Echocardiogram assisted report generation method, system, device and storage medium

Abstract

The embodiment of the application discloses an echocardiogram auxiliary report generation method, system, device and storage medium, and relates to the technical field of medical treatment. The echocardiogram auxiliary report generation method first aligns the input echocardiogram video through automatic registration, then identifies the starting time of each cardiac cycle in the echocardiogram video, generates the echocardiogram video frame of each cardiac cycle based on a preset frame number, divides the heart structure in the echocardiogram video frame into N target images of target regions by using a medical image segmentation model, obtains a structure parameter, inputs the structure parameter into an echocardiogram knowledge graph to automatically generate an auxiliary report, including an auxiliary diagnosis report and / or an auxiliary interpretation report. Therefore, doctors do not need to manually measure and fill in related clinical parameters or diagnosis opinions, and the health status of patients can be understood, and the working pressure of doctors is effectively reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical treatment, and in particular to an echocardiogram auxiliary report generation method, system, device and storage medium. BACKGROUND

[0002] At present, China is facing the dual pressures of population aging and the prevalence of metabolic risk factors, and the incidence and prevalence of cardiovascular diseases continue to increase, which has become the leading cause of death of residents. Heart disease mainly includes heart failure, arrhythmia, coronary heart disease, coronary artery abnormalities and the like, has high incidence and strong harmfulness, and has the characteristics of gradualness and suddenness, so regular examination of heart function is very important for heart health.

[0003] The current heart function examination means is usually ultrasonic examination. Ultrasonic examination is a medical imaging diagnostic technique based on ultrasonic waves. Through ultrasound, the size, structure and pathological lesions of muscles and internal organs can be directly observed. However, the ultrasonic examination in the related art usually needs to manually measure and fill in the relevant clinical parameters, and then manually input the diagnostic opinions according to the parameter values and the ultrasonic video, so as to generate a complete examination report, thereby increasing the workload of doctors. SUMMARY

[0004] The present application aims to at least solve one of the technical problems existing in the prior art. To this end, the present application provides an echocardiogram auxiliary report generation method, system, device and storage medium, which can automatically generate an auxiliary report based on an echocardiogram video, thereby effectively improving work efficiency.

[0005] In a first aspect, an echocardiogram auxiliary report generation method is provided, comprising:

[0006] automatically registering and aligning an input echocardiogram video; the echocardiogram video comprises at least one complete cardiac cycle;

[0007] identifying the start time of each cardiac cycle, and generating an ultrasonic video frame of each cardiac cycle based on a preset frame number;

[0008] based on the ultrasonic video frame, using a medical image segmentation model to segment the heart structure into a target image of N target regions, and obtaining a structure parameter based on each target image; wherein N is an integer greater than 1;

[0009] generating an auxiliary report based on the structure parameter and a pre-constructed echocardiogram knowledge graph; the auxiliary report comprises an auxiliary diagnosis report and / or an auxiliary interpretation report.

[0010] In some embodiments of the present application, the automatic registration and alignment of the input echocardiogram video further comprises:

[0011] detect key point information of a plurality of the ultrasound video frames;

[0012] select a standard section and a non-standard section in the echocardiogram video based on the key point information;

[0013] automatically register the non-standard section with the standard section to obtain an aligned echocardiogram video.

[0014] In some embodiments of the present application, based on the ultrasound video frames, the heart structure is segmented into N target regions by a medical image segmentation model to obtain a target image, and a structure parameter is obtained based on each target image, further comprising:

[0015] segmenting the heart structure in the ultrasound video frames by a medical image segmentation model to obtain the target image of N target regions;

[0016] performing image binarization on each target image to obtain a binarized image;

[0017] performing edge detection on the binarized image to obtain position information of each target region;

[0018] obtaining the structure parameter based on each position information.

[0019] In some embodiments of the present application, the structure parameter comprises at least one of the following: left ventricular end-diastolic volume, left ventricular end-systolic volume, interventricular septal thickness, left ventricular posterior wall thickness, right ventricular outflow tract, and ejection fraction, wherein the ejection fraction is calculated according to the left ventricular end-diastolic volume and the left ventricular end-systolic volume.

[0020] In some embodiments of the present application, the auxiliary report is generated based on the structure parameter and an echocardiogram knowledge graph, further comprising:

[0021] inputting the structure parameter into an echocardiogram knowledge graph to obtain a graph query path;

[0022] obtaining an auxiliary diagnosis conclusion based on the graph query path;

[0023] generating an auxiliary diagnosis report based on the auxiliary diagnosis conclusion, a preset diagnosis template, and the ultrasound video frames.

[0024] In some embodiments of the present application, further comprising:

[0025] obtaining an interpretation conclusion of the auxiliary diagnosis conclusion based on a preset diagnosis suggestion;

[0026] generating the auxiliary interpretation report according to the interpretation conclusion and a preset interpretation template.

[0027] In some embodiments of the present application, the structure parameters include at least one of the following: left ventricular end-diastolic volume, left ventricular end-systolic volume, interventricular septum thickness, left ventricular posterior wall thickness, right ventricular outflow tract, ejection fraction; the echocardiogram knowledge graph includes the structure parameters, target node information, and auxiliary diagnosis conclusion, the target node information is determined according to the structure parameters, and the auxiliary diagnosis conclusion is obtained according to the target node information and a preset diagnosis standard; before the auxiliary report is automatically generated based on the structure parameters and the echocardiogram knowledge graph, the method further includes:

[0028] constructing the echocardiogram knowledge graph, and the construction process includes:

[0029] if the left ventricular posterior wall thickness is greater than a first parameter, the target node information is left ventricular posterior wall thickening, and the auxiliary diagnosis conclusion obtained based on the target node information and a preset diagnosis standard is hypertensive heart disease or aortic valve stenosis;

[0030] if the left ventricular posterior wall thickness is less than a second parameter, the target node information is left ventricular posterior wall thinning, and the auxiliary diagnosis conclusion obtained based on the target node information and a preset diagnosis standard is dilated cardiomyopathy;

[0031] if the interventricular septum thickness is greater than a third parameter, the target node information is interventricular septum thickening, and the auxiliary diagnosis conclusion obtained based on the target node information and a preset diagnosis standard is coronary heart disease;

[0032] if the interventricular septum thickness is less than a fourth parameter, the target node information is interventricular septum thinning, and the auxiliary diagnosis conclusion obtained based on the target node information and a preset diagnosis standard is dilated cardiomyopathy;

