Method and apparatus for predicting and diagnosing the risk of cardiovascular disease by type

Abstract

A cardiovascular disease type-specific risk prediction and diagnostic device according to one embodiment comprises at least one processor, the at least one processor predicts the risk of cardiovascular disease type using a diagnostic model, the diagnostic model includes a first model that predicts cardiovascular risk by analyzing input fundus images, and a second model that predicts the risk of cardiovascular disease type by receiving input of the cardiovascular risk score and disease type-specific factors predicted by the first model, and the at least one processor may be characterized by providing diagnostic information based on the cardiovascular risk predicted by the first model and the risk of cardiovascular disease type predicted by the second model.

Description

Technical Field

[0001] This application relates to a diagnostic method and apparatus for predicting the type-specific risks of cardiovascular diseases using machine learning that has learned fundus images and cardiovascular biomarkers.

Background Art

[0002] A fundus image (i.e., a fundus image or a retinal image) is an image taken of the back surface of the retina and includes the retina, retinal blood vessels, optic nerve, macula, choroid, etc. Fundus images have conventionally been mainly used for the diagnosis of ophthalmic diseases. However, in recent years, due to the rapid development of artificial intelligence (AI) technology, their scope of application has been greatly expanded to even evaluate the risks of systemic diseases such as cardiovascular diseases. This technology is not only non-invasive and cost-effective but is also recognized as a very useful tool particularly in the process of disease screening for large population groups. AI has demonstrated excellent performance in analyzing changes in retinal blood vessels related to cardiovascular diseases, thereby forming a new paradigm in the field of medical image analysis. As technologies related to methods for assisting in the diagnosis of cardiovascular diseases using existing fundus images, the applicant has filed Korean patent applications such as Application Nos. 10-2018-0166720, 10-2018-0166721, 10-2018-0166722, etc. The said patent applications are technologies that provide auxiliary information related to the integrated risk assessment for cardiovascular diseases, focusing on more accurately predicting the risks of cardiovascular diseases through fundus images and supporting clinical decisions based thereon.

[0003] On the other hand, cardiovascular diseases include detailed categories such as myocardial infarction (MI), stroke, peripheral artery disease (PAD), heart failure (HF), and atrial fibrillation (AF). While the applicant's existing patent application mentioned earlier aims to predict the overall risk of cardiovascular disease, if it were possible to predict the individual risk for each detailed category of cardiovascular disease, this would provide even more detailed supplementary information. This could provide important support for healthcare professionals to make more precise clinical decisions based on each patient's condition. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Korean Patent Application No. 10-2018-0166720 [Patent Document 2] Korean Patent Application No. 10-2018-0166721 [Patent Document 3] Korean Patent Application No. 10-2018-0166722 [Overview of the project] [Problems that the invention aims to solve]

[0005] One problem that the present invention aims to solve is to provide a method and an apparatus that implements this method, which can predict the risk of a patient's type of cardiovascular disease by receiving an input of a patient's fundus image, using a machine learning model that has learned the correlation between fundus images and the risk of different types of cardiovascular disease. Specifically, the present invention aims to provide an apparatus that predicts the risk of different types of cardiovascular disease by additionally considering disease type-specific factors, including age, sex, blood biomarkers, smoking status, and whether or not the patient is obese, in an integrated cardiovascular risk score based on features extracted from fundus images by a machine learning model.

[0006] One problem that this invention aims to solve is to provide a device that predicts the risk of different types of cardiovascular disease by additionally considering the type of cardiovascular plaque in the target patient, i.e., whether it is soft plaque or hard plaque.

[0007] One problem that the present invention aims to solve is to provide a cardiovascular disease type-specific risk prediction device that includes a model for determining what clinical tests are recommended for a patient based on the risk of each type of cardiovascular disease.

