Acute myocardial infarction prediction method and system

Abstract

The invention relates to an acute myocardial infarction prediction method and system, and the method comprises the steps: obtaining a to-be-processed coronary artery CTA image; performing image segmentation on the coronary artery CTA image to obtain a target segmentation image; wherein the target segmented image comprises a plurality of coronary artery anatomical structures; performing image characteristic analysis on the target segmented image to obtain characteristic parameters for representing coronary atherosclerotic plaque characteristics; performing regression prediction by using the characteristic parameters to obtain the relevancy between the characteristic parameters and acute myocardial infarction; and based on the correlation degree, predicting the probability of acute myocardial infarction in a future preset time. Compared with the prior art, the method has the advantages of high prediction accuracy, high reliability and the like.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to an acute myocardial infarction prediction method and system. Background Art

[0002] Coronary artery disease (CAD), also known as coronary atherosclerotic heart disease, is a metabolic disorder caused by the combined effects of genes and the environment. Acute myocardial infarction (AMI) is often caused by the rupture of non-obstructive atherosclerotic plaques in the coronary arteries, which causes coronary artery lumen stenosis or occlusion, leading to myocardial ischemia, hypoxia and necrosis. Since there are no early signs, how to more accurately assess the risk of AMI patients and predict the probability of AMI in patients in the later stage is a difficult problem that needs to be solved in clinical practice.

[0003] Currently, the existing schemes for predicting AMI are based on traditional stratification tools such as high-risk plaque characteristics and basic clinical features. However, due to the complexity of AMI and the inevitable disadvantages of traditional stratification tools, such as high-risk plaque characteristics reflecting the morphological characteristics of atherosclerotic plaques and the basic clinical characteristics information being too single, traditional stratification tools are not suitable for the prediction of AMI, resulting in low accuracy of the prediction results obtained by existing AMI prediction devices. Summary of the invention

[0004] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a method and system for predicting acute myocardial infarction with high prediction accuracy and strong reliability.

[0005] The purpose of the present invention can be achieved by the following technical solutions:

[0006] A method for predicting acute myocardial infarction comprises: obtaining a coronary artery CTA image to be processed; performing image segmentation on the coronary artery CTA image to obtain a target segmentation image; wherein the target segmentation image contains a plurality of coronary artery anatomical structures; performing image characteristic analysis on the target segmentation image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques; performing regression prediction using the characteristic parameters to obtain the correlation between the characteristic parameters and the occurrence of acute myocardial infarction; and based on the correlation, predicting the probability of acute myocardial infarction occurring within a preset time in the future.

[0007] Preferably, the acquiring of the target segmentation image comprises: acquiring a target segmentation image including at least one segmentation image of a right coronary artery segmentation image, a posterior descending branch segmentation image, a left ventricular posterior branch segmentation image, a left main trunk segmentation image, a left anterior descending branch segmentation image, a left circumflex branch segmentation image, a diagonal branch segmentation image, an intermediate branch segmentation image and an obtuse marginal branch segmentation image.

[0008] Preferably, the image characteristic analysis of the target segmented image includes: determining the positions of the vascular intima and the vascular adventitia in the target segmented image; determining a region of interest of coronary atherosclerotic plaques in the target segmented image based on the positions of the vascular intima and the vascular adventitia; and performing image characteristic analysis of the target segmented image based on the region of interest.

[0009] More preferably, the characteristic parameters include high-risk plaque characteristics; the high-risk plaque characteristics include: at least one of low-density plaque, napkin ring sign, positive reconstruction and punctate calcification;

[0010] The image characteristic analysis of the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques includes: judging whether the target segmented image has the low-density plaque based on the CT value of the target segmented image; and / or judging whether the target segmented image has the napkin ring sign based on the CT value of the target segmented image; and / or judging whether the target segmented image has the positive reconstruction by determining the maximum blood vessel diameter of the lesion segment and the average diameter of the reference blood vessels at the proximal and distal ends of the lesion based on the target segmented image; and / or judging whether the target segmented image has the point calcification based on the section diameter of the calcification inside the plaque in the target segmented image.