[0033] if the right ventricular outflow tract is greater than a fifth parameter, the target node information is right ventricular outflow tract diameter too large, and the auxiliary diagnosis conclusion obtained based on the target node information and a preset diagnosis standard is right ventricular hypertrophy or right ventricular enlargement;

[0034] if the ejection fraction is greater than a sixth parameter, the target node information is left ventricular systolic function measurement too high, and the auxiliary diagnosis conclusion obtained based on the target node information and a preset diagnosis standard is ejection fraction pathologically too high;

[0035] if the interventricular septum thickness is less than a seventh parameter, the target node information is left ventricular systolic function measurement too low, and the auxiliary diagnosis conclusion obtained based on the target node information and a preset diagnosis standard is left ventricular systolic function reduction;

[0036] If the septal wall thickness is greater than the third parameter and the left ventricular posterior wall thickness is greater than the first parameter, the auxiliary diagnosis conclusion is hypertrophic cardiomyopathy based on the target node information and a preset diagnosis standard.

[0037] The atlas query path is generated according to the structure parameters, the target node information and the auxiliary diagnosis conclusion.

[0038] In a second aspect, the embodiments of the present application further provide an echocardiogram auxiliary report generation system, comprising:

[0039] An automatic registration module is configured to automatically register and align an input echocardiogram video, wherein the echocardiogram video comprises at least one complete cardiac cycle;

[0040] A preprocessing module is configured to identify a start time of each cardiac cycle and generate an echocardiogram frame of each cardiac cycle based on a preset frame number;

[0041] A segmentation and extraction module is configured to segment a heart structure into a target image of N target regions by using a medical image segmentation model and obtain structure parameters based on each target image, wherein N is an integer greater than 1.

[0042] A report generation module is configured to generate an auxiliary report based on the structure parameters and a pre-constructed echocardiogram knowledge atlas, wherein the auxiliary report comprises an auxiliary diagnosis report and / or an auxiliary interpretation report.

[0043] In a third aspect, the embodiments of the present application further provide an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the echocardiogram auxiliary report generation method according to the first aspect of the present application.

[0044] In a fourth aspect, the embodiments of the present application further provide a computer readable storage medium, wherein the storage medium stores a program, and the program is executed by a processor to implement the echocardiogram auxiliary report generation method according to the first aspect of the present application.

[0045] The embodiments of the present application at least have the following beneficial effects: the embodiments of the present application provide an echocardiogram assisted report generation method, system, device and storage medium, wherein the echocardiogram assisted report generation method first aligns the input echocardiogram video through automatic registration, then identifies the starting time of each cardiac cycle in the echocardiogram video, and generates the echocardiogram video frame of each cardiac cycle based on a preset frame number, segments the heart structure in the echocardiogram video frame into N target images of target regions by using a medical image segmentation model, obtains a structure parameter from the target images, and inputs the structure parameter into an echocardiogram knowledge graph to automatically generate an assisted report, including an assisted diagnosis report and / or an assisted interpretation report. Therefore, doctors do not need to manually measure and fill in related clinical parameters or diagnosis opinions, and patients can understand their own health status, thereby effectively reducing the work pressure of doctors.

[0046] Additional aspects and advantages of the application will be set forth in part in the description which follows, and in part will become apparent to those skilled in the art upon examination of the following and / or realization of the application. BRIEF DESCRIPTION OF DRAWINGS

[0047] The above and / or additional aspects and advantages of the present application will become apparent and be readily appreciated from the description of the embodiments, given in conjunction with the accompanying drawings, in which:

[0048] Figure 1 FIG. 1 is a flowchart of an echocardiogram assisted report generation method according to an embodiment of the present application;

[0049] Figure 2 FIG. 2 is a flowchart of step S101 in FIG. 1; Figure 1

[0050] Figure 3 FIG. 3 is a flowchart of step S103 in FIG. 1; Figure 1

[0051] Figure 4 FIG. 4 is a flowchart of step S104 in FIG. 1; Figure 1

[0052] Figure 5 FIG. 5 is an echocardiogram knowledge graph according to an embodiment of the present application;

[0053] Figure 6 FIG. 6 is an echocardiogram assisted diagnosis report according to an embodiment of the present application;

[0054] Figure 7 FIG. 7 is a flowchart after step S403 in FIG. 4; Figure 4

[0055] Figure 8 FIG. 8 is an echocardiogram assisted interpretation report according to an embodiment of the present application; and ​​​​

[0056] Figure 9 is a schematic diagram of an echocardiogram auxiliary report generation system module provided in an embodiment of the present application;

[0057] Figure 10 is a schematic diagram of an electronic device provided in an embodiment of the present application.

[0058] The reference signs: automatic registration module 100, preprocessing module 200, segmentation extraction module 300, report generation module 400, electronic device 1000, processor 1001, memory 1002. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application is further described in detail below in combination with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and cannot be used to limit the present application.

[0060] The embodiments of the present application are described in detail below, and examples of the embodiments are shown in the drawings, wherein the same or similar reference signs represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and are only used to explain the present application and cannot be understood as limiting the present application.

[0061] In the description of the present application, it should be understood that the orientation description, such as the orientation or position relationship indicated by up, down, front, back, left, right, etc. is based on the orientation or position relationship shown in the drawings, which is only for the convenience of describing the present application and simplifying the description, and does not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0062] In the description of the present application, the meaning of several is one or more, and the meaning of multiple is more than two. Greater than, less than, more than, etc. are understood as not including the number, and above, below, etc. are understood as including the number. If it is described as first, second, it is only used for the purpose of distinguishing technical features, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features or the sequence of indicated technical features.

[0063] In the description of the present application, unless otherwise explicitly limited, the words such as setting, installing, connecting, etc. should be broadly understood, and the person skilled in the art can reasonably determine the specific meaning of the above words in the present application in combination with the specific content of the technical scheme.

[0064] In order to better understand the technical scheme provided by the present application, the terms appearing in the text are explained as follows:

[0065] Cardiac cycle: refers to the process experienced by the cardiovascular system from the start of one heartbeat to the start of the next. When the heart relaxes, the internal pressure decreases, and blood flows back into the heart from the vena cava. When the heart contracts, the internal pressure increases, pumping blood to the arteries. Each time the heart contracts and relaxes, it constitutes a cardiac cycle. In a cardiac cycle, the two atria contract first, with the right atrium contracting slightly before the left atrium. After the atria begin to relax, the two ventricles contract, with the left ventricle contracting slightly before the right ventricle. In the late stage of ventricular relaxation, the atria begin to contract again.

[0066] Heart failure: referred to as heart failure, refers to the obstruction of the contraction and / or diastolic function of the heart, which cannot fully discharge the volume of blood from the body circulation into the heart, leading to the accumulation of blood in the venous system and insufficient perfusion of blood in the arterial system, resulting in a disease of cardiac circulation disorder.

[0067] Ejection fraction: refers to the percentage of stroke volume to ventricular end-diastolic volume (i.e. cardiac preload), with a normal value of 50-70%. It can be checked by cardiac ultrasound and is one of the important indicators for determining the type of heart failure.

[0068] At present, China is facing the dual pressures of population aging and the prevalence of metabolic risk factors, and the incidence and prevalence of cardiovascular diseases continue to increase, which has become the leading cause of death among residents. The latest "China Cardiovascular Health and Disease Report 2021" shows that in 2019, cardiovascular diseases accounted for 46.74% and 44.26% of deaths in rural and urban areas, respectively, and 2 out of every 5 deaths were due to cardiovascular diseases. In 2021, the number of people with cardiovascular diseases is expected to reach 330 million, of which more than 30 million people have heart-related diseases. Heart disease mainly includes heart failure, arrhythmia, coronary heart disease, and coronary artery abnormalities, with high incidence and strong harmfulness, and characteristics of gradualness and suddenness. Therefore, regular examination of heart function is very important for heart health.

[0069] Current heart function examination methods mainly include electrocardiogram, coronary CT, coronary angiography, and ultrasound examination. Among them, electrocardiogram can only be used to find some functional abnormalities, and cannot see the overall morphology, coronary CT and coronary angiography need to inject contrast medium, which is harmful to the human body and not suitable for regular examination. Therefore, non-invasive and high-accuracy ultrasound examination has become the focus of research. Ultrasound examination is a medical imaging diagnostic technique based on ultrasound waves, which can directly observe the size, structure and pathological lesions of muscles and internal organs. However, the current ultrasound examination usually requires the doctor to manually measure and fill in the relevant clinical parameters according to his own knowledge after obtaining the ultrasound video scanned by the machine, and then manually input the diagnosis opinion according to the parameter values and the ultrasound video to generate a complete examination report. Moreover, patients usually have difficulty in understanding the medical terms in the report. Therefore, in order to reduce the burden and work pressure of doctors, it is of great significance to develop an ultrasound echocardiogram video analysis system for automatically generating auxiliary reports.

[0070] Based on this, the embodiments of the present application provide an ultrasound echocardiogram auxiliary report generation method, system, device and storage medium, wherein the ultrasound echocardiogram auxiliary report generation method first aligns the input ultrasound echocardiogram video by automatic registration, then identifies the start time of each cardiac cycle in the ultrasound echocardiogram video, and generates ultrasound video frames of each cardiac cycle based on a preset frame number, and then separates the heart structure in the ultrasound video frames into N target regions by using a medical image segmentation model to obtain structure parameters, and then inputs the structure parameters into an ultrasound echocardiogram knowledge graph to automatically generate an auxiliary report, including an auxiliary diagnosis report and / or an auxiliary interpretation report. Thus, doctors do not need to manually measure and fill in relevant clinical parameters or diagnosis opinions, and patients can understand their own health status, effectively reducing the work pressure of doctors.

[0071] Referring to Figure 1 The embodiments of the present application provide an ultrasound echocardiogram auxiliary report generation method, including but not limited to the following steps S101 to S104.

[0072] Step S101, automatically registering and aligning the input ultrasound echocardiogram video.

[0073] In some embodiments, the ultrasound echocardiogram video includes at least one complete cardiac cycle, after reading in the data of the ultrasound echocardiogram video, the standard section is automatically found, and the non-standard section is automatically registered based on the standard section, so that the doctor does not need to manually move the ultrasound probe to find the standard section, and the aligned ultrasound echocardiogram video can be obtained.

[0074] Step S102, identifying the start time of each cardiac cycle, and generating ultrasound video frames of each cardiac cycle based on a preset frame number.

[0075] In some embodiments, the starting time of each cardiac cycle in the aligned echocardiogram video file is automatically identified, and ultrasound video frames of each cardiac cycle are generated based on a preset frame number. For example, the preset frame number can be 32, i.e., 32 ultrasound video frames are generated for each cardiac cycle in the video file. For another example, the preset frame number can be 64, i.e., 64 ultrasound video frames are generated for each cardiac cycle in the video file, which is not limited in the present embodiment.

[0076] In step S103, based on the ultrasound video frames, the cardiac structure is segmented into target images of N target regions by using a medical image segmentation model, and structure parameters are obtained based on each target image.

[0077] In some embodiments, the cardiac structure in the ultrasound video frames is segmented into target images of N target regions by using a medical image segmentation model, where N is an integer greater than 1. For example, when N is 9, the cardiac structure can be segmented into the following 9 target regions: left ventricular endocardium, left ventricular epicardium, right ventricular endocardium, right ventricular epicardium, tricuspid valve, pulmonary valve, mitral valve, aortic valve, and interventricular septum. For another example, when N is 10, the cardiac structure can be segmented into the following 10 target regions: right ventricular anterior wall, right ventricular, interventricular septum, left ventricular, left ventricular posterior wall, left ventricular outflow tract, aortic valve, aortic sinus, proximal ascending aorta, and left atrium, which is not limited in the present embodiment.

[0078] In some embodiments, after the cardiac structure is segmented into different target regions, the background region and the target regions are assigned values. The background region is assigned a value of 0, and the different target regions are respectively assigned values of integers from 1 to N. For example, when N is 9, the 9 target regions are respectively assigned values of integers from 1 to 9. Specifically, the left ventricular endocardium is assigned a value of 1, the left ventricular epicardium is assigned a value of 2, the right ventricular endocardium is assigned a value of 3, the right ventricular epicardium is assigned a value of 4, the tricuspid valve is assigned a value of 5, the pulmonary valve is assigned a value of 6, the mitral valve is assigned a value of 7, the aortic valve is assigned a value of 8, and the interventricular septum is assigned a value of 9.

[0079] It can be understood that after the cardiac structure in the ultrasound video frames is segmented into N target regions, there are N target images, and structure parameters corresponding to the target regions are obtained based on each target image.

[0080] In step S104, an auxiliary report is generated based on the structure parameters and a pre-constructed echocardiogram knowledge graph.

[0081] In some embodiments, the auxiliary report includes an auxiliary diagnosis report and an auxiliary interpretation report. Specifically, the structural parameters are input into a pre-constructed echocardiogram knowledge graph to automatically obtain an auxiliary diagnosis conclusion, thereby generating an auxiliary diagnosis report and an auxiliary interpretation report, assisting doctors in diagnosis, without the need for doctors to manually measure and fill in relevant clinical parameters or diagnosis opinions, and assisting patients in understanding their own health status, effectively reducing the work pressure of doctors.

[0082] Referring to Figure 2 As shown in the figure, in some embodiments of the present application, the above step S101 can further include but is not limited to the following steps S201 to S203.

[0083] Step S201, detecting key point information of the plurality of ultrasound video frames.

[0084] In some embodiments, the automatic registration process includes feature point detection, feature matching, and image transformation. First, the feature point detection detects the key point information of the ultrasound video frames. Specifically, the AKAZE algorithm is used to detect the key point information in the ultrasound video frame image.

[0085] Step S202, selecting a standard section and a non-standard section in the echocardiogram video based on the key point information.

[0086] In some embodiments, the feature matching is to match the key point information in the two images of the standard section and the non-standard section in the echocardiogram video using the BFMatcher matching algorithm of OpenCV, thereby selecting the standard section and the non-standard section in the echocardiogram video based on the key point information.

[0087] Step S203, automatically registering the non-standard section with the standard section to obtain an aligned echocardiogram video.

[0088] In some embodiments, the image transformation is to automatically register the non-standard section with the standard section. Illustratively, the findHomography function of OpenCV is used to calculate the Homographies matrix, and the warpPerspective function is used to transform the non-standard section in the echocardiogram video into a standard section based on the Homographies matrix, thereby obtaining an aligned echocardiogram video.

[0089] Referring to Figure 3 As shown in the figure, in some embodiments of the present application, the above step S103 can further include but is not limited to the following steps S301 to S304.

[0090] Step S301, segmenting the heart structure of the ultrasound video frame using a medical image segmentation model to obtain the target image of the N target regions.

[0091] In some embodiments, the heart structure of the ultrasound video frame is segmented into N target regions by using the medical image segmentation model to obtain a target image, where N is an integer greater than 1. For example, when N is 9, the heart structure can be segmented into the following 9 target regions: left ventricular endocardium, left ventricular epicardium, right ventricular endocardium, right ventricular epicardium, tricuspid valve, pulmonary valve, mitral valve, aortic valve, and interventricular septum, thereby obtaining a corresponding 9 target images; for another example, when N is 10, the heart structure can be segmented into the following 10 target regions: right ventricular anterior wall, right ventricular, interventricular septum, left ventricular, left ventricular posterior wall, left ventricular outflow tract, aortic valve, aortic sinus, proximal ascending aorta, and left atrium, thereby obtaining a corresponding 10 target images.

[0092] In step S302, image binarization is performed on each target image to obtain a binarized image.

[0093] It can be understood that image binarization is a process of setting the gray value of a pixel point on an image to 0 or 255, that is, the entire image presents a clear black and white effect, and the binarization of the image is beneficial to the further processing of the image, makes the image simple, and reduces the data amount, and highlights the outline of the target of interest. In some embodiments, a related function in the OpenCV tool library is used to perform image binarization on each target image to obtain a binarized image.

[0094] In step S303, edge detection is performed on the binarized image to obtain position information of each target region.

[0095] In some embodiments, edge detection is performed on the binarized image, and the position information of each target region is obtained by using an edge detection function. For example, the edge detection function is used to find the specific positions of the left atrium, right atrium, left ventricle, and right ventricle of the heart and the valves, and then a polygon fitting function is used to analyze the geometric data of the four cavities, including the long diameter and the transverse diameter, and finally the number of target pixel points is counted to estimate the area of each cavity. The size of the heart chamber and the cardiac function assessment are the basis of the echocardiography.

[0096] In step S304, a structure parameter is obtained based on each position information.

[0097] In some embodiments, the structure parameters include at least one of left ventricular end-diastolic volume (EDV), left ventricular end-systolic volume (ESV), interventricular septal thickness (IVST), left ventricular posterior wall thickness (LVPWT), right ventricular outflow tract (RVOT), and ejection fraction (EF), wherein the ejection fraction is calculated according to the left ventricular end-diastolic volume and the left ventricular end-systolic volume.

[0098] In some embodiments, the corresponding structure parameters can be obtained based on the position information of each target region. For example, for a heart failure diagnosis task, two structure parameters, i.e., left ventricular end-systolic volume and left ventricular end-diastolic volume, need to be measured. The cardiac cycle refers to the process experienced by the cardiovascular system from the start of one heartbeat to the start of the next heartbeat, including the systole and diastole. When the ventricle diastole, the blood fills the ventricle. When the ventricle diastole to the maximum, the blood volume filled in the left ventricle is called left ventricular end-diastolic volume. When the ventricle systole, the blood flows out of the ventricle. When the ventricle systole to the minimum, the blood volume remaining in the left ventricle is called left ventricular end-systolic volume. In the ultrasound video frame of the left ventricular endocardium segmentation, the images of left ventricular end-systole and left ventricular end-diastole are selected, and the area surrounded by the left ventricular endocardium is calculated based on the position information of the target region to be used as the left ventricular end-systolic volume and the left ventricular end-diastolic volume, thereby obtaining the two structure parameters required for the diagnosis task.

[0099] Referring to Figure 4 As shown in the above step S104, in some embodiments of the present application, the above step S104 can further include but is not limited to the following steps S401 to S403.

[0100] In step S401, the structure parameters are input into the echocardiogram knowledge graph to obtain a graph query path.

[0101] In some embodiments, the pre-constructed echocardiogram knowledge graph is stored in a Neo4j database, the extracted structure parameters are input into the echocardiogram knowledge graph, and the graph query path is obtained by graph Cypher query path.