[0008] The problems that this application seeks to solve are not limited to those described above, and any problems not mentioned can be clearly understood by a person with ordinary skill in the art to which this application pertains from this specification and the accompanying drawings. [Means for solving the problem]

[0009] According to one aspect of this application, a cardiovascular disease type-specific risk prediction and diagnostic device according to one embodiment includes: a first model (cardiovascular risk prediction model) that analyzes input fundus images to predict cardiovascular risk; and a second model (cardiovascular disease type-specific risk prediction model) that receives additional input of the cardiovascular risk score and disease type-specific factors predicted by the first model to predict the risk of cardiovascular disease type; and provides each cardiovascular disease type-specific risk grade and related information by combining the cardiovascular risk predicted by the first model and the risk of cardiovascular disease type predicted by the second model.

[0010] According to one aspect of this application, a cardiovascular disease type-specific risk prediction and diagnostic device according to one embodiment includes: a first model that analyzes input fundus images to predict cardiovascular risk; a second model that receives input of the cardiovascular risk score and disease type-specific factors predicted by the first model to predict the risk of cardiovascular disease type; and a third model (recommended test judgment model) that determines recommended clinical tests for the patient based on the cardiovascular disease type-specific risk predicted by the second model. The device provides each cardiovascular disease type-specific risk grade and related information, as well as recommended test-related information, by comprehensively combining the cardiovascular risk predicted by the first model and the cardiovascular disease type-specific risk predicted by the second model.

[0011] According to one aspect of this application, a cardiovascular disease type-specific risk prediction and diagnostic device according to one embodiment includes: a first model that analyzes input fundus images to predict cardiovascular risk and cardiovascular plaque type; a second model that receives additional input of the cardiovascular risk score and disease type-specific factors predicted by the first model to predict the risk of cardiovascular disease type; and a third model that determines recommended clinical tests for the patient based on the cardiovascular disease type-specific risk predicted by the second model. The device provides each cardiovascular disease type-specific risk grade and related information, as well as recommended test-related information, by comprehensively combining the cardiovascular risk and plaque type predicted by the first model and the cardiovascular disease type-specific risk predicted by the second model. [Effects of the Invention]

[0012] According to the present invention, by using a machine learning model that has learned the correlation between fundus images and the risk of different types of cardiovascular disease, it is possible to receive fundus images from a target patient and predict not only the patient's integrated cardiovascular risk score but also the risk of different types of cardiovascular disease. Furthermore, the machine learning model can predict the type of cardiovascular plaque in the patient. This makes it possible to formulate an individualized treatment plan optimized for each patient, and to identify patients at high risk of specific cardiovascular diseases early and strengthen preventive measures.

[0013] The effects of this application are not limited to those described above, and any effects not mentioned can be clearly understood by a person with ordinary skill in the art to which this application pertains from this specification and the accompanying drawings. [Brief explanation of the drawing]

[0014] [Figure 1] This is a block diagram illustrating a diagnostic device according to one embodiment. [Figure 2] This diagram illustrates how cardiovascular risk (Reti-CVD) can serve as a biomarker for predicting the individual risks of cardiovascular stroke, myocardial infarction, and atrial fibrillation. [Figure 3] This figure illustrates the correlation between Reti-CVD and various types of cardiovascular disease. [Figure 4] This figure illustrates the correlation between Reti-CVD and various types of cardiovascular disease. [Figure 5] This figure illustrates the cardiovascular disease prediction performance for diabetic patients according to this embodiment. [Figure 6] This figure illustrates the cardiovascular disease prediction performance for diabetic patients according to this embodiment. [Figure 7] This is a diagram illustrating a method for predicting the risk of different types of cardiovascular disease according to one embodiment. [Figure 8] This figure illustrates the predictive performance of atrial fibrillation risk according to one embodiment. [Figure 9]It is a diagram for explaining a method for predicting risks by type of cardiovascular disease according to an embodiment. [Figure 10] It is a diagram for explaining a method for predicting risks by type of cardiovascular disease according to another embodiment.

Embodiments for Carrying Out the Invention

[0015] The above-mentioned objects, features, and advantages of the present application will become clearer through the following detailed description related to the attached drawings. However, since the present application can be modified in various ways and can have various embodiments, specific embodiments will be illustrated in the drawings and described in detail below.

