Method for establishing HCM patient endpoint event prediction model based on LGE radiomics features and clinical features, and method and device for predicting SCD risk

Abstract

The invention relates to the technical field of medical big data, in particular to a method for establishing an HCM patient endpoint event prediction model based on LGE radiomics features and clinical prediction features, and a method and device for predicting SCD risks based on the HCM patient endpoint event prediction model. According to the method for establishing the model, LGE radiomics and clinical features are innovatively combined, and a plurality of radiomics-clinical feature combined prediction models corresponding to each of a cardiovascular death patient group, a heart failure inpatient group and a secondary endpoint event patient group are established by adopting a plurality of machine learning methods; and then, screening out the optimal radiomics-clinical feature combined prediction model of each patient group to serve as the HCM patient endpoint event prediction model of each patient group. The present disclosure enables the prediction of various cardiovascular endpoint events for HCM patients, not limited to sudden cardiac death, and exhibiting good predictive efficacy in various predictions.

Description

Technical Field

[0001] The present disclosure relates to the field of medical big data technology, and specifically to a method for establishing an endpoint event prediction model for patients with hypertrophic cardiomyopathy (HCM) based on delayed gadolinium enhancement (LGE) imaging features and clinical prediction features, and a method and device for predicting the risk of sudden cardiac death (SCD) based on the endpoint event prediction model for HCM patients. Background Art

[0002] HCM is a common hereditary heart disease, which is inherited in an autosomal dominant manner and is characterized by significant myocardial hypertrophy (commonly involving the left ventricular wall) with diastolic dysfunction, usually accompanied by symptoms such as chest pain, dyspnea, syncope, and sudden death. Studies have shown that the prevalence of HCM in ordinary adults is 0.16% to 0.23%, and the annual mortality rate is 1.5% to 4%, or even higher. In fact, the clinical course of HCM is diverse. 47% to 50% of HCM patients can maintain a benign course with long-term asymptomatic or no obvious symptoms throughout their lives. On the other hand, 40% to 50% of HCM patients will experience at least one adverse cardiovascular event, including sudden cardiac death, heart failure, atrial fibrillation / stroke, etc., causing a serious social health burden.

[0003] Due to the high risk of SCD and extremely poor prognosis of SCD in HCM patients, current research on the prognosis of HCM patients focuses on the early prediction and prevention of SCD, and has achieved good results. However, heart failure is the main cause of HCM-related morbidity and mortality and the largest unmet treatment need for adult HCM patients, and is gradually becoming a research focus.

[0004] However, there is still a lack of relevant risk prediction research on heart failure, cardiovascular death, etc. in HCM patients. Therefore, it is an urgent problem to find predictive factors and establish prediction models for HCM patients to predict the risk of cardiovascular endpoint events (cardiovascular death, heart failure, re-hospitalization rate, etc.) in HCM patients and guide the formulation of treatment plans for HCM patients. Summary of the invention

[0005] In order to solve the problems in the related art, the embodiments of the present disclosure provide a method for establishing an endpoint event prediction model for HCM patients based on LGE radiomics features and clinical prediction features, and a method and device for predicting SCD risk based on the endpoint event prediction model for HCM patients.

[0006] In a first aspect, the present disclosure provides a method for establishing an endpoint event prediction model for HCM patients based on LGE radiomics features and clinical prediction features, the method being implemented by a computer and comprising:

[0007] Obtaining an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively;

[0008] According to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group, the first clinical characteristic prediction factor of the cardiovascular death patient group is obtained; according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group, the second clinical characteristic prediction factor of the heart failure hospitalized patient group is obtained; according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group, the third clinical characteristic prediction factor of the secondary endpoint event patient group is obtained; wherein, the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class;

[0009] According to the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, obtaining the post-screening imaging omics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group; calculating the imaging omics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group according to the corresponding post-screening imaging omics features;

[0010] For each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: based on the imaging feature scores of each patient group and the corresponding clinical feature predictors, a variety of machine learning methods are used to establish corresponding multiple imaging-clinical feature joint prediction models; the multiple imaging-clinical feature joint prediction models are verified, and the optimal imaging-clinical feature joint prediction model is selected as the HCM patient endpoint event prediction model for each patient group.

[0011] According to an embodiment of the present disclosure, obtaining a first clinical characteristic prediction factor of the cardiovascular death patient group based on the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group includes: inputting the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group into a multivariate survival analysis model using a stepwise regression method to obtain the first clinical characteristic prediction factor;

[0012] According to the baseline clinical prediction characteristics of the heart failure inpatient group and the second HCM patient control group, obtaining a second clinical characteristic prediction factor of the heart failure inpatient group, including: inputting the baseline clinical prediction characteristics of the heart failure inpatient group and the second HCM patient control group into a multivariate survival analysis model using a stepwise regression method to obtain the second clinical characteristic prediction factor;

[0013] According to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group, obtaining the third clinical characteristic prediction factor of the secondary endpoint event patient group, including: inputting the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group into a multivariate survival analysis model using a stepwise regression method to obtain the third clinical characteristic prediction factor;

[0014] The baseline clinical prediction features include: basic patient information, echocardiogram information, electrocardiogram information, and basic cardiac magnetic resonance imaging parameters.

[0015] According to an embodiment of the present disclosure, the multi-factor survival analysis model is any one or more of the following:

[0016] Cox proportional hazards model, Weibull distribution model, competing risks model;

[0017] The stepwise regression method includes any one or more of the following:

[0018] Forward stepwise regression, backward stepwise regression, and bidirectional stepwise regression.

[0019] According to an embodiment of the present disclosure, the HCM endpoint events of the HCM patients in the cardiovascular death patient group include sudden cardiac death and / or death from heart failure;

[0020] The HCM endpoint events of HCM patients in the group of patients hospitalized with heart failure included hospitalization for heart failure;

[0021] The HCM endpoint events of HCM patients in the secondary endpoint event patient group include any one or more of the following: unplanned readmission due to cardiovascular disease, unplanned readmission due to other reasons, and death from non-cardiovascular disease;

[0022] No sudden cardiac death or death from heart failure occurred in the HCM patients in the first HCM patient control group;

[0023] No HCM patients in the second HCM patient control group were hospitalized for heart failure;

[0024] HCM patients in the third HCM patient control group did not experience unplanned readmission due to cardiovascular disease, unplanned readmission due to other reasons, and non-cardiovascular disease death.