[0011] More preferably, the characteristic parameters include stenosis degree and plaque load;

[0012] The image characteristic analysis of the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques includes: determining the minimum vessel lumen diameter of the lesion segment and the average diameter of the reference vessel lumens at the proximal and distal ends of the lesion based on the target segmented image, and determining the stenosis degree of the target segmented image; determining the minimum vessel lumen area of ​​the lesion segment and the average area of ​​the reference vessel lumens at the proximal and distal ends of the lesion based on the target segmented image, and determining the plaque load of the target segmented image.

[0013] More preferably, the characteristic parameter includes lesion length;

[0014] The performing of image characteristic analysis on the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques includes: determining the length from the proximal end of the lesion to the distal end of the lesion based on the target segmented image, and determining the lesion length of the target segmented image.

[0015] More preferably, the characteristic parameter includes coronary perivascular fat attenuation index;

[0016] The performing of image characteristic analysis on the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques includes: based on the target segmented image, according to a preset pericoronary fat CT threshold, with the vascular adventitia as the inner boundary and the length perpendicular to the vascular lumen diameter as the outer boundary, determining the pericoronary fat attenuation index of the target segmented image.

[0017] More preferably, the characteristic parameters include: parameters of influencing factors affecting the measurement of coronary atherosclerotic plaques;

[0018] The method further includes: classifying the influencing factors to obtain quantitative indicators and counting indicators; comparing the quantitative indicators and the counting indicators between groups to determine the correlation between the quantitative indicators and the counting indicators and the occurrence of acute myocardial infarction; if the correlation is less than a preset critical value, determining a target threshold value of the characteristic parameter; wherein the preset critical value is used to characterize the sensitivity of the characteristic parameter; based on the target threshold value and the correlation, selecting a target characteristic parameter from a plurality of the characteristic parameters;

[0019] The predicting of the probability of acute myocardial infarction occurring within a future preset time based on the correlation includes: predicting the probability of acute myocardial infarction occurring within a future preset time based on the correlation between the target characteristic parameter and the occurrence of acute myocardial infarction.

[0020] Preferably, the using the characteristic parameters to perform regression prediction to obtain the correlation between the characteristic parameters and the occurrence of acute myocardial infarction includes:

[0021] Using a target regression prediction model to perform regression prediction on the characteristic parameters to obtain the correlation between the characteristic parameters and the occurrence of acute myocardial infarction; wherein the target regression prediction model is a regression prediction model with the largest statistic selected from multiple candidate regression prediction models;

[0022] The expression of the target regression prediction model is:

[0023] h=1-h0(t) expβ

[0024] Among them, h represents the correlation between the characteristic parameter and the occurrence of acute myocardial infarction within the future preset time t; h0(t) represents the average survival probability within the future preset time t; β represents the prognostic index, which is the sum of the product of the characteristic parameter and the corresponding preset parameter coefficient.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] High prediction accuracy and strong reliability: The acute myocardial infarction prediction method provided by the present invention obtains characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques from multiple dimensions, thereby obtaining the correlation between each characteristic parameter and acute myocardial infarction. On this basis, combined with the basic clinical information, the probability of acute myocardial infarction occurring at a preset time in the future is predicted, thereby obtaining a more accurate prediction probability, thereby improving the prediction accuracy of the acute myocardial infarction prediction method based on coronary artery CTA imaging for the prediction probability of acute myocardial infarction occurring at a preset time in the future. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 A schematic diagram of a flow chart of a method for predicting acute myocardial infarction provided by an embodiment of the present invention;

[0028] Figure 2 A schematic diagram showing the comparison of the area under the curve of the regression prediction model provided in the embodiment of the present application;

[0029] Figure 3 A coronary artery CTA image of a patient in the application case provided in the embodiment of the present application;

[0030] Figure 4 A schematic diagram of the structure of an acute myocardial infarction prediction system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0031] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions 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 ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0032] like Figure 1 As shown, a method for predicting acute myocardial infarction comprises:

[0033] S110: Acquire a coronary artery CTA image to be processed;

[0034] S120: performing image segmentation on the coronary artery CTA image to obtain a target segmented image; wherein the target segmented image includes a plurality of coronary artery anatomical structures;

[0035] S130: performing image characteristic analysis on the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques;

[0036] S140: performing regression prediction using the characteristic parameters to obtain the correlation between the characteristic parameters and the occurrence of acute myocardial infarction;

[0037] S150: Based on the correlation, predict the probability of acute myocardial infarction occurring within a preset time in the future.