[0016] Throughout the specification, the same reference numerals generally indicate the same components. Also, components with the same functions within the scope of the same concept appearing in the drawings of each embodiment will be described using the same reference signs, and redundant descriptions thereof will be omitted.

[0017] When it is determined that a specific description of a known function or configuration related to the present application may unnecessarily obscure the gist of the present application, the detailed description thereof will be omitted. Also, the numbers (for example, the first, the second, etc.) used in the description process of this specification are merely identification symbols for distinguishing one component from another component.

[0018] Also, the suffixes "module" and "unit" for the components used in the following embodiments are given or mixed only for ease of specification writing, and do not have meanings or roles that distinguish them from each other by themselves.

[0019] In the following embodiments, the singular form includes the plural form unless the context clearly indicates a different meaning.

[0020] In the following embodiments, terms such as "includes" or "has" mean that the features or components described in the specification are present, and do not preclude the possibility of the addition of one or more other features or components.

[0021] In drawings, components may be shown with their size exaggerated or reduced for illustrative purposes. For example, the size and thickness of each component shown in the drawings are arbitrarily shown for illustrative purposes and the present invention is not necessarily limited to those shown.

[0022] Where a particular embodiment can be implemented separately, the order of certain processes may differ from the order described. For example, two processes described consecutively may occur substantially simultaneously, or they may proceed in the reverse order of the description.

[0023] In the following embodiments, when we say that components are connected, this includes not only cases where components are directly connected, but also cases where components are indirectly connected with other components interposed between them. For example, when we say that components are electrically connected in this specification, this includes not only cases where components are directly electrically connected, but also cases where components are indirectly electrically connected with other components interposed between them.

[0024] The following describes diagnostic devices and methods for predicting the risk of different types of cardiovascular disease based on eye images to assist healthcare professionals in their decision-making. In this specification, the term "diagnosis" may mean a diagnostic aid that assists in the diagnosis of a disease, rather than directly diagnosing the disease itself. For the sake of clarity, the term "diagnosis" is used below, but it may refer to a diagnostic aid. The machine learning models described herein may be designed based on various machine learning libraries.

[0025] Cardiovascular diseases include hypertension, atherosclerosis, coronary artery disease (CAD), valvular heart disease, pericardial disease, myocarditis, cardiomyopathy, arrhythmia, heart failure (HF), rheumatic heart disease, congenital heart disease, aortic aneurysm, peripheral vascular disease, cerebrovascular disease, and pulmonary vascular disease. More specifically, these include angina pectoris, myocardial infarction (MI), and pulmonary artery hypertension. Hypertension (PAH), Cor Pulmonale, Left Heart Failure, Right Heart Failure, Aortic Valve Stenosis (AS), Aortic Valve Regurgitation (AR), Pulmonary Valve Stenosis (PS), Pulmonary Valve Regurgitation (PR), Tricuspid Valve Stenosis (TS), Tricuspid Valve Regurgitation (TR), Mitral Valve Stenosis (MS), Mitral Valve Regurgitation, Peripheral Artery Disease Diseases (PAD), aortic aneurysm (e.g., abdominal aortic aneurysm, thoracic aortic aneurysm), pericarditis, stroke, cerebral infarctionCardiovascular disease may include at least one of the following: infarction, cerebral hemorrhage, cerebral aneurysm, thrombotic disease (e.g., deep vein thrombosis, pulmonary embolism), atrial fibrillation (AF), atrial flutter (AT), paroxysmal supraventricular tachycardia (PSVT), ventricular tachycardia (VT), bradyarrhythmia, ischemic stroke, hemorrhagic stroke, or transient ischemic attack (TIA). Cardiovascular disease may include complications. Furthermore, complications may include heart attack, death due to cardiovascular disease, heart attack, cardiogenic shock, kidney disease, aspiration pneumonia, dysphagia, decreased motor function, decreased language function, cognitive decline, sleep disorders, emotional disorders, neuropathic pain, urinary tract infection, malnutrition, deep vein thrombosis, pressure ulcers, falls, pain, seizures, depression, etc. Additionally, detailed subtypes of cardiovascular disease may be included. For example, heart failure can be classified into heart failure with reduced left ventricular ejection fraction (HFrEF), heart failure with mildly reduced left ventricular ejection fraction (HFmrEF), and heart failure with preserved left ventricular ejection fraction (HFpEF). Stroke can be classified into ischemic stroke, thrombotic cerebral infarction (cerebral thrombosis), cerebral embolism (embolic cerebral infarction), lacunar infarction, hemorrhagic stroke, intracerebral hemorrhage, subarachnoid hemorrhage, transient ischemic attack, brainstem stroke, stroke of unknown cause, etc. Atrial fibrillation can be classified into detailed subtypes including paroxysmal atrial fibrillation, persistent atrial fibrillation, long-term persistent atrial fibrillation, and permanent atrial fibrillation. Myocardial infarction can be classified into detailed subtypes including type 1 (spontaneous myocardial infarction), type 2 (myocardial infarction due to ischemic disequilibrium), type 3 (myocardial infarction resulting in death when biomarker values ​​are unavailable), type 4a (myocardial infarction associated with percutaneous coronary intervention (PCI)), type 4b (myocardial infarction associated with stent thrombosis), and type 5 (myocardial infarction associated with coronary artery bypass grafting (CABG)).