[0025] According to an embodiment of the present disclosure, the LGE image group of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group is used to obtain the post-screening radiomics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, including:

[0026] For each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, an LGE image group of each patient group is obtained, wherein the LGE image group includes LGE images of HCM patients in each patient group; the left ventricular myocardial short-axis section of the LGE image in the LGE image group is used as a region of interest, and the pre-screening imaging omics features of each patient group are extracted from the region of interest;

[0027] The pre-screening radiomics features of each patient group were input into the least absolute shrinkage and selection operator LASSO model for correlation analysis to obtain the post-screening radiomics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group.

[0028] According to an embodiment of the present disclosure, the calculating of the imaging feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group according to the corresponding post-screening imaging feature comprises:

[0029] According to the characteristic distribution and LASSO coefficient of the corresponding post-screening imaging features, the imaging feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group are calculated respectively.

[0030] According to an embodiment of the present disclosure, the machine learning method includes any one of the following:

[0031] Logistic regression, K-nearest neighbor method, elastic network method, naive Bayes method.

[0032] According to an embodiment of the present disclosure, the multiple radiomics-clinical feature joint prediction models are verified and the optimal radiomics-clinical feature joint prediction model is selected, including: calculating the area under the ROC curve of the multiple radiomics-clinical feature joint prediction models, and taking the radiomics-clinical feature joint prediction model with the largest area under the ROC curve as the optimal radiomics-clinical feature joint prediction model.

[0033] In a second aspect, the present disclosure provides a method for predicting the risk of sudden cardiac death (SCD) based on an endpoint event prediction model for HCM patients, the method being implemented by a computer and comprising:

[0034] Obtaining an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively;

[0035] According to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group, the first clinical characteristic prediction factor of the cardiovascular death patient group is obtained; according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group, the second clinical characteristic prediction factor of the heart failure hospitalized patient group is obtained; according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group, the third clinical characteristic prediction factor of the secondary endpoint event patient group is obtained; wherein, the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class;

[0036] According to the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, obtaining the post-screening imaging omics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group; calculating the imaging omics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group according to the corresponding post-screening imaging omics features;

[0037] For each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: according to the imaging omics feature scores of each patient group and the corresponding clinical feature predictors, a plurality of corresponding imaging omics-clinical feature joint prediction models are established by using a plurality of machine learning methods; the plurality of imaging omics-clinical feature joint prediction models are verified, and the optimal imaging omics-clinical feature joint prediction model is selected as the HCM patient endpoint event prediction model for each patient group;

[0038] Acquire a data group of HCM patients to be tested, wherein the data group of HCM patients to be tested includes data of multiple HCM patients to be tested;

[0039] The HCM patient data group to be tested is input into the HCM patient endpoint event prediction model of the cardiovascular death patient group to obtain the corresponding SCD risk prediction result.

[0040] According to an embodiment of the present disclosure, the machine learning method includes any one of the following:

[0041] Logistic regression, K-nearest neighbor method, elastic network method, naive Bayes method.

[0042] In a third aspect, the present disclosure provides an apparatus for establishing an endpoint event prediction model for HCM patients based on LGE radiomics features and clinical prediction features, wherein the apparatus is set in a computer and comprises:

[0043] A first acquisition module is configured to acquire an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively;

[0044] The second acquisition module is configured to acquire the first clinical characteristic prediction factor of the cardiovascular death patient group according to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group; acquire the second clinical characteristic prediction factor of the heart failure hospitalized patient group according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group; acquire the third clinical characteristic prediction factor of the secondary endpoint event patient group according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group; wherein the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class;

[0045] The first calculation module is configured to obtain the post-screening imaging omics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group; and calculate the imaging omics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the corresponding post-screening imaging omics features;

[0046] The first model establishment selection module is configured to: for each patient group among the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: based on the imaging feature scores of each patient group and the corresponding clinical feature predictors, use multiple machine learning methods to establish corresponding multiple imaging feature joint prediction models; verify the multiple imaging feature joint prediction models, and select the optimal imaging feature joint prediction model as the HCM patient endpoint event prediction model for each patient group.

[0047] In a fourth aspect, the present disclosure provides a device for predicting the risk of sudden cardiac death (SCD) based on an endpoint event prediction model for HCM patients, comprising:

[0048] A third acquisition module is configured to acquire an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively;

[0049] The fourth acquisition module is configured to acquire the first clinical characteristic prediction factor of the cardiovascular death patient group according to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group; acquire the second clinical characteristic prediction factor of the heart failure hospitalized patient group according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group; acquire the third clinical characteristic prediction factor of the secondary endpoint event patient group according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group; wherein the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class;

[0050] The second calculation module is configured to obtain the post-screening imaging omics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group; and calculate the imaging omics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the corresponding post-screening imaging omics features;

[0051] The second model establishment selection module is configured to: for each patient group of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: according to the imaging omics feature score of each patient group and the corresponding clinical feature predictor, use multiple machine learning methods to establish corresponding multiple imaging omics-clinical feature joint prediction models; verify the multiple imaging omics-clinical feature joint prediction models, and select the optimal imaging omics-clinical feature joint prediction model as the HCM patient endpoint event prediction model for each patient group;

[0052] A data acquisition module is configured to acquire a data group of HCM patients to be tested, wherein the data group of HCM patients to be tested includes data of multiple HCM patients to be tested;

[0053] The SCD prediction module is configured to input the HCM patient data group to be tested into the HCM patient endpoint event prediction model of the cardiovascular death patient group to obtain the corresponding SCD risk prediction result.

[0054] In a fifth aspect, an embodiment of the present disclosure provides an electronic device, comprising a memory and a processor; wherein the memory is used to store computer instructions, wherein the computer instructions are executed by the processor to implement a method as described in any one of the first aspect or the second aspect.

[0055] In a sixth aspect, a computer-readable storage medium is provided in an embodiment of the present disclosure, on which computer instructions are stored. When the computer instructions are executed by a processor, the method described in any one of the first aspect or the second aspect is implemented.

[0056] In a seventh aspect, a computer program product is provided in an embodiment of the present disclosure, comprising computer instructions, which, when executed by a processor, implement the method as described in any one of the first aspect or the second aspect.

[0057] According to the technical solution provided in the embodiments of the present disclosure, LGE radiomics is combined with clinical characteristics, and based on the radiomics feature scores of each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, as well as the corresponding clinical feature prediction factors, a variety of machine learning methods are used to establish corresponding multiple radiomics-clinical feature joint prediction models; and the multiple radiomics-clinical feature joint prediction models are verified, and the optimal radiomics-clinical feature joint prediction model is selected as the HCM patient endpoint event prediction model for each patient group.

[0058] The present invention innovatively combines LGE imaging genomics with clinical characteristics, so that the prediction accuracy of the established prediction model is better than that of the model established by only using clinical prediction indicators, and it can not only predict the cardiovascular endpoint event of sudden cardiac death, but also predict other cardiovascular endpoint events of HCM patients (death from heart failure, hospitalization for heart failure, unplanned readmission due to cardiovascular disease, unplanned readmission due to other reasons, and death from non-cardiovascular disease). It realizes the prediction of endpoint events such as hospitalization for heart failure that are not given special attention by existing schemes, and shows good predictive efficacy in the prediction of various cardiovascular endpoint events.

[0059] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Other features, objectives and advantages of the present disclosure will become more apparent through the following detailed description of non-limiting embodiments in conjunction with the accompanying drawings. In the accompanying drawings:

[0061] Figure 1 A flowchart showing a method for establishing an HCM patient endpoint event prediction model based on LGE radiomics features and clinical prediction features according to an embodiment of the present disclosure;

[0062] Figure 2 A schematic diagram showing the distribution of endpoint events in HCM patients according to an embodiment of the present disclosure;

[0063] Figure 3 A schematic diagram showing the area under the ROC curve of multiple radiomics-clinical characteristics joint prediction models for a cardiovascular death patient group in an embodiment of the present disclosure;

[0064] Figure 4 A schematic diagram showing the area under the ROC curve of multiple radiomics-clinical characteristics joint prediction models for a group of hospitalized patients with heart failure in an embodiment of the present disclosure;

[0065] Figure 5 A schematic diagram showing the area under the ROC curve of multiple radiomics-clinical characteristics joint prediction models for the secondary endpoint event patient group in the embodiments of the present disclosure;

[0066] Figure 6 A flowchart showing a method for predicting SCD risk based on an HCM patient endpoint event prediction model according to an embodiment of the present disclosure;

[0067] Figure 7 A structural block diagram of a device for establishing an HCM patient endpoint event prediction model based on LGE radiomics features and clinical prediction features according to an embodiment of the present disclosure is shown;

[0068] Figure 8 A structural block diagram of a device for predicting SCD risk based on an HCM patient endpoint event prediction model according to an embodiment of the present disclosure is shown;

[0069] Fig. 9 A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown;

[0070] Fig.10 A schematic diagram showing the structure of a computer system suitable for implementing the method according to an embodiment of the present disclosure is shown. DETAILED DESCRIPTION

[0071] Hereinafter, exemplary embodiments of the present disclosure will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily implement them. In addition, for the sake of clarity, parts not related to the description of the exemplary embodiments are omitted in the accompanying drawings.

[0072] In the present disclosure, it should be understood that terms such as "include" or "have" are intended to indicate the presence of features, numbers, steps, behaviors, components, parts, or a combination thereof disclosed in the present specification, and are not intended to exclude the possibility that one or more other features, numbers, steps, behaviors, components, parts, or a combination thereof exist or are added.

[0073] It should also be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present disclosure may be combined with each other. The present disclosure will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0074] In the present disclosure, if it involves operations of obtaining user information or user data or displaying user information or user data to others, the operations are all authorized and confirmed by the user, or actively selected by the user.

[0075] As mentioned above, early prediction and identification of the risk of cardiovascular endpoint events are of great significance for HCM patients. After investigation and in-depth analysis, the inventors found that although existing studies have revealed that factors such as gender, brain natriuretic peptide levels, left atrial diameter, critical decrease in left ventricular ejection fraction, and left ventricular outflow tract pressure gradient are considered as predictive indicators of heart failure and adverse cardiovascular events in HCM patients, these indicators generally face the limitations of insufficient sensitivity or have not been fully verified by large-scale clinical trials in practical applications. Therefore, how to explore more accurate and highly sensitive predictive factors and build a more reliable and accurate prediction model based on this, in order to achieve accurate assessment and prediction of the disease progression and prognostic risk of HCM patients, is an urgent problem to be solved.

[0076] With the development of imaging technology, cardiac magnetic resonance imaging (CMR) has become the standard of non-invasive imaging tools in the cardiovascular field. LGE, as a CMR technology, can indicate the presence and degree of myocardial fibrosis by observing the concentration of gadolinium contrast agent remaining in the scar area. The inventors found that although there have been studies exploring the significant correlation between LGE and cardiovascular endpoints, there are still some issues that cannot be ignored when applying CMR-LGE examination results to the prediction of cardiovascular endpoints in HCM patients:

[0077] First, the threshold of LGE% is currently set at 15% when predicting the risk of cardiovascular endpoint events. However, some clinical studies have found that even HCM patients with LGE% below 15% have a higher risk of cardiovascular endpoint events. Therefore, relying solely on LGE% as a predictor of cardiovascular endpoint events may not be sensitive enough.

[0078] Furthermore, LGE enhancement is usually based on the degree of normal myocardial enhancement as a reference standard. In the absence of normal myocardial controls in patients with extensive myocardial fibrosis, the measurement of LGE% may have a large error. Therefore, more robust predictive factors are still needed in clinical practice.

[0079] Therefore, CMR-LGE has problems such as poor sensitivity and large measurement errors in practical applications. The inventors noticed that with the continuous development of artificial intelligence, radiomics has received certain attention in the cardiovascular field, and the inventors innovatively combined CMR-LGE radiomics with clinical characteristics to achieve a method for establishing an endpoint event prediction model for HCM patients based on LGE radiomics features and clinical prediction features, the method being implemented by a computer and comprising:

[0080] Obtain an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively; according to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group, obtain a first clinical characteristic prediction factor of the cardiovascular death patient group; according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group, obtain a second clinical characteristic prediction factor of the heart failure hospitalized patient group; according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group, obtain a third clinical characteristic prediction factor of the secondary endpoint event patient group; wherein, the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, and the second clinical characteristic prediction factor includes: N YHA cardiac function grade, non-sustained ventricular tachycardia, the third clinical feature predictor includes: age, NYHA cardiac function grade; according to the LGE image group of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, obtain the post-screening imaging genomics characteristics corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group; calculate the imaging genomics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group according to the corresponding post-screening imaging genomics features; for each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: according to the imaging genomics feature scores of each patient group and the corresponding clinical feature predictors, use multiple machine learning methods to establish corresponding multiple imaging genomics-clinical feature joint prediction models; verify the multiple imaging genomics-clinical feature joint prediction models, and select the optimal imaging genomics-clinical feature joint prediction model as the HCM patient endpoint event prediction model for each patient group.

[0081] This paper preliminarily explores the feasibility of using machine learning to combine LGE imaging genomics and clinical characteristics to predict endpoint events in HCM patients. It can not only predict the cardiovascular endpoint event of sudden cardiac death, but also predict other endpoint events in HCM patients. The prediction accuracy is better than the model established by using only clinical prediction indicators, and it shows good predictive efficacy in the prediction of various endpoint events, with good universality.

[0082] Figure 1 A flowchart showing a method for establishing an HCM patient endpoint event prediction model based on LGE radiomics features and clinical prediction features according to an embodiment of the present disclosure is shown. The method is implemented by a computer, such as Figure 1 As shown, the method includes the following steps S101 to S104:

[0083] In step S101, an HCM patient endpoint event group and an HCM patient control group are obtained, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively; wherein the cardiovascular death patient group, the heart failure hospitalized patient group, the secondary endpoint event patient group, the first HCM patient control group, the second HCM patient control group, and the third HCM patient control group all include multiple HCM patients.

[0084] According to an embodiment of the present disclosure, the HCM endpoint events of the HCM patients in the cardiovascular death patient group include sudden cardiac death (SCD) and / or death from heart failure.

[0085] According to an embodiment of the present disclosure, the HCM endpoint event of the HCM patients in the group of patients hospitalized with heart failure includes hospitalization for heart failure.

[0086] According to an embodiment of the present disclosure, the HCM endpoint events of the HCM patients in the secondary endpoint event patient group include any one or more of the following: unplanned readmission due to cardiovascular disease, unplanned readmission due to other reasons, and death from non-cardiovascular disease.

[0087] According to an embodiment of the present disclosure, the HCM patients in the first HCM patient control group did not suffer from sudden cardiac death and death from heart failure.

[0088] According to an embodiment of the present disclosure, the HCM patients in the second HCM patient control group did not suffer from hospitalization for heart failure.

[0089] According to an embodiment of the present disclosure, the HCM patients in the third HCM patient control group did not experience unplanned readmission due to cardiovascular disease, unplanned readmission due to other reasons, and non-cardiovascular disease death.

[0090] like Figure 2 As shown, in a specific embodiment, 104 HCM patients were enrolled and followed up for an average of 2.78 years (1.16 to 5.28 years). Among them, 10 patients (9.6%) suffered cardiovascular death, and they were included in the cardiovascular death patient group, and the corresponding first HCM patient control group was 94 cases. 15 patients (14.4%) were hospitalized for heart failure, and they were included in the heart failure hospitalization patient group, and the corresponding second HCM patient control group was 89 cases. A total of 29 patients (27.9%) had secondary endpoint events, including 25 patients (24.0%) with unplanned readmissions due to cardiovascular and cerebrovascular diseases, and 4 patients (3.8%) with unplanned readmissions due to other reasons, and they were included in the secondary endpoint event patient group, and the corresponding third HCM patient control group was 75 cases.

[0091] In step S102, the first clinical characteristic predictor of the cardiovascular death patient group is obtained according to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group; the second clinical characteristic predictor of the heart failure hospitalized patient group is obtained according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group; the third clinical characteristic predictor of the secondary endpoint event patient group is obtained according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group; wherein, the first clinical characteristic predictor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic predictor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic predictor includes: age, NYHA heart function class.

[0092] According to an embodiment of the present disclosure, the baseline clinical predictive characteristics of the cardiovascular death patient group, the heart failure hospitalized patient group, the secondary endpoint event patient group, the first HCM patient control group, the second HCM patient control group, and the third HCM patient control group all include: basic patient information, echocardiography information, electrocardiogram information, and basic cardiac magnetic resonance parameters.

[0093] Specifically, the basic information of the patient includes: gender, age, height, weight, NYHA heart function classification, family history of sudden cardiac death, etc. Echocardiographic information includes: ventricular thickest wall thickness, left ventricular ejection fraction, left atrial diameter, left ventricular end-diastolic diameter, left ventricular posterior wall diastolic thickness, right ventricular end-diastolic diameter, etc. Electrocardiographic information includes: non-sustained ventricular tachycardia, atrial tachycardia, total number of ventricular premature beats in 24 hours, total number of paired ventricular premature beats in 24 hours, etc. Basic parameters of cardiac magnetic resonance imaging include: left ventricular myocardial mass, left ventricular end-diastolic volume, left ventricular end-systolic volume, ventricular thickest wall thickness, LGE enhancement percentage, etc.

[0094] Among them, the NYHA heart function classification divides the impaired heart function into four levels according to the degree of activity that induces heart failure symptoms: Level I: daily activities are not restricted; Level II: physical activities are slightly restricted; Level III: physical activities are significantly restricted; Level IV: cannot engage in any physical activities.

[0095] According to an embodiment of the present disclosure, the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group are input into a multivariate survival analysis model using a stepwise regression method to obtain the first clinical characteristic prediction factor.

[0096] According to an embodiment of the present disclosure, the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group are input into a multivariate survival analysis model using a stepwise regression method to obtain the second clinical characteristic prediction factor.

[0097] According to an embodiment of the present disclosure, the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group are input into a multivariate survival analysis model using a stepwise regression method to obtain the third clinical characteristic prediction factor.

[0098] In detail, for the baseline clinical predictive characteristics, these variables were entered into the multivariate survival analysis model using the stepwise regression method to obtain the clinical characteristic predictors corresponding to each of the patient groups.

[0099] According to an embodiment of the present disclosure, the multi-factor survival analysis model is any one or more of the following: Cox proportional hazard model, Weibull distribution model, competing risk model.

[0100] According to an embodiment of the present disclosure, the stepwise regression method includes any one or more of the following: forward stepwise regression method, backward stepwise regression method, and bidirectional stepwise regression method.

[0101] In a specific embodiment, the baseline clinical prediction characteristics are subjected to multivariate analysis using the forward stepwise regression method (p value is 0.05 for entry and 0.1 for exit), and the relevant clinical characteristic predictors in the cardiovascular death patient group, heart failure hospitalized patient group, and secondary endpoint event patient group are screened out respectively.