[0102] In some embodiments, the construction of the echocardiographic knowledge graph is based on clinical diagnostic experience. It constructs an echocardiographic knowledge graph with entities representing cardiac-related clinical structural parameters, pathological structural and functional analyses, and cardiac disease diagnostic results. This is done according to preset diagnostic standards, such as the medical gold standard (which refers to the most reliable method for diagnosing diseases currently recognized in clinical medicine). The graph then maps the relationships between these entities. Figure 5 The echocardiography knowledge graph shown includes structural parameters, target node information, and auxiliary diagnostic conclusions. First, target node information is obtained based on the structural parameters. Then, auxiliary diagnostic conclusions are derived based on the target node information and preset diagnostic criteria. Its construction process includes:

[0103] If the thickness of the left ventricular posterior wall is greater than the first parameter, which is 11 mm, i.e., LVPWT>11 mm, then the target node information is left ventricular posterior wall thickening. Based on the target node information and the preset diagnostic criteria, the auxiliary diagnostic conclusion is hypertensive heart disease or aortic stenosis.

[0104] If the thickness of the left ventricular posterior wall is less than the second parameter, which is 6 mm, i.e., LVPWT < 6 mm, then the target node information is that the left ventricular posterior wall is thin. Based on the target node information and the preset diagnostic criteria, the auxiliary diagnostic conclusion is dilated cardiomyopathy.

[0105] If the interventricular septal thickness is greater than the third parameter, which is 11 mm (IVST>11 mm), then the target node information is interventricular septal thickening. Based on the target node information and the preset diagnostic criteria, the auxiliary diagnostic conclusion is coronary heart disease.

[0106] If the interventricular septal thickness is less than the fourth parameter, which is 6 mm (IVST < 6 mm), then the target node information indicates a thin interventricular septum. Based on the target node information and the preset diagnostic criteria, the auxiliary diagnostic conclusion is dilated cardiomyopathy.

[0107] If the right ventricular outflow tract is larger than the fifth parameter, which is 30 mm (i.e., RVOT>30 mm), then the target node information is that the inner diameter of the right ventricular outflow tract is too large. Based on the target node information and the preset diagnostic criteria, the auxiliary diagnostic conclusion is right ventricular hypertrophy or right ventricular enlargement.

[0108] If the ejection fraction is greater than the sixth parameter, which is 80%, i.e., EF>80%, then the target node information is that the left ventricular systolic function measurement value is too high. Based on the target node information and the preset diagnostic criteria, the auxiliary diagnostic conclusion is that the ejection fraction is pathologically high.

[0109] If the interventricular septal thickness is less than the seventh parameter, and the seventh parameter is 50%, i.e. EF < 50%, then the target node information is that the left ventricular systolic function measurement value is too low. Based on the target node information and the preset diagnostic criteria, the auxiliary diagnostic conclusion is that the left ventricular systolic function is reduced.

[0110] If the interventricular septal thickness is greater than the third parameter, and the left ventricular posterior wall thickness is greater than the first parameter, that is, IVST>11mm and LVPWT>11mm, the auxiliary diagnosis conclusion based on the target node information and the preset diagnosis standard is hypertrophic cardiomyopathy.

[0111] It can be understood that the echocardiogram knowledge graph in the above Figure 5 is only an example, and the echocardiogram knowledge graph can be expanded and refined based on the medical gold standard in actual application.

[0112] In addition, the atlas query path is generated according to the structure parameter, the target node information, and the auxiliary diagnosis conclusion. For example, referring to Figure 5 and Figure 6 , the EF calculated by measuring the two structure parameters of the patient's ESV and EDV is 45.3%, that is, EF<50%, the target node information based on the echocardiogram knowledge graph is left ventricular systolic function measurement is too low, and the auxiliary diagnosis conclusion based on the target node information and the preset diagnosis standard is left ventricular systolic function is reduced. Therefore, the atlas query path is generated, so as to automatically generate the diagnosis opinion in the auxiliary diagnosis report, including left ventricular systolic function is reduced.

[0113] Step S402, obtaining an auxiliary diagnosis conclusion based on the atlas query path.

[0114] In some embodiments, in the echocardiogram knowledge graph, according to the atlas query path, the auxiliary diagnosis conclusion corresponding to the structure parameter can be automatically obtained. For example, referring to the echocardiogram knowledge graph shown in Figure 5 , according to the left ventricular end-systolic volume ESV and the left ventricular end-diastolic volume EDV branch in the knowledge graph, the ejection fraction EF can be calculated. The ejection fraction refers to the percentage of each stroke output volume to the ventricular end-diastolic volume, and the normal value is 50-70%, which is one of the important indications for judging the type of heart failure. The formula for calculating the ejection fraction is: EF=(EDV-ESV)*100% / EDV. According to the knowledge graph, if EF<50%, the target node information of left ventricular systolic function measurement is too low is obtained, and the auxiliary diagnosis conclusion is left ventricular systolic function is reduced; if EF>80%, the target node information of left ventricular systolic function measurement is too high is obtained, and the auxiliary diagnosis conclusion is ejection fraction is pathologically too high.

[0115] Step S403, generating an auxiliary diagnosis report based on the auxiliary diagnosis conclusion, the preset diagnosis template, and the ultrasound video frame.

[0116] In some embodiments, referring to Figure 6The shown ultrasound auxiliary diagnosis report is generated based on the auxiliary diagnosis conclusion, the preset diagnosis template and the ultrasound video frame. The upper part of the preset diagnosis module is the personal information of the patient. The left side is the ultrasound image chart result area. The right side is the specific content of the auxiliary diagnosis report, including the numerical value of each structure parameter, the auxiliary diagnosis conclusion and the diagnosis opinion of the ultrasound findings, which effectively assist the doctor in diagnosis. It can be understood that the preset diagnosis template can be adjusted according to the needs. The present embodiment is only illustrative.

[0117] Specifically, referring to Figure 6 The shown auxiliary diagnosis report is an echocardiogram (cardiac ultrasound) of a 67-year-old patient. The left ventricular systolic function measured by ultrasound is: EF: 45.3%, the measured interval heart cycle variance is 0.23, no abnormality is found, and the wall motion index is: WMSI: 2.17. The diagnosis opinion obtained therefrom has the following three points: 1. Left ventricular systolic function is reduced, 2. Left ventricular enlargement, 3. Follow-up. It can be understood that the auxiliary diagnosis report is automatically generated based on the ultrasound results and is used for reference by the doctor.

[0118] In some embodiments of the present application, the echocardiogram is also removed from the image noise, and the image is enhanced to assist the doctor in subjective judgment. Specifically, the Real-ESRGAN model is used to perform super-resolution enhancement on the input echocardiogram video file. The enhancement effect is as follows: Figure 6 The ultrasound findings result area, wherein the left graph is the original graph, and the right graph is the enhanced result graph.

[0119] Referring to Figure 7 In some embodiments of the present application, after the above step S403, the following steps S501 to S502 can be included but are not limited thereto.

[0120] Step S501, obtaining the interpretation conclusion of the auxiliary diagnosis conclusion based on the preset diagnosis suggestion.

[0121] It can be understood that since there are many medical terms in the auxiliary diagnosis report, it is not convenient for the patient to understand his own health status, which increases the workload and work pressure of the doctor. Therefore, the echocardiogram auxiliary report generation method of the present application can obtain the interpretation conclusion of the auxiliary diagnosis conclusion based on the preset diagnosis suggestion, and interpret the auxiliary diagnosis report with medical terms.

[0122] Step S502, generating an auxiliary interpretation report according to the interpretation conclusion and the preset interpretation template.

[0123] In some embodiments, the auxiliary interpretation report is generated according to the interpretation conclusion and a preset interpretation module, specifically, the auxiliary interpretation report is generated by using a RIS diagnosis knowledge base-automatic report algorithm, wherein the top of the preset interpretation template is the brief personal information of the patient, including the name, age, gender and diagnosis report number, the main body is the detailed explanation and suggestion of each diagnosis opinion in the auxiliary diagnosis report, including medical term explanation, etiology, adverse effects and clinical suggestions, which facilitates the patient to preliminarily understand the diagnosis result. It can be understood that the preset interpretation template can be adjusted according to the needs, and the embodiment is only illustrative.

[0124] It can be understood that, Figure 8 The auxiliary interpretation report is for Figure 6 The three diagnosis opinions in the auxiliary diagnosis report are fully explained, for example, for the first point of left ventricular systolic dysfunction, the auxiliary interpretation report shows that left ventricular systolic dysfunction usually refers to the decrease of cardiac ejection volume, and the left ventricular ejection fraction of cardiac color Doppler ultrasound is a clinical index for measuring left ventricular systolic function. If the ejection fraction is less than 50%, it means that the systolic function is decreased. Acute myocardial infarction, long-term poor control of hypertension, various types of cardiomyopathy such as hypertrophic cardiomyopathy, restrictive cardiomyopathy and dilated cardiomyopathy, acute severe myocarditis can also cause myocardial damage and significant decrease of systolic function in a short period of time. If you have a history of related diseases, please do not panic, and go to a doctor in time and tell the doctor the truth. It is helpful for the patient to understand his own health status and test results.

[0125] The echocardiogram auxiliary report generation method first automatically registers and aligns the input echocardiogram video, then identifies the start time of each cardiac cycle in the echocardiogram video, and generates an echocardiogram video frame for each cardiac cycle based on a preset frame number, and then uses a medical image segmentation model to segment the heart structure in the echocardiogram video frame into N target regions to obtain a target image, and obtains a structure parameter from the target image, and inputs the structure parameter into an echocardiogram knowledge graph to automatically generate an auxiliary diagnosis report and / or an auxiliary interpretation report. Therefore, doctors do not need to manually measure and fill in related clinical parameters or diagnosis opinions, and patients can understand their own health status, which effectively reduces the work pressure of doctors.

[0126] The echocardiogram auxiliary report generation system provided in the embodiments of the present application can realize the echocardiogram auxiliary report generation method described above, and refer to Figure 9 As shown in the figure, in some embodiments of the present application, the echocardiogram auxiliary report generation system comprises:

[0127] An automatic registration module 100 is configured to automatically register and align the input echocardiogram video; the echocardiogram video comprises at least one complete cardiac cycle;

[0128] The preprocessing module 200 is configured to identify the start time of each cardiac cycle and generate ultrasound video frames of each cardiac cycle based on a preset frame number.

[0129] The segmentation extraction module 300 is configured to segment the heart structure into a target image of N target regions by using a medical image segmentation model, and obtain a structure parameter based on each target image; wherein N is an integer greater than 1.

[0130] The report generation module 400 is configured to generate an auxiliary report based on the structure parameter and a pre-constructed echocardiogram knowledge graph; the auxiliary report includes an auxiliary diagnosis report and / or an auxiliary interpretation report.

[0131] The following will take the heart failure diagnosis task as an example to specifically illustrate the structure and use process of the echocardiogram auxiliary report generation system of the present application. It should be understood that the input and output format standards, specific clinical parameters, segmentation models, parameter extraction algorithms, etc. involved in the present embodiment are specific methods implemented for the heart failure diagnosis embodiment, and are not used for limitation.

[0132] Firstly, the echocardiogram auxiliary report generation system receives a user uploaded echocardiogram video file containing a complete cardiac cycle apical four-chamber standard section data, the system decodes the video file and automatically intercepts the standard section, automatically registers the non-standard section through the automatic registration module 100, and then stores the aligned echocardiogram video as a set of continuous ultrasound video frames through the preprocessing module 200. For the heart failure diagnosis task, the left ventricular endocardium needs to be segmented to measure the left ventricular end-diastolic volume (EDV) and end-systolic volume (ESV). Through the segmentation extraction module 300, the left ventricular endocardium dynamic segmentation model EchoNet model trained on the apical four-chamber standard section echocardiogram dataset EchoNet-Dynamic is used to input the continuous ultrasound video frames into the EchoNet model to complete the forward propagation, and the left ventricular endocardium segmentation result is obtained.

[0133] It can be understood that, for the heart failure diagnosis task, mainly by calculating the ejection fraction, it is judged whether there is a heart failure problem according to the ejection fraction. According to the end-systolic volume (ESV) and end-diastolic volume (EDV) branches in the echocardiogram knowledge graph, the ejection fraction (EF) can be calculated, so it is necessary to measure the two important clinical parameters of ESV and EDV. When the ventricle diastoles, the blood fills the ventricle, and when the ventricle diastoles to the maximum, the left ventricle fills the blood volume, which is called end-diastolic volume (EDV). When the ventricle contracts, the blood flows out of the ventricle, and when the ventricle contracts to the minimum, the left ventricle residual blood volume is called end-systolic volume (ESV). On the left ventricular endocardium segmentation result, the pictures of end-systole and end-diastole in the ultrasound video frame are selected, and the area surrounded by the left ventricular endocardium is calculated as the end-systolic volume and the end-diastolic volume.

[0134] The ejection fraction refers to the percentage of each stroke output volume to the ventricular end-diastolic volume, and the normal value is 50%-70%, which is one of the important indications for judging the type of heart failure. The formula for calculating the ejection fraction is: EF=(EDV-ESV)*100% / EDV. If EF<50%, the target node information of left ventricular systolic function measurement is too low, and the auxiliary diagnosis conclusion is left ventricular systolic dysfunction; if EF>80%, the target node information of left ventricular systolic function measurement is too high, and the auxiliary diagnosis conclusion is ejection fraction pathologically high. Finally, through the report generation module 400, the system automatically generates an ultrasound auxiliary diagnosis report, and then uses the RIS diagnosis knowledge base-automatic report generation algorithm to automatically generate an ultrasound auxiliary interpretation report. For each diagnosis opinion in the auxiliary diagnosis report, detailed explanations and suggestions are given, including medical term explanation, etiology, adverse effects, and clinical suggestions, which facilitates patients to preliminarily understand their own health status and ultrasound reports.

[0135] The specific implementation of the echocardiogram auxiliary report generation system of the embodiment is basically the same as the specific implementation of the echocardiogram auxiliary report generation method described above, and will not be repeated here.

[0136] Figure 10 An electronic device 1000 provided by an embodiment of the application is shown. The electronic device 1000 includes a processor 1001, a memory 1002, and a computer program stored in the memory 1002 and executable on the processor 1001, and the computer program is used to execute the echocardiogram auxiliary report generation method described above when running.

[0137] The processor 1001 and the memory 1002 can be connected by a bus or other means.

[0138] The memory 1002, as a kind of non-transient computer readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs, such as the echocardiogram assisted report generation method described in the embodiments of the present application.The processor 1001 realizes the echocardiogram assisted report generation method described above by running the non-transient software programs and instructions stored in the memory 1002.

[0139] The memory 1002 can include a program storage area and a data storage area, wherein the program storage area can store an operating system and at least one application required by a function; the data storage area can store the echocardiogram assisted report generation method described above. In addition, the memory 1002 can include a high-speed random access memory 1002, and can also include a non-transient memory 1002, such as at least one storage device, a flash memory device or other non-transient solid-state storage device. In some embodiments, the memory 1002 can optionally include a memory 1002 remotely disposed relative to the processor 1001, and these remote memories 1002 can be connected to the electronic device 1000 through 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.

[0140] The non-transient software programs and instructions required to implement the echocardiogram assisted report generation method described above are stored in the memory 1002, and when executed by one or more processors 1001, the echocardiogram assisted report generation method described above is executed, for example, the method steps S101 to S104 in Figure 1 , the method steps S201 to S203 in Figure 2 , the method steps S301 to S304 in Figure 3 , the method steps S401 to S403 in Figure 4 , the method steps S501 to S502 in Figure 7 .

[0141] The embodiments of the present application also provide a storage medium, which is a computer readable storage medium, and the storage medium stores a computer program, and the computer program is executed by a processor to implement the echocardiogram assisted report generation method described above. The memory, as a kind of non-transient computer readable storage medium, can be used to store non-transient software programs and non-transient computer executable programs. In addition, the memory can include a high-speed random access memory, and can 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 embodiments, the memory can optionally include a memory remotely disposed relative to the processor, and these remote memories can be connected to the processor through 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.

[0142] The echocardiogram assisted report generation method, system, device and storage medium provided by the embodiments of the present application, wherein the echocardiogram assisted report generation method first automatically registers and aligns the input echocardiogram video, then identifies the starting time of each cardiac cycle in the echocardiogram video, and generates the echocardiogram video frames of each cardiac cycle based on a preset frame number, segments the heart structure in the echocardiogram video frames into N target images by using a medical image segmentation model, and obtains the structure parameters therefrom, and then inputs the structure parameters into the echocardiogram knowledge graph to automatically generate an auxiliary diagnosis report and / or an auxiliary interpretation report. Thus, the processing of the echocardiogram video, the measurement and extraction of the clinical index structure parameters are automatically completed, and the comprehensive analysis of the echocardiogram video is completed according to the echocardiogram knowledge graph to generate a complete echocardiogram assisted report. The doctor does not need to manually measure and fill in the related clinical parameters or diagnosis opinions, and the patient can understand his own health status, which effectively reduces the work pressure of the doctor.

[0143] The above-described embodiments are merely illustrative for describing the present application and the units described as separate parts can or can not be physically separate, that is, can be located in one place or distributed on multiple network units. Part or all of the modules can be selected as needed to achieve the purpose of the embodiments.

[0144] Those of ordinary skill in the art can understand that all or some of the steps in the above disclosed method and system can be implemented as software, firmware, hardware and appropriate combinations thereof. Some or all of the physical components can be implemented as software executed by a processor such as a central processing unit, a digital signal processor or a microprocessor, or as hardware, or as an integrated circuit such as an application specific integrated circuit. Such software can be distributed on a computer readable medium, which can include computer storage media (or non-transitory media) and communication media (or transitory media). As known to those of ordinary skill in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tapes, storage device storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, as known to those of ordinary skill in the art, communication media typically includes computer readable instructions, data structures, program modules or other data in modulated data signals such as carrier waves or other transport mechanisms, and can include any information delivery medium.

[0145] It should also be appreciated that various embodiments provided by the present application can be combined in any manner to achieve different technical effects. The above is a specific description of the preferred embodiments of the present application, but the present application is not limited to the above embodiments, and those skilled in the art can make various equivalent modifications or replacements without departing from the spirit of the present application.

Claims

1. An echocardiography assisted report generation method, characterized by, The method comprises: automatically registering and aligning input echocardiogram videos; the echocardiogram videos comprise at least one complete cardiac cycle; identifying the start time of each cardiac cycle, and generating echocardiogram frames of each cardiac cycle based on a preset frame number; based on the echocardiogram frames, segmenting the cardiac structure into target images of N target regions by using a medical image segmentation model, and obtaining structural parameters based on each target image; wherein N is an integer greater than 1; generating an auxiliary report based on the structural parameters and a pre-constructed echocardiogram knowledge graph; the auxiliary report comprises an auxiliary diagnosis report and / or an auxiliary interpretation report; The automatic registration and alignment of the input echocardiogram videos further comprises: detecting key point information of a plurality of echocardiogram frames; selecting standard sections and non-standard sections in the echocardiogram videos based on the key point information; automatically registering the non-standard sections with the standard sections to obtain aligned echocardiogram videos; The method of generating an auxiliary report based on the structural parameters and an echocardiogram knowledge graph further comprises: inputting the structural parameters into the echocardiogram knowledge graph to obtain a graph query path; obtaining an auxiliary diagnosis conclusion based on the graph query path; generating an auxiliary diagnosis report based on the auxiliary diagnosis conclusion, a preset diagnosis template, and the echocardiogram frames, and the auxiliary interpretation report comprises detailed explanations and suggestions for each diagnosis opinion.

2. The echocardiography assisted report generation method of claim 1, wherein, The method of segmenting the cardiac structure into target images of N target regions based on the echocardiogram frames and using a medical image segmentation model further comprises: segmenting the cardiac structure in the echocardiogram frames by using a medical image segmentation model to obtain the target images of the N target regions; performing image binarization on each target image to obtain a binarized image; performing edge detection on the binarized image to obtain position information of each target region; obtaining the structural parameters based on each position information.

3. The echocardiography assisted report generation method of any of claims 1 to 2, wherein, The structural parameters comprise at least one of left ventricular end-diastolic volume, left ventricular end-systolic volume, interventricular septal thickness, left ventricular posterior wall thickness, right ventricular outflow tract, and ejection fraction, wherein the ejection fraction is calculated based on the left ventricular end-diastolic volume and the left ventricular end-systolic volume.

4. The echocardiography assisted report generation method of claim 1, wherein, The method further comprises: obtaining an interpretation conclusion of the auxiliary diagnosis conclusion based on a preset diagnosis suggestion; generating the auxiliary interpretation report according to the interpretation conclusion and a preset interpretation template.

5. The echocardiography assisted report generation method of claim 1, wherein, The structural parameters comprise at least one of left ventricular end-diastolic volume, left ventricular end-systolic volume, interventricular septal thickness, left ventricular posterior wall thickness, right ventricular outflow tract, and ejection fraction; the echocardiogram knowledge graph comprises the structural parameters, target node information, and an auxiliary diagnosis conclusion, the target node information is determined based on the structural parameters, and the auxiliary diagnosis conclusion is obtained based on the target node information and a preset diagnosis standard; Before automatically generating the auxiliary report based on the structural parameters and the pre-constructed echocardiogram knowledge graph, the method further comprises: constructing the echocardiogram knowledge graph, the construction process comprising: If the left ventricular posterior wall thickness is greater than a first parameter, the target node information is left ventricular posterior wall thickening, and based on the target node information and a preset diagnostic standard, the auxiliary diagnostic conclusion is hypertension heart disease or aortic valve stenosis; If the left ventricular posterior wall thickness is less than a second parameter, the target node information is left ventricular posterior wall thinning, and based on the target node information and a preset diagnostic standard, the auxiliary diagnostic conclusion is dilated cardiomyopathy; If the interventricular septum thickness is greater than a third parameter, the target node information is interventricular septum thickening, and based on the target node information and a preset diagnostic standard, the auxiliary diagnostic conclusion is coronary heart disease; If the interventricular septum thickness is less than a fourth parameter, the target node information is interventricular septum thinning, and based on the target node information and a preset diagnostic standard, the auxiliary diagnostic conclusion is dilated cardiomyopathy; If the right ventricular outflow tract is greater than a fifth parameter, the target node information is right ventricular outflow tract diameter too large, and based on the target node information and a preset diagnostic standard, the auxiliary diagnostic conclusion is right ventricular hypertrophy or right ventricular enlargement; If the ejection fraction is greater than a sixth parameter, the target node information is left ventricular systolic function measurement too high, and based on the target node information and a preset diagnostic standard, the auxiliary diagnostic conclusion is ejection fraction pathologically too high; If the interventricular septum thickness is less than a seventh parameter, the target node information is left ventricular systolic function measurement too low, and based on the target node information and a preset diagnostic standard, the auxiliary diagnostic conclusion is left ventricular systolic function reduction; If the interventricular septum thickness is greater than the third parameter, and the left ventricular posterior wall thickness is greater than the first parameter, based on the target node information and a preset diagnostic standard, the auxiliary diagnostic conclusion is hypertrophic cardiomyopathy; The graph query path is generated according to the structure parameters, the target node information and the auxiliary diagnostic conclusion.

6. An echocardiogram assisted report generation system, characterized by, Comprise: An automatic registration module for automatically registering and aligning an input echocardiogram video; the echocardiogram video comprises at least one complete cardiac cycle; A preprocessing module for identifying the start time of each cardiac cycle, and generating an echocardiogram frame of each cardiac cycle based on a preset frame number; A segmentation and extraction module for segmenting the heart structure into a target image of N target regions using a medical image segmentation model, and obtaining structure parameters based on each target image; Wherein N is an integer greater than 1; A report generation module for generating an auxiliary report based on the structure parameters and a pre-constructed echocardiogram knowledge graph; The auxiliary report includes: an auxiliary diagnosis report and / or an auxiliary interpretation report; The automatic registration and alignment of the input echocardiogram video further comprises: Detecting key point information of a plurality of echocardiogram frames; Selecting a standard section and a non-standard section in the echocardiogram video based on the key point information; Using the standard section to automatically register the non-standard section to obtain an aligned echocardiogram video; The auxiliary report generated based on the structure parameters and the echocardiogram knowledge graph further comprises: Input the structure parameter into an echocardiogram knowledge graph to obtain a graph query path; Obtain an auxiliary diagnosis conclusion based on the graph query path; Generate an auxiliary diagnosis report based on the auxiliary diagnosis conclusion, a preset diagnosis template, and the echocardiogram video frame.

7. An electronic device, comprising: The memory stores a computer program, and the processor implements the echocardiogram auxiliary report generation method in any one of claims 1 to 5 when executing the computer program.

8. A computer-readable storage medium, characterized in that, The storage medium stores a program, and the program is executed by the processor to implement the echocardiogram auxiliary report generation method in any one of claims 1 to 5.

Images