[0026] 1. Biomarkers for detailed cardiovascular disease types

[0027] Fundus imaging provides a non-invasive method for observing the vascular structure of the human body, thereby establishing a foundation for predicting the likelihood of developing cardiovascular disease (CVD). Machine learning models trained on fundus images and cardiovascular biomarkers are used to predict the risk of future cardiovascular disease. Key cardiovascular biomarkers include coronary artery calcium (CAC) score, pooled cohort equation (PCE) score, QRISK score, modified Framingham score (FRS), carotid intima-media thickness (CIMT) score, brachial-ankle pulse wave velocity (baPWV) score, and ankle-brachial index (ABI). The machine learning model generates a cardiovascular risk score, Reti-CVD, for input fundus images. Based on this score, individuals are classified into three groups: low risk, medium risk, and high risk. This classification criterion uses the baseline values ​​established in the applicant's previous study, "Deep-learning-based cardiovascular risk stratification using coronary artery calcium scores predicted from retinal photographs," The Lancet Digital Health 3(5). Through previous research, the applicant demonstrated that the Reti-CVD score can be used as a biomarker to predict the risk of future cardiovascular disease. In this invention, the Reti-CVD score is newly revealed to be closely associated with the risk of various types of future cardiovascular disease, and a detailed algorithm is provided to predict the risk of different types of cardiovascular disease based on the Reti-CVD score.

[0028] The diagnostic device can use machine learning models to predict various types of information and risks as defined herein, such as cardiovascular risk, risk by type of cardiovascular disease, and risk by subtype of cardiovascular disease.

[0029] Figure 1 is a block diagram illustrating a diagnostic device according to one embodiment. Referring to Figure 1, the diagnostic device (10) may include a processor (12), a storage module (11), and a communication module (13).

[0030] The processor (12) may include one or more of the following: a CPU (Central Processing Unit), RAM (Random Access Memory), a GPU (Graphics Processing Unit), one or more microprocessors, and other electronic components capable of processing input data according to a predetermined logic. Furthermore, the processor (12) may consist of at least one component.

[0031] The processor (12) can read system programs and various processing programs stored in the memory module (11). For example, the processor (12) can load processes, methods, etc. for performing the diagnosis described later onto RAM and perform various processing according to the loaded programs. For example, the processor (12) can process the algorithms described herein. The processor (12) can predict cardiovascular risk using machine learning models. Furthermore, the processor (12) can implement various methods described herein.

[0032] The memory module (11) can store diagnostic models. The memory module (11) can store machine learning models, parameters of machine learning models, variables, algorithms described herein, etc.

[0033] The memory module (11) can be embodied in non-volatile semiconductor memory, hard disk, flash memory, RAM, ROM (Read Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), or other tangