present invention correspond to the heart disease prediction method based on deep learning. Therefore, for the beneficial effects of the heart disease prediction device based on deep learning, the computer device and the computer-readable storage medium, please refer to the description of the corresponding beneficial effects of the heart disease prediction method based on deep learning above, which will not be repeated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 is a flowchart of a method for predicting heart disease based on deep learning according to an embodiment of the present invention;

[0038] Figure 2 is a schematic diagram of a process for generating an analysis report according to an embodiment of the present invention;

[0039] Figure 3 is a structural block diagram of a heart disease prediction device based on deep learning according to an embodiment of the present invention;

[0040] Figure 4 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0042] According to an embodiment of the present invention, an embodiment of a heart disease prediction method based on deep learning is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in an order different from that shown here.

[0043] In this embodiment, a method for predicting heart disease based on deep learning is provided, which can be executed by a server, a terminal, a mobile terminal and other devices. Figure 1 is a flowchart of a method for predicting heart disease based on deep learning according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0044] Step S101, obtaining the patient's to-be-detected echocardiogram, which is acquired by a cardiac color ultrasound instrument. The cardiac color ultrasound instrument can be used to acquire the patient's echocardiogram, which can be any type of ultrasound instrument, and high-quality cardiac ultrasound images are preferred.

[0045] Step S102, input the to-be-detected echocardiogram image into a pre-established heart disease prediction model, and output a heart disease prediction result, which includes: the disease type and the probability of the disease occurring within a preset time in the future. The disease type may include myocardial ischemia, coronary heart disease, etc., and the probability of the disease occurring in the future may be the probability value of a patient developing a specific type of heart disease in the next few months or years, for example: the probability of a patient suffering from myocardial ischemia within the next six months is 90%.

[0046] In the diagnosis of cardiovascular diseases, resting cardiac ultrasound examination is one of the essential links in the diagnostic process. Cardiac ultrasound examination has the advantages of being non-invasive, convenient and relatively inexpensive. Therefore, in this embodiment, the cardiac ultrasound images collected by a non-invasive cardiac color ultrasound instrument are used as the data to be detected, and the cardiac ultrasound images to be detected are input into the pre-established heart disease prediction model to obtain the heart disease prediction results. The use of cardiac ultrasound images for heart disease risk prediction is not only simple to operate, but also can effectively reduce the inspection cost and improve the patient's diagnostic experience.

[0047] In some optional embodiments, the heart disease prediction model is established by the following steps:

[0048] Step S201, obtaining a set of echocardiographic video images of the patient. An echocardiographic machine generally used for echocardiography will output ultrasound videos of several heart cross sections, which can be collected and integrated into an echocardiographic video set. The echocardiographic video set can include a large number of color ultrasound cardiac image videos of patients with heart diseases, including healthy patients and patients who actually suffer from heart diseases. These image videos should also cover various types of heart diseases, such as myocardial lesions, valvular diseases, etc.

[0049] Step S202, marking each frame of the echocardiogram in the echocardiogram video set, and dividing the echocardiogram into a training set, a test set, and a validation set. The marked content may include the lesion status, the position and contour of the cardiac chamber wall and the ventricular endocardium, etc.

[0050] Step S203, identifying the cardiac chamber wall and the ventricular endocardium in each frame of the echocardiogram video set. The video may be firstly subjected to a mixed image recognition algorithm, such as a star algorithm, adaptive directional filtering, Snake, graph matching, etc., to identify the chamber wall and the endocardium, etc., frame by frame, and then dynamically tracked according to the changes in the positions of the identified chamber wall and the endocardium and learned as features.

[0051] Step S204, dynamically tracking and extracting features according to the position changes of the cardiac chamber wall and the ventricular endocardium, to obtain a feature set corresponding to each frame of the echocardiogram.

[0052] Specifically, the object tracking technology in computer vision can be used to track the position changes of the heart chamber wall and the ventricular endocardium in the video sequence. At the same time, features can be extracted from these position changes, which may include myocardial motion amplitude, myocardial strain parameters, motion trajectory and other information.

[0053] Step S205, inputting the labeled echocardiogram images and the feature sets corresponding to the echocardiogram images into a deep learning model for feature learning, and obtaining a heart disease prediction model after feature learning.

[0054] Among them, deep learning models, such as convolutional neural networks, take the labeled echocardiogram and its corresponding feature set as input for feature learning and model training. The model will learn to extract valuable features from echocardiograms for automatic identification and prediction of heart diseases.

[0055] In this embodiment, a new computer vision and unsupervised machine learning model is used to automatically analyze the patient's cardiac ultrasound images, quantify the patient's myocardial motion characteristics, and use the machine learning model to calculate the patient's risk of cardiovascular disease. The heart disease prediction model can effectively improve the ability to diagnose heart disease in the early stage and provide more auxiliary information for clinicians.

[0056] In some optional embodiments, after obtaining the heart disease prediction result, the method further includes:

[0057] Obtain the patient's clinical information and the imaging parameters when collecting the echocardiogram to be tested.

[0058] Generate analysis reports based on patient clinical information, imaging parameters, and heart disease prediction results.

[0059] Specifically, the patient's clinical information may include the patient's identity information, the patient's medical history, and other examination results, etc. The imaging parameters may include ultrasound probe parameters, ultrasound parameters, ultrasound Doppler parameters, etc., which may be used to assist in evaluating the cardiac tissue structure, blood flow dynamics, etc.

[0060] Reference Figure 2 As shown, a complete heart disease diagnosis process can be as follows: the ultrasound doctor first takes an X-ray, and then the recorder or prediction system imports the patient's clinical information from the hospital medical record system, and the recorder or prediction system imports the taken echocardiogram images and image parameters from the ultrasound system, and then the prediction system automatically generates an analysis report containing the heart disease prediction results based on the patient's clinical information, image parameters and echocardiogram images and stores it in the hospital medical record system so that it can be called up and adjusted at any time in the hospital medical record system.

[0061] In this embodiment, the analysis report generated by the patient's clinical data and prediction results can provide doctors with clearer and more detailed auxiliary diagnosis information, which can help doctors better understand the patient's information and condition, while improving the quality of patient treatment and management.

[0062] In some optional implementations, marking each frame of the echocardiogram in the echocardiogram video set includes:

[0063] In case of lesions on echocardiogram, the lesion area is marked;

[0064] In the case that there is no lesion in the echocardiogram, the echocardiogram is marked as normal;

[0065] Mark the locations of the heart chamber walls and ventricular lining.

[0066] Specifically, before training the deep learning model, a large number of samples need to be labeled. For example, areas with lesions in echocardiograms need to be marked, and echocardiograms without lesions also need to be marked as normal images. Then, the position and contour of the chamber wall, the position and morphology of the ventricular endocardium, etc. in each frame of echocardiograms need to be marked. This labeling can be done manually or with automatic and semi-automatic tools to ensure accurate labeling information.

[0067] In this embodiment, each frame of the echocardiogram in the echocardiogram video set is marked, which can effectively improve the accuracy of model establishment.

[0068] In some optional embodiments, the heart disease prediction results also include: myocardial motion indicators corresponding to the disease type, wherein the myocardial motion indicators include: the movement of the heart chamber wall during heart contraction or relaxation, myocardial strain, and ejection fraction.

[0069] The movement of the heart chamber wall during heart contraction, including the contraction force and range of the myocardium. The movement of the heart chamber wall during heart diastole, including the functional state of the myocardium when it is relaxed. Myocardial strain can include the deformation of the heart chamber wall in the radial and axial directions of the heart, which can be used to evaluate the deformation capacity of the heart in various directions. The ejection fraction indicates the proportion of blood ejected from the ventricle during each heart contraction, which can effectively evaluate the heart's pumping function.

[0070] Furthermore, for cardiomyopathy or coronary heart disease, myocardial motion indicators also include: echocardiography showing asymmetric thickening of the ventricular septum or ventricular wall, smaller ventricular cavity, and possible increase (in non-obstructed areas) or decrease (in obstructed areas) in the amplitude of motion. Variability imaging can reveal local myocardial dysfunction, especially areas with increased myocardial stiffness, enlarged ventricular cavity, and generally weakened myocardial motion. Myocardial strain is used to measure the degree of deformation of myocardial fibers during cardiac contraction and relaxation to reflect the local functional state of the heart. It provides more detailed and sensitive information than traditional ejection fraction (EF), and can detect subtle changes in myocardial function at an early stage, which is of great value for the early diagnosis and prognosis of heart disease.

[0071] In this embodiment, in terms of auxiliary diagnosis of heart disease, not only can a heart disease risk prediction based on myocardial motion characteristics be given, but more accurate and detailed myocardial motion indicators can also be automatically given, avoiding the limitations of a single decision and making the final result more stable and reliable, thereby further assisting doctors in diagnosis and avoiding diagnostic errors caused by individual subjective factors and human errors of doctors.

[0072] In some optional embodiments, the disease types include: myocardial ischemia, coronary heart disease, hypertrophic cardiomyopathy, and myocardial amyloidosis.

[0073] In this embodiment, the provided heart disease prediction model can predict various types of diseases and has broad application prospects.

[0074] In this embodiment, a heart disease prediction device based on deep learning is also provided, which is used to implement the above embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0075] This embodiment provides a heart disease prediction device based on deep learning, such as Figure 3 As shown, including:

[0076] An acquisition module 201 is used to acquire the patient's to-be-detected echocardiogram, where the to-be-detected echocardiogram is acquired by a cardiac color doppler ultrasound instrument;

[0077] The prediction module 203 is used to input the to-be-detected echocardiogram image into a pre-established heart disease prediction model and output a heart disease prediction result, which includes: the disease type and the probability of the disease occurring within a preset time in the future.

[0078] In some optional embodiments, the device further comprises:

[0079] The training module is used to obtain a video set of echocardiograms of patients; mark each frame of the echocardiograms in the video set of echocardiograms, and divide the echocardiograms into a training set, a test set, and a validation set; identify the cardiac chamber walls and the ventricular endocardium in each frame of the echocardiograms in the video set of echocardiograms; dynamically track the position changes of the cardiac chamber walls and the ventricular endocardium, and perform feature extraction to obtain a feature set corresponding to each frame of the echocardiograms; input the marked echocardiograms and the feature sets corresponding to the echocardiograms into a deep learning model for feature learning to obtain a heart disease prediction model after feature learning.

[0080] The generation module is used to obtain the patient's clinical information and the image parameters when collecting the echocardiogram to be tested; based on the patient's clinical information, image parameters and heart disease prediction results, an analysis report is generated.

[0081] The marking module is used to mark the lesion area when there is a lesion in the echocardiogram; to mark the echocardiogram as a normal state when there is no lesion in the echocardiogram; and to mark the position of the heart chamber wall and the ventricular endocardium.

[0082] The deep learning-based heart disease prediction device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0083] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0084] The embodiment of the present invention also provides a computer device having the above Figure 3 The deep learning-based heart disease prediction device shown.

[0085] See also Figure 4 , Figure 4 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 4As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 4 A processor 10 is taken as an example.

[0086] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0087] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0088] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0089] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0090] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0091] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0092] A part of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the existence of the computer program instruction in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc., and accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium accessible to the computer.

[0093] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A heart disease prediction method based on deep learning, characterized in that: The method comprises: Acquire the patient's to-be-detected echocardiogram, wherein the to-be-detected echocardiogram is acquired by a cardiac color doppler ultrasound instrument; The to-be-detected echocardiogram image is input into a pre-established heart disease prediction model, and a heart disease prediction result is output, wherein the heart disease prediction result includes: the disease type and the probability of the disease occurring within a preset time in the future.

2. The method according to claim 1, characterized in that The heart disease prediction model is established by the following steps: Obtain the patient's cardiac ultrasound video set; Marking each frame of the echocardiogram in the echocardiogram video set, and dividing the echocardiogram into a training set, a test set, and a validation set; Identify the cardiac chamber wall and the ventricular endocardium in each frame of the echocardiogram video set; Dynamically track the position changes of the cardiac chamber wall and the ventricular endocardium, and extract features to obtain a feature set corresponding to each frame of echocardiogram; The labeled echocardiogram images and the feature set corresponding to the echocardiogram images are input into a deep learning model for feature learning to obtain a heart disease prediction model after feature learning.

3. The method according to claim 1, characterized in that After obtaining the heart disease prediction result, the method further includes: Acquiring clinical information of the patient and imaging parameters when collecting the echocardiogram to be detected; An analysis report is generated based on the patient's clinical information, the imaging parameters, and the heart disease prediction results.

4. The method according to claim 2, characterized in that: The step of marking each frame of the echocardiogram image in the echocardiogram video set includes: In case of lesions on echocardiogram, the lesion area is marked; In the case that there is no lesion in the echocardiogram, the echocardiogram is marked as normal; The locations of the heart chamber wall and the ventricular endothelium are marked.

5. The method according to claim 1, characterized in that The heart disease prediction result also includes: a myocardial motion index corresponding to the disease type, wherein the myocardial motion index includes: the motion of the heart chamber wall during heart contraction or relaxation, myocardial strain, and ejection fraction.

6. The method according to claim 1, characterized in that The disease types include: myocardial ischemia, coronary heart disease, hypertrophic cardiomyopathy, and myocardial amyloidosis.

7. A heart disease prediction device based on deep learning, characterized in that: The device comprises: An acquisition module is used to acquire the patient's to-be-detected echocardiogram image, wherein the to-be-detected echocardiogram image is acquired by a cardiac color doppler ultrasound instrument; The prediction module is used to input the to-be-detected echocardiogram into a pre-established heart disease prediction model and output a heart disease prediction result, wherein the heart disease prediction result includes: the disease type and the probability of the disease occurring within a preset time in the future.

8. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the deep learning-based heart disease prediction method according to any one of claims 1-6 by executing the computer instructions.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the deep learning-based heart disease prediction method described in any one of claims 1-6.

10. A computer program product, characterized in that It comprises computer instructions, which are used to enable a computer to execute the heart disease prediction method based on deep learning as described in any one of claims 1-6.

Citation Information

Cited By

  • Processing method and device for hypertrophic cardiomyopathy risk prediction model

    CN120565056A

  • Processing method and device of hypertrophic cardiomyopathy risk prediction model

    CN120565056B