[0102] according to Figure 2 The example shown further illustrates the clinical characteristic predictors obtained by forward stepwise regression in multivariate Cox regression, as shown in Table 1:

[0103] Table 1

[0104] Predictors p-value Cardiovascular death left atrial diameter 0.004 Family history of sudden cardiac death 0.060 Hospitalization for heart failure NYHA cardiac function classification 0.049 Non-sustained ventricular tachycardia 0.007 Secondary endpoint events age 0.023 NYHA cardiac function classification 0.033

[0105] Among them, in the multivariate Cox regression, the forward stepwise regression method was used to find the best combination of clinical prediction features by gradually adding baseline clinical prediction features to the Cox proportional hazard model and re-evaluating the goodness of fit of the model at each step. The p value was used as the entry (0.05) and exit (0.1) criteria. If the exit of a clinical prediction feature resulted in a significant decrease in the model goodness of fit (judged by a p value less than 0.05), the clinical prediction feature was retained in the model. Otherwise, it could be considered to be removed from the model (judged by a p value greater than 0.1). Thus, the clinical characteristic predictors corresponding to the cardiovascular death patient group were obtained, including left atrial diameter and family history of sudden cardiac death (although the variable "family history of sudden cardiac death" was not statistically significant, it was included in subsequent studies considering that it was a recognized risk factor for the occurrence of cardiovascular endpoint events in HCM patients and was only marginally unrelated). The clinical characteristic predictors corresponding to the heart failure hospitalized patient group included NYHA heart function class and non-sustained ventricular tachycardia. The clinical characteristic predictors corresponding to the secondary endpoint event patient group included age and NYHA heart function class.

[0106] In step S103, based on the LGE image groups of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, the post-screening imaging genomics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group are obtained; and the imaging genomics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group are calculated based on the corresponding post-screening imaging genomics features.

[0107] Specifically, for each patient group among the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, an LGE image group of each patient group is obtained, and the LGE image group includes the LGE images of the HCM patients in each patient group; the left ventricular myocardial short-axis section of the LGE image in the LGE image group is used as a region of interest, and the pre-screening imaging omics features of each patient group are extracted from the region of interest.

[0108] The pre-screening radiomics features of each patient group were input into the Least Absolute Shrinkage and Selection Operator (LASSO) model for correlation analysis to obtain the post-screening radiomics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group.

[0109] Furthermore, according to the characteristic distribution and LASSO coefficient of the corresponding post-screening imaging features, the imaging feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group are calculated respectively.

[0110] Specifically, the LGE images in the LGE image group of each patient group are outlined, and the left ventricular ventricular muscle is circled as the region of interest ROI on the LGE image of the short axis of the heart. Then, the imaging features of the left ventricular myocardium are extracted based on the ROI. Under the original image and multiple filters, morphological features, first-order statistical features, second-order and high-order texture features are extracted as pre-screening imaging features. The pre-screening imaging features are then input into the LASSO model for correlation analysis, and the imaging features with the LASSO coefficient ≠ 0 are selected based on the λ value that minimizes the average cross-validation error.

[0111] exist Figure 2 In the example shown, after screening, the group of patients with cardiovascular death included 7 post-screening imaging features, which are:

[0112] Logarithm-gldm-Small DependenceEmphasis; wavelet transform LHL-adjacent gray level difference matrix-strength (wavelet-LHL-ngtdm-Strength); wavelet transform HLH-gray level co-occurrence matrix feature-total entropy (wavelet-HLH-glcm-Sum Entropy); wavelet transform LHH-first order feature-kurtosis (wavelet-LHH-first order-Kurtosis); wavelet transform HHH-gray level co-occurrence matrix feature-difference entropy (wavelet-HHH-glcm-Difference Entropy); wavelet transform HLL-gray level size area matrix feature-small size area distribution measure (wavelet-HLL-glszm-Small Area Emphasis); wavelet transform HHH-first order feature-mean (wavelet-HHH-first order-Mean).

[0113] There were three post-screening imaging features in the heart failure hospitalization group:

[0114] Wavelet transform HLH-first order feature-mean (wavelet-HLH-first order-Mean); local binary pattern (3D image)-grayscale run-length matrix-run entropy (lbp-3D-k-glrlm-Run Entropy); wavelet transform LHH-grayscale size area matrix feature-size area non-uniformity normalized (wavelet-LHH-glszm-Size Zone NonUniformity Normalized).

[0115] The secondary endpoint event group included 6 post-screening radiomics features, namely:

[0116] Local binary pattern (3D image)-first order feature-minimum eigenvalue (lbp-3D-k-first order-Minimum); wavelet transform HHH-first order feature-mean (wavelet-HHH-first order-Mean); wavelet transform HHH-grayscale size zone matrix feature-zone variance (wavelet-HHH-glszm-Zone Variance); wavelet transform HLH-first order feature-mean (wavelet-HLH-first order-Mean); wavelet transform LHH-grayscale size zone matrix feature-size zone non-uniformity normalized (wavelet-LHH glszm-Size Zone Non UniformityNormalized); wavelet transform LHL-grayscale co-occurrence matrix feature-difference square (wavelet-LHL-glcm-DifferenceVariance).

[0117] According to the embodiments of the present disclosure, the radiomics feature scores (Rad-scores) corresponding to each patient group are calculated based on the distribution of each included post-screening radiomics feature and the corresponding characteristic coefficients of different radiomics features in LASSO regression. Rad-score integrates multiple radiomics features into one numerical value to characterize the imaging characteristics of interest, which facilitates the application of complex radiomics information in clinical decision-making and research.

[0118] In step S104, for each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: based on the imaging feature scores of each patient group and the corresponding clinical feature prediction factors, a plurality of corresponding imaging feature-clinical feature joint prediction models are established using a variety of machine learning methods; the plurality of imaging feature-clinical feature joint prediction models are verified, and the optimal imaging feature-clinical feature joint prediction model is selected as the HCM patient endpoint event prediction model for each patient group.

[0119] According to an embodiment of the present disclosure, the machine learning method includes any one of the following: logistic regression method, K nearest neighbor method, elastic network method, naive Bayes method.

[0120] According to an embodiment of the present disclosure, the areas under the ROC curves of the multiple radiomics-clinical feature joint prediction models are calculated, and the radiomics-clinical feature joint prediction model with the largest area under the ROC curve is taken as the optimal radiomics-clinical feature joint prediction model.

[0121] In the present disclosure, the clinical prediction index is combined with the Rad-Score of each endpoint event, and four machine learning methods, namely, Logistic Regression (LR), K-Nearest Neighbors (KNN), Elastic Net (EN), and Naive Bayes, are used to establish the corresponding radiomics-clinical feature joint prediction model. After each prediction model is established, the area under the ROC curve (AUC) of each model is calculated, and the radiomics-clinical feature joint prediction model with the largest area under the ROC curve is taken as the optimal radiomics-clinical feature joint prediction model.

[0122] Figure 3 A schematic diagram showing the area under the ROC curve of multiple radiomics-clinical characteristics joint prediction models for a cardiovascular death patient group in an embodiment of the present disclosure; Figure 4 A schematic diagram showing the area under the ROC curve of multiple radiomics-clinical characteristics joint prediction models for a group of hospitalized patients with heart failure in an embodiment of the present disclosure; Figure 5 A schematic diagram showing the area under the ROC curve of multiple radiomics-clinical characteristics joint prediction models for the secondary endpoint event patient group in an embodiment of the present disclosure.

[0123] like Figure 3-5 As shown in the figures, the optimal radiomics-clinical characteristics joint prediction model in the cardiovascular death patient group was the model based on logistic regression training, the optimal radiomics-clinical characteristics joint prediction model in the heart failure hospitalized patient group was the model based on K nearest neighbor training, and the optimal radiomics-clinical characteristics joint prediction model in the secondary endpoint event patient group was the model based on naive Bayes training.

[0124] This paper innovatively combines machine learning with imaging omics and clinical characteristics and applies them to the prediction of endpoint events in HCM patients, improving the prediction efficiency of multiple endpoint events in HCM patients. The prediction accuracy of the established prediction model is better than that of the model established by using only clinical prediction indicators. At the same time, this paper focuses on important endpoint events such as cardiovascular death and heart failure hospitalization in HCM patients, which have not been given special attention in previous studies, and finds the potential value of LGE imaging omics features in predicting these endpoint events, providing a more comprehensive way to assess the risk of endpoint events in HCM patients.

[0125] Figure 6 A flowchart of a method for predicting SCD risk based on an HCM patient endpoint event prediction model according to an embodiment of the present disclosure is shown. The method is implemented by a computer and includes steps S601 to S606.

[0126] In step S601, an HCM patient endpoint event group and an HCM patient control group are obtained, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively.

[0127] In step S602, the first clinical characteristic predictor of the cardiovascular death patient group is obtained according to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group; the second clinical characteristic predictor of the heart failure hospitalized patient group is obtained according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group; the third clinical characteristic predictor of the secondary endpoint event patient group is obtained according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group; wherein, the first clinical characteristic predictor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic predictor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic predictor includes: age, NYHA heart function class.

[0128] In step S603, based on the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, the post-screening imaging genomics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group are obtained; and the imaging genomics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group are calculated based on the corresponding post-screening imaging genomics features.

[0129] In step S604, for each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: based on the imaging feature scores of each patient group and the corresponding clinical feature prediction factors, a plurality of corresponding imaging feature-clinical feature joint prediction models are established using a variety of machine learning methods; the plurality of imaging feature-clinical feature joint prediction models are verified, and the optimal imaging feature-clinical feature joint prediction model is selected as the HCM patient endpoint event prediction model for each patient group.

[0130] In step S605, a data group of HCM patients to be tested is obtained, wherein the data group of HCM patients to be tested includes data of multiple HCM patients to be tested.

[0131] In step S606, the HCM patient data group to be tested is input into the HCM patient endpoint event prediction model of the cardiovascular death patient group to obtain the corresponding SCD risk prediction result.

[0132] In the present disclosure, the SCD risk of HCM patients is predicted by utilizing the endpoint event prediction model of the cardiovascular death patient group. The model can output a risk score or probability value for each patient to assist doctors in assessing the patient's risk status and formulating corresponding prevention and treatment strategies.

[0133] Figure 7 A structural block diagram of a device for establishing an HCM patient endpoint event prediction model based on LGE radiomics features and clinical prediction features according to an embodiment of the present disclosure is shown. The device can be implemented as part or all of an electronic device through software, hardware, or a combination of both.

[0134] like Figure 7 As shown, the device 700 includes: a first acquisition module 701, a second acquisition module 702, a first calculation module 703 and a first model establishment selection module 704.

[0135] The first acquisition module 701 is configured to acquire an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively;

[0136] The second acquisition module 702 is configured to acquire the first clinical characteristic prediction factor of the cardiovascular death patient group according to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group; acquire the second clinical characteristic prediction factor of the heart failure hospitalized patient group according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group; acquire the third clinical characteristic prediction factor of the secondary endpoint event patient group according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group; wherein the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class;

[0137] The first calculation module 703 is configured to obtain the post-screening radiomics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group; and calculate the radiomics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the corresponding post-screening radiomics features;

[0138] The first model establishment selection module 704 is configured to: for each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: based on the imaging feature scores of each patient group and the corresponding clinical feature predictors, use multiple machine learning methods to establish corresponding multiple imaging feature joint prediction models; verify the multiple imaging feature joint prediction models, and select the optimal imaging feature joint prediction model as the HCM patient endpoint event prediction model for each patient group.

[0139] Figure 8 A structural block diagram of a device for predicting SCD risk based on an HCM patient endpoint event prediction model according to an embodiment of the present disclosure is shown, wherein the device can be implemented as part or all of an electronic device through software, hardware, or a combination of both.

[0140] like Figure 8 As shown, the device 800 includes: a third acquisition module 801 , a fourth acquisition module 802 , a second calculation module 803 , a second model establishment selection module 804 , a data acquisition module 805 and an SCD prediction module 806 .

[0141] The third acquisition module 801 is configured to acquire an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively.

[0142] The fourth acquisition module 802 is configured to obtain the first clinical characteristic prediction factor of the cardiovascular death patient group based on the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group; obtain the second clinical characteristic prediction factor of the heart failure hospitalized patient group based on the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group; obtain the third clinical characteristic prediction factor of the secondary endpoint event patient group based on the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group; wherein, the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class.

[0143] The second calculation module 803 is configured to obtain the post-screening imaging genomics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group; and calculate the imaging genomics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the corresponding post-screening imaging genomics features.

[0144] The second model establishment selection module 804 is configured to: for each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: based on the imaging feature scores of each patient group and the corresponding clinical feature predictors, use multiple machine learning methods to establish corresponding multiple imaging feature joint prediction models; verify the multiple imaging feature joint prediction models, and select the optimal imaging feature joint prediction model as the HCM patient endpoint event prediction model for each patient group.

[0145] The data acquisition module 805 is configured to acquire a data group of HCM patients to be tested, wherein the data group of HCM patients to be tested includes data of multiple HCM patients to be tested.

[0146] The SCD prediction module 806 is configured to input the HCM patient data group to be tested into the HCM patient endpoint event prediction model of the cardiovascular death patient group to obtain the corresponding SCD risk prediction result.

[0147] The present disclosure also discloses an electronic device, Fig. 9 A structural block diagram of an electronic device according to an embodiment of the present disclosure is shown.

[0148] like Fig. 9 As shown, the electronic device includes a memory and a processor, wherein the memory is used to store computer instructions, wherein the computer instructions are executed by the processor to implement the method according to an embodiment of the present disclosure.

[0149] The method is:

[0150] Obtain an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively.

[0151] According to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group, the first clinical characteristic prediction factor of the cardiovascular death patient group is obtained; according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group, the second clinical characteristic prediction factor of the heart failure hospitalized patient group is obtained; according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group, the third clinical characteristic prediction factor of the secondary endpoint event patient group is obtained; wherein, the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class;

[0152] According to the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, obtaining the post-screening imaging omics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group; calculating the imaging omics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group according to the corresponding post-screening imaging omics features;

[0153] For each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: based on the imaging feature scores of each patient group and the corresponding clinical feature predictors, a variety of machine learning methods are used to establish corresponding multiple imaging-clinical feature joint prediction models; the multiple imaging-clinical feature joint prediction models are verified, and the optimal imaging-clinical feature joint prediction model is selected as the HCM patient endpoint event prediction model for each patient group.

[0154] According to an embodiment of the present disclosure, a data group of HCM patients to be tested is obtained, wherein the data group of HCM patients to be tested includes data of multiple HCM patients to be tested; the data group of HCM patients to be tested is input into an endpoint event prediction model of HCM patients in the cardiovascular death patient group to obtain a corresponding SCD risk prediction result.

[0155] Fig.10 A schematic diagram showing the structure of a computer system suitable for implementing the method according to an embodiment of the present disclosure is shown.

[0156] like Fig.10 As shown, the computer system includes a processing unit, which can perform the various methods in the above-mentioned embodiments according to the program stored in the read-only memory (ROM) or the program loaded from the storage part into the random access memory (RAM). In the RAM, various programs and data required for the operation of the computer system are also stored. The processing unit, ROM and RAM are connected to each other through a bus. The input / output (I / O) interface is also connected to the bus.

[0157] The following components are connected to the I / O interface: an input part including a keyboard, a mouse, etc.; an output part including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker; a storage part including a hard disk, etc.; and a communication part including a network interface card such as a LAN card, a modem, etc. The communication part performs a communication process via a network such as the Internet. The drive is also connected to the I / O interface as needed. Removable media, such as disks, optical disks, magneto-optical disks, semiconductor memories, etc., are installed on the drive as needed so that the computer program read therefrom is installed into the storage part as needed. Among them, the processing unit can be implemented as a processing unit such as a CPU, a GPU, a TPU, an FPGA, an NPU, etc.

[0158] In particular, according to an embodiment of the present disclosure, the method described above can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program tangibly contained on a machine-readable medium, and the computer program includes a program code for executing the above method. In such an embodiment, the computer program can be downloaded and installed from a network through a communication part, and / or installed from a removable medium.

[0159] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flowchart, and the combination of boxes in the block diagram and / or flowchart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0160] The units or modules involved in the embodiments described in the present disclosure may be implemented by software or programmable hardware. The units or modules described may also be set in a processor, and the names of these units or modules do not constitute limitations on the units or modules themselves in some cases.

[0161] As another aspect, the present disclosure further provides a computer-readable storage medium, which may be a computer-readable storage medium included in the electronic device or computer system in the above embodiment; or a computer-readable storage medium that exists independently and is not assembled into a device. The computer-readable storage medium stores one or more programs, and the programs are used by one or more processors to execute the method described in the present disclosure.

[0162] The above description is only a preferred embodiment of the present disclosure and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the present disclosure is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present disclosure (but not limited to) by each other.

Claims

1. A method for establishing a prediction model for endpoint events in patients with hypertrophic cardiomyopathy (HCM) based on delayed gadolinium-enhanced LGE imaging features and clinical prediction features, characterized in that: The method is implemented by a computer and comprises: Obtaining an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively; According to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group, the first clinical characteristic prediction factor of the cardiovascular death patient group is obtained; according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group, the second clinical characteristic prediction factor of the heart failure hospitalized patient group is obtained; according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group, the third clinical characteristic prediction factor of the secondary endpoint event patient group is obtained; wherein, the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class; According to the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, obtaining the post-screening imaging omics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group; calculating the imaging omics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group according to the corresponding post-screening imaging omics features; For each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: based on the imaging feature scores of each patient group and the corresponding clinical feature predictors, a variety of machine learning methods are used to establish corresponding multiple imaging-clinical feature joint prediction models; the multiple imaging-clinical feature joint prediction models are verified, and the optimal imaging-clinical feature joint prediction model is selected as the HCM patient endpoint event prediction model for each patient group.

2. The method according to claim 1, characterized in that: According to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group, obtaining the first clinical characteristic prediction factor of the cardiovascular death patient group, including: inputting the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group into a multivariate survival analysis model using a stepwise regression method to obtain the first clinical characteristic prediction factor; According to the baseline clinical prediction characteristics of the heart failure inpatient group and the second HCM patient control group, obtaining a second clinical characteristic prediction factor of the heart failure inpatient group, including: inputting the baseline clinical prediction characteristics of the heart failure inpatient group and the second HCM patient control group into a multivariate survival analysis model using a stepwise regression method to obtain the second clinical characteristic prediction factor; According to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group, obtaining the third clinical characteristic prediction factor of the secondary endpoint event patient group, including: inputting the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group into a multivariate survival analysis model using a stepwise regression method to obtain the third clinical characteristic prediction factor; The baseline clinical prediction features include: basic patient information, echocardiogram information, electrocardiogram information, and basic cardiac magnetic resonance imaging parameters.

3. The method according to claim 2, characterized in that: The multivariate survival analysis model is any one or more of the following: Cox proportional hazards model, Weibull distribution model, competing risks model; The stepwise regression method includes any one or more of the following: Forward stepwise regression, backward stepwise regression, and bidirectional stepwise regression.

4. The method according to claim 2, characterized in that: The HCM endpoint events of the HCM patients in the cardiovascular death patient group include sudden cardiac death and / or death from heart failure; The HCM endpoint events of HCM patients in the group of patients hospitalized with heart failure included hospitalization for heart failure; The HCM endpoint events of HCM patients in the secondary endpoint event patient group include any one or more of the following: unplanned readmission due to cardiovascular disease, unplanned readmission due to other reasons, and death from non-cardiovascular disease; No sudden cardiac death or death from heart failure occurred in the HCM patients in the first HCM patient control group; No HCM patients in the second HCM patient control group were hospitalized for heart failure; HCM patients in the third HCM patient control group did not experience unplanned readmission due to cardiovascular disease, unplanned readmission due to other reasons, and non-cardiovascular disease death.

5. The method according to claim 1, characterized in that The method of obtaining the post-screening imaging omics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the LGE image group of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group includes: For each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, an LGE image group of each patient group is obtained, wherein the LGE image group includes LGE images of HCM patients in each patient group; the left ventricular myocardial short-axis section of the LGE image in the LGE image group is used as a region of interest, and the pre-screening imaging omics features of each patient group are extracted from the region of interest; The pre-screening radiomics features of each patient group were input into the least absolute shrinkage and selection operator LASSO model for correlation analysis to obtain the post-screening radiomics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group.

6. The method according to claim 5, characterized in that Calculating the imaging feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group according to the corresponding post-screening imaging feature comprises: According to the characteristic distribution and LASSO coefficient of the corresponding post-screening imaging features, the imaging feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group are calculated respectively.

7. The method according to claim 1, characterized in that The machine learning method includes any one of the following: Logistic regression, K-nearest neighbor method, elastic network method, naive Bayes method.

8. The method according to claim 1, characterized in that: The verifying of the multiple radiomics-clinical features joint prediction models and selecting the optimal radiomics-clinical features joint prediction model comprises: The areas under the ROC curves of the multiple radiomics-clinical feature joint prediction models are calculated, and the radiomics-clinical feature joint prediction model with the largest area under the ROC curve is taken as the optimal radiomics-clinical feature joint prediction model.

9. A method for predicting the risk of sudden cardiac death (SCD) based on an endpoint event prediction model for HCM patients, characterized in that: The method is implemented by a computer and comprises: Obtaining an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively; According to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group, the first clinical characteristic prediction factor of the cardiovascular death patient group is obtained; according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group, the second clinical characteristic prediction factor of the heart failure hospitalized patient group is obtained; according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group, the third clinical characteristic prediction factor of the secondary endpoint event patient group is obtained; wherein, the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class; According to the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group, obtaining the post-screening imaging omics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group; calculating the imaging omics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group according to the corresponding post-screening imaging omics features; For each patient group in the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: according to the imaging omics feature scores of each patient group and the corresponding clinical feature predictors, a plurality of corresponding imaging omics-clinical feature joint prediction models are established by using a plurality of machine learning methods; the plurality of imaging omics-clinical feature joint prediction models are verified, and the optimal imaging omics-clinical feature joint prediction model is selected as the HCM patient endpoint event prediction model for each patient group; Acquire a data group of HCM patients to be tested, wherein the data group of HCM patients to be tested includes data of multiple HCM patients to be tested; The HCM patient data group to be tested is input into the HCM patient endpoint event prediction model of the cardiovascular death patient group to obtain the corresponding SCD risk prediction result.

10. The method according to claim 9, characterized in that The machine learning method includes any one of the following: Logistic regression, K-nearest neighbor method, elastic network method, naive Bayes method.

11. A device for establishing a prediction model for endpoint events in HCM patients based on LGE radiomics features and clinical prediction features, characterized in that: The device is arranged in a computer and comprises: A first acquisition module is configured to acquire an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively; The second acquisition module is configured to acquire the first clinical characteristic prediction factor of the cardiovascular death patient group according to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group; acquire the second clinical characteristic prediction factor of the heart failure hospitalized patient group according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group; acquire the third clinical characteristic prediction factor of the secondary endpoint event patient group according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group; wherein the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class; The first calculation module is configured to obtain the post-screening imaging omics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group; and calculate the imaging omics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the corresponding post-screening imaging omics features; The first model establishment selection module is configured to: for each patient group among the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: based on the imaging feature scores of each patient group and the corresponding clinical feature predictors, use multiple machine learning methods to establish corresponding multiple imaging feature joint prediction models; verify the multiple imaging feature joint prediction models, and select the optimal imaging feature joint prediction model as the HCM patient endpoint event prediction model for each patient group.

12. A device for predicting SCD risk based on an endpoint event prediction model for HCM patients, characterized in that: The device is arranged in a computer and comprises: A third acquisition module is configured to acquire an HCM patient endpoint event group and an HCM patient control group, wherein the HCM patient endpoint event group includes a cardiovascular death patient group, a heart failure hospitalized patient group, and a secondary endpoint event patient group, and the HCM patient control group includes a first HCM patient control group, a second HCM patient control group, and a third HCM patient control group corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group, respectively; The fourth acquisition module is configured to acquire the first clinical characteristic prediction factor of the cardiovascular death patient group according to the baseline clinical prediction characteristics of the cardiovascular death patient group and the first HCM patient control group; acquire the second clinical characteristic prediction factor of the heart failure hospitalized patient group according to the baseline clinical prediction characteristics of the heart failure hospitalized patient group and the second HCM patient control group; acquire the third clinical characteristic prediction factor of the secondary endpoint event patient group according to the baseline clinical prediction characteristics of the secondary endpoint event patient group and the third HCM patient control group; wherein the first clinical characteristic prediction factor includes: left atrial diameter, family history of sudden cardiac death, the second clinical characteristic prediction factor includes: NYHA heart function class, non-sustained ventricular tachycardia, and the third clinical characteristic prediction factor includes: age, NYHA heart function class; The second calculation module is configured to obtain the post-screening imaging omics features corresponding to the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the LGE images of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group; and calculate the imaging omics feature scores of the cardiovascular death patient group, the heart failure hospitalized patient group, and the secondary endpoint event patient group based on the corresponding post-screening imaging omics features; The second model establishment selection module is configured to: for each patient group of the cardiovascular death patient group, the heart failure hospitalized patient group and the secondary endpoint event patient group: according to the imaging omics feature score of each patient group and the corresponding clinical feature predictor, use multiple machine learning methods to establish corresponding multiple imaging omics-clinical feature joint prediction models; verify the multiple imaging omics-clinical feature joint prediction models, and select the optimal imaging omics-clinical feature joint prediction model as the HCM patient endpoint event prediction model for each patient group; A data acquisition module is configured to acquire a data group of HCM patients to be tested, wherein the data group of HCM patients to be tested includes data of multiple HCM patients to be tested; The SCD prediction module is configured to input the HCM patient data group to be tested into the HCM patient endpoint event prediction model of the cardiovascular death patient group to obtain the corresponding SCD risk prediction result.

13. An electronic device, characterized in that: The method comprises a memory and a processor; wherein the memory is used to store computer instructions, wherein the computer instructions are executed by the processor to implement the method according to any one of claims 1 to 10.

14. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the computer instructions are executed by a processor, the method according to any one of claims 1 to 10 is implemented.

15. A computer program product, comprising computer instructions, which, when executed by a processor, implement the method according to any one of claims 1 to 10.

Images