[0038] The coronary CTA (coronary CT angiography, CTA) image in the above step S110 is a non-invasive imaging method for evaluating coronary arteries, which can not only be used to evaluate the anatomical structure and pathological conditions of the coronary arteries, but also to analyze relevant parameters of structures such as pericoronary adipose tissue (PCAT).

[0039] The above step S120 may use a more mature image segmentation technology in the related art to perform image segmentation on the coronary artery CTA image. The acquired target segmented image may include the anatomical structures of multiple coronary arteries, and the target segmented image may include: at least one segmented image of the right coronary artery segmentation image, the posterior descending branch segmentation image, the left posterior ventricular branch segmentation image, the left main trunk segmentation image, the left anterior descending branch segmentation image, the left circumflex branch segmentation image, the diagonal branch segmentation image, the intermediate branch segmentation image and the obtuse marginal branch segmentation image.

[0040] The above-mentioned step S130 includes: determining the positions of the vascular intima and the vascular adventitia in the target segmented image; determining the region of interest of the coronary atherosclerotic plaque in the target segmented image based on the positions of the vascular intima and the vascular adventitia; and performing image characteristic analysis on the target segmented image based on the region of interest.

[0041] It is understandable that the more mature technology in the art can be used to obtain the region of interest. For its specific implementation method, please refer to the relevant technology, and the embodiments of this application will not be repeated.

[0042] The following describes the scheme of performing image characteristic analysis on the image in step S130 from the perspective of characteristic parameters:

[0043] (1) Characteristic parameters include high-risk plaque characteristics;

[0044] High-risk plaque characteristics include at least one of the following: low-density plaque, napkin ring sign, positive remodeling, and punctate calcification;

[0045] At this time, the above step S130 includes:

[0046] Based on the CT value of the target segmentation image, determine whether there is a low-density plaque in the target segmentation image;

[0047] and / or, judging whether the target segmentation image has a napkin ring sign based on the CT value of the target segmentation image;

[0048] and / or, based on the target segmented image, determining the maximum blood vessel diameter of the lesion segment and the average diameter of the reference blood vessels proximal and distal to the lesion, and judging whether there is positive reconstruction in the target segmented image;

[0049] And / or, based on the diameter of the section of the calcification inside the plaque in the target segmented image, it is determined whether there is point-like calcification in the target segmented image.

[0050] (2) characteristic parameters include stenosis degree and plaque burden;

[0051] At this time, the above step S130 includes:

[0052] Based on the target segmentation image, the minimum vessel lumen diameter of the lesion segment and the average vessel lumen diameter of the reference vessel proximal and distal to the lesion are determined to determine the stenosis degree of the target segmentation image;

[0053] Based on the target segmentation image, the minimum vessel lumen area of ​​the lesion segment and the average lumen area of ​​the reference vessels proximal and distal to the lesion are determined to determine the plaque burden of the target segmentation image.

[0054] (3) Characteristic parameters include lesion length;

[0055] At this time, the above step S130 includes:

[0056] Based on the target segmented image, the length from the proximal end of the lesion to the distal end of the lesion is determined, and the lesion length of the target segmented image is determined.

[0057] (4) Characteristic parameters include coronary perivascular fat attenuation index;

[0058] At this time, the above step S130 includes:

[0059] Based on the target segmented image, according to a preset pericoronary fat CT threshold, with the adventitia as the inner boundary and the length perpendicular to the vascular lumen diameter as the outer boundary, the pericoronary fat attenuation index of the target segmented image is determined.

[0060] Pericoronary adipose tissue (PCAT) is a special visceral fat close to the adventitia of the coronary artery. The inflammatory response of pericoronary adipose tissue is involved in the development of atherosclerotic plaques on the one hand, and can reduce the stability of plaques on the other hand, which may eventually lead to acute myocardial infarction. Therefore, the inflammatory response of pericoronary fat plays an important role in the development of acute myocardial infarction. The perivascular fat attenuation index (FAI) is the average CT value of PCAT obtained by post-processing software using traditional coronary CTA data. Studies have shown that inflammatory factors can inhibit the differentiation of preadipocytes and lipid accumulation in perivascular adipose tissue, causing its density on CT to change from the lipid phase (close to -190HU) to the water phase (close to -30HU), so vascular inflammation can be indirectly assessed by measuring FAI. Integrating FAI into current coronary CTA analysis will help identify individuals at risk of future acute myocardial infarction before macroscopic changes in the coronary artery wall structure occur.

[0061] (5) Characteristic parameters include: parameters of influencing factors affecting the measurement of coronary atherosclerotic plaques;

[0062] At this time, the above-mentioned acute myocardial infarction prediction method also includes: classifying the influencing factors to obtain quantitative indicators and counting indicators; comparing the quantitative indicators and counting indicators between groups to determine the correlation between the quantitative indicators and counting indicators and the occurrence of acute myocardial infarction; if the correlation is less than a preset critical value, determining the target threshold of the characteristic parameter; wherein the preset critical value is used to characterize the sensitivity and specificity of the characteristic parameter; based on the target threshold and the correlation, selecting the target characteristic parameter from multiple characteristic parameters;

[0063] Based on the correlation, the probability of acute myocardial infarction occurring within a preset time in the future is predicted, including:

[0064] Based on the correlation between the target characteristic parameter and the occurrence of acute myocardial infarction, the probability of acute myocardial infarction occurring within a preset time in the future is predicted. This implementation example is:

[0065] The results of univariate Cox analysis are shown in Table 1, and the results of multivariate Cox analysis are shown in Table 2. The preset value was used as the boundary for evaluating whether the factor was associated with acute myocardial infarction, where the preset value was set to 0.05. When the factor probability obtained by analysis was greater than the preset value, it was considered that the factor was not associated with acute myocardial infarction. If the preset value was less than 0.05, it was considered that the factor was associated with acute myocardial infarction. For example, age, gender, body mass index, hyperlipidemia, smoking history, lesion length, plaque load, calcified plaque volume, punctate calcification, and napkin ring sign were not associated with acute myocardial infarction; diabetes, hypertension, total plaque volume, non-calcified plaque volume, low-density plaque, positive remodeling, and perivascular fat attenuation index were significantly associated with acute myocardial infarction.

[0066] To prevent the model from overfitting, Figure 2 As shown, multiple characteristic parameters are combined, wherein the maximum area under the curve (AUC) of regression prediction model C is 0.82. Therefore, in this embodiment, model C can be selected as the target regression prediction model.

[0067] Table 1 Results of univariate Cox analysis

[0068] Hazard ratio [95% confidence interval] P-value age 1.05[1.03,1.07] <0.001 gender 1.20[0.50–2.87] 0.682 Body Mass Index 2.19[0.99–4.85] 0.052 diabetes 4.65[2.07–10.48] <0.001 hypertension 2.99[1.36–6.59] 0.006 High cholesterol 1.44[0.67–3.10] 0.351 Smoking history 1.02[0.98–1.05] 0.422 Coronary CTA plaque characteristics Lesion length 1.32[0.12–15.00] 0.823 Plaque burden 1.79[0.91–3.53] 0.092 Total plaque volume 5.95[1.84–19.31] 0.003 Calcified plaque volume 0.96[0.91–1.01] 0.133 Non-calcified plaque volume 3.33[1.07–10.36] 0.037 High-risk plaque characteristics Low-density plaque 9.41[2.56–34.52] 0.001 Spot calcification 1.07[0.19–6.03] 0.973 Positive Reconstruction 4.01[1.28–12.56] 0.017 Napkin ring sign 1.63[0.48–5.56] 0.438 Perivascular fat attenuation index 14.00[3.46–56.64] <0.001

[0069] Table 2 Results of multivariate Cox analysis

[0070] Hazard ratio [95% confidence interval] P-value age 1.05[1.03,1.06] <0.001 diabetes 4.15[1.97–9.58] <0.001 hypertension 1.82[1.21–6.55] 0.016 Total plaque volume 5.05[1.45–7.82] 0.032 Non-calcified plaque volume 3.10[1.01–9.52] 0.029 Low-density plaque 7.21[1.49–31.01] 0.021 Positive Reconstruction 3.11[1.11–12.14] 0.040 Perivascular fat attenuation index 8.91[2.61–18.15] <0.001

[0071] The above scheme can screen the characteristic parameters through the above scheme, so as to select the characteristic parameters with a higher correlation with the occurrence of acute myocardial infarction, thereby improving the prediction accuracy of the predicted probability of acute myocardial infarction occurring within a preset time in the future.

[0072] The above step S140 includes: using the target regression prediction model to perform regression prediction on the characteristic parameters to obtain the correlation between the characteristic parameters and the occurrence of acute myocardial infarction; wherein the target regression prediction model is a regression prediction model with the largest statistic selected from multiple candidate regression prediction models;

[0073] The expression of the target regression prediction model is:

[0074] h=1-h0(t) expβ

[0075] Among them, h represents the correlation between the characteristic parameter and the occurrence of acute myocardial infarction within the future preset time t; h0(t) represents the average survival probability within the future preset time t; β represents the prognostic index, which is the sum of the product of the characteristic parameter and the corresponding preset parameter coefficient.

[0076] The present application example also provides a specific application case:

[0077] like Figure 3 As shown, Figure 3 The coronary artery CTA image of a patient with stable angina pectoris is shown. The patient is 65 years old, has diabetes, hypertension, and a plaque in the proximal segment of the left anterior descending artery, with a moderate stenosis of about 50% ( Figure 3 A), the lesion has a low-density plaque component (shown in blue, Figure 3 B), the total plaque volume is 879.77 mm 3 The volume of non-calcified plaque is 842.25 mm 3 ( Figure 3 C), the perivascular fat attenuation index is -68HU ( Figure 3 D) in the above example is substituted into the regression model to calculate the probability of the patient having a greater acute myocardial infarction in the later stage. The above example proves that the method provided in the embodiment of the present application can effectively support the prediction of the prognosis of patients after thrombectomy.

[0078] like Figure 4 As shown, based on the same inventive concept, the embodiment of the present application also provides an acute myocardial infarction prediction system 200, which includes:

[0079] An image acquisition module 210 is used to acquire a coronary artery CTA image to be processed;

[0080] An image segmentation module 220 is used to perform image segmentation on the coronary artery CTA image to obtain a target segmented image; wherein the target segmented image contains a plurality of coronary artery anatomical structures;

[0081] An image analysis module 230 is used to perform image characteristic analysis on the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques;

[0082] An image processing module 240 is used to perform regression prediction using the characteristic parameters to obtain the correlation between the characteristic parameters and the occurrence of acute myocardial infarction;

[0083] The prediction module 250 is used to predict the probability of acute myocardial infarction occurring within a preset time in the future based on the correlation.

[0084] Optionally, the image segmentation module 220 acquires a target segmentation image, including: acquiring a target segmentation image including at least one segmentation image of a right coronary artery segmentation image, a posterior descending branch segmentation image, a left ventricular posterior branch segmentation image, a left main trunk segmentation image, a left anterior descending branch segmentation image, a left circumflex branch segmentation image, a diagonal branch segmentation image, an intermediate branch segmentation image, and an obtuse marginal branch segmentation image.

[0085] Optionally, the characteristic parameters include high-risk plaque characteristics; the high-risk plaque characteristics include: at least one of low-density plaque, napkin ring sign, positive reconstruction and punctate calcification;

[0086] The image analysis module 230 is specifically used for:

[0087] Based on the CT value of the target segmented image, determining whether the target segmented image has the low-density plaque;

[0088] and / or, judging whether the target segmented image has the napkin ring sign based on the CT value of the target segmented image;

[0089] and / or, based on the target segmented image, determining the maximum blood vessel diameter of the lesion segment and the average diameter of reference blood vessels proximal and distal to the lesion, and judging whether the target segmented image has the positive reconstruction;

[0090] And / or, based on the diameter of the section of the calcification inside the plaque in the target segmented image, it is determined whether the target segmented image has the point-like calcification.

[0091] Optionally, the characteristic parameters include stenosis degree and plaque load;

[0092] The image analysis module 230 is specifically used for:

[0093] Based on the target segmented image, determining the minimum vessel lumen diameter of the lesion segment and the average vessel lumen diameter of the reference vessel proximal and distal to the lesion, and determining the stenosis degree of the target segmented image;

[0094] Based on the target segmented image, the minimum vessel lumen area of ​​the lesion segment and the average vessel lumen area of ​​the reference vessel proximal and distal to the lesion are determined to determine the plaque load of the target segmented image.

[0095] Optionally, the characteristic parameter includes lesion length;

[0096] The image analysis module 230 is specifically used for:

[0097] Based on the target segmented image, the length from the proximal end of the lesion to the distal end of the lesion is determined, and the lesion length of the target segmented image is determined.

[0098] Optionally, the characteristic parameter includes a coronary perivascular fat attenuation index;

[0099] The image analysis module 230 is specifically used for:

[0100] Based on the target segmented image, according to a preset pericoronary fat CT threshold, with the adventitia as the inner boundary and the length perpendicular to the vascular lumen diameter as the outer boundary, the pericoronary fat attenuation index of the target segmented image is determined.

[0101] Optionally, the characteristic parameters include: parameters of influencing factors affecting the measurement of coronary atherosclerotic plaques;

[0102] The acute myocardial infarction prediction system 200 further includes:

[0103] A target characteristic parameter acquisition module is used to classify the influencing factors to obtain quantitative indicators and counting indicators; compare the quantitative indicators and the counting indicators between groups to determine the correlation between the quantitative indicators and the counting indicators and the occurrence of acute myocardial infarction; if the correlation is less than a preset critical value, determine the target threshold of the characteristic parameter; wherein the preset critical value is used to characterize the sensitivity of the characteristic parameter; based on the target threshold and the correlation, select the target characteristic parameter from the multiple characteristic parameters.

[0104] At this time, the prediction module 250 is specifically used to predict the probability of acute myocardial infarction occurring within a preset time in the future based on the correlation between the target characteristic parameter and the occurrence of acute myocardial infarction.

[0105] Optionally, the image processing module 240 is specifically used for:

[0106] Using a target regression prediction model to perform regression prediction on the characteristic parameters to obtain the correlation between the characteristic parameters and the occurrence of acute myocardial infarction; wherein the target regression prediction model is a regression prediction model with the largest statistic selected from multiple candidate regression prediction models;

[0107] The expression of the target regression prediction model is:

[0108] h=1-h0(t) expβ

[0109] Among them, h represents the correlation between the characteristic parameter and the occurrence of acute myocardial infarction within the future preset time t; h0(t) represents the average survival probability within the future preset time t; β represents the prognostic index, which is the sum of the product of the characteristic parameter and the corresponding preset parameter coefficient.

[0110] The present application also provides a computer-readable storage medium, which stores computer program instructions, and when the computer program instructions are read and executed by a computer processor, the acute myocardial infarction prediction method provided by the present application is executed. For example, the computer-readable storage medium can be implemented as a memory in an electronic device.

[0111] In the description of the embodiments of the present application, the term "and / or" is only a description of the association relationship of the associated objects, indicating that there may be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0112] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for predicting acute myocardial infarction, characterized in that: The method comprises: Acquire the coronary artery CTA image to be processed; Performing image segmentation on the coronary artery CTA image to obtain a target segmented image; wherein the target segmented image includes a plurality of coronary artery anatomical structures; Performing image characteristic analysis on the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques; Using the characteristic parameters to perform regression prediction, and obtaining the correlation between the characteristic parameters and the occurrence of acute myocardial infarction; Based on the correlation, the probability of acute myocardial infarction occurring within a preset time in the future is predicted.

2. The method for predicting acute myocardial infarction according to claim 1, characterized in that: The step of obtaining a target segmented image comprises: Acquire a target segmentation image including at least one of the right coronary artery segmentation image, the posterior descending branch segmentation image, the left ventricular posterior branch segmentation image, the left main trunk segmentation image, the left anterior descending branch segmentation image, the left circumflex branch segmentation image, the diagonal branch segmentation image, the intermediate branch segmentation image and the obtuse marginal branch segmentation image.

3. The method for predicting acute myocardial infarction according to claim 1, characterized in that: The performing image characteristic analysis on the target segmented image comprises: Determine the position of the vascular intima and the vascular adventitia in the target segmented image; Determining a region of interest of coronary atherosclerotic plaque in the target segmented image based on the positions of the vascular intima and the vascular adventitia; Image characteristic analysis is performed on the target segmented image based on the region of interest.

4. The method for predicting acute myocardial infarction according to claim 2, characterized in that: The characteristic parameters include high-risk plaque characteristics; the high-risk plaque characteristics include: at least one of low-density plaque, napkin ring sign, positive reconstruction and punctate calcification; The performing image characteristic analysis on the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques includes: Based on the CT value of the target segmented image, determining whether the target segmented image has the low-density plaque; and / or, judging whether the target segmented image has the napkin ring sign based on the CT value of the target segmented image; and / or, based on the target segmented image, determining the maximum blood vessel diameter of the lesion segment and the average diameter of reference blood vessels proximal and distal to the lesion, and judging whether the target segmented image has the positive reconstruction; And / or, based on the diameter of the section of the calcification inside the plaque in the target segmented image, it is determined whether the target segmented image has the point-like calcification.

5. The method for predicting acute myocardial infarction according to claim 2, characterized in that: The characteristic parameters include stenosis degree and plaque load; The performing image characteristic analysis on the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques includes: Based on the target segmented image, determining the minimum vessel lumen diameter of the lesion segment and the average vessel lumen diameter of the reference vessel proximal and distal to the lesion, and determining the stenosis degree of the target segmented image; Based on the target segmented image, the minimum vessel lumen area of ​​the lesion segment and the average vessel lumen area of ​​the reference vessel proximal and distal to the lesion are determined to determine the plaque load of the target segmented image.

6. The method for predicting acute myocardial infarction according to claim 2, characterized in that: The characteristic parameters include lesion length; The performing image characteristic analysis on the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques includes: Based on the target segmented image, the length from the proximal end of the lesion to the distal end of the lesion is determined, and the lesion length of the target segmented image is determined.

7. The method for predicting acute myocardial infarction according to claim 2, characterized in that: The characteristic parameters include coronary perivascular fat attenuation index; The performing image characteristic analysis on the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques includes: Based on the target segmented image, according to a preset pericoronary fat CT threshold, with the adventitia as the inner boundary and the length perpendicular to the vascular lumen diameter as the outer boundary, the pericoronary fat attenuation index of the target segmented image is determined.

8. The method for predicting acute myocardial infarction according to claim 2, characterized in that: The characteristic parameters include: parameters of influencing factors affecting the measurement of coronary atherosclerotic plaques; The method further comprises: Classify the influencing factors to obtain quantitative indicators and counting indicators; Comparing the quantitative index and the counting index between groups to determine the correlation between the quantitative index and the counting index and the occurrence of acute myocardial infarction; If the correlation is less than a preset critical value, determining a target threshold of the characteristic parameter; wherein the preset critical value is used to characterize the sensitivity of the characteristic parameter; Selecting a target characteristic parameter from the plurality of characteristic parameters based on the target threshold and the correlation; The predicting, based on the correlation, of the probability of acute myocardial infarction occurring within a preset future time period includes: Based on the correlation between the target characteristic parameter and the occurrence of acute myocardial infarction, the probability of acute myocardial infarction occurring within a preset time in the future is predicted.

9. The method for predicting acute myocardial infarction according to any one of claims 1 to 8, characterized in that: The step of performing regression prediction using the characteristic parameters to obtain the correlation between the characteristic parameters and the occurrence of acute myocardial infarction includes: Using a target regression prediction model to perform regression prediction on the characteristic parameters to obtain the correlation between the characteristic parameters and the occurrence of acute myocardial infarction; wherein the target regression prediction model is a regression prediction model with the largest statistic selected from multiple candidate regression prediction models; The expression of the target regression prediction model is: h=1-h0(t) expβ Among them, h represents the correlation between the characteristic parameter and the occurrence of acute myocardial infarction within the future preset time t; h0(t) represents the average survival probability within the future preset time t; β represents the prognostic index, which is the sum of the product of the characteristic parameter and the corresponding preset parameter coefficient.

10. An acute myocardial infarction prediction system, characterized in that: The system comprises: An image acquisition module, used for acquiring a coronary artery CTA image to be processed; An image segmentation module, used for performing image segmentation on the coronary artery CTA image to obtain a target segmented image; wherein the target segmented image contains a plurality of coronary artery anatomical structures; An image analysis module, used for performing image characteristic analysis on the target segmented image to obtain characteristic parameters for characterizing the characteristics of coronary atherosclerotic plaques; An image processing module, used for performing regression prediction using the characteristic parameters to obtain the correlation between the characteristic parameters and the occurrence of acute myocardial infarction; The prediction module is used to predict the probability of acute myocardial infarction occurring within a preset time in the future based on the correlation.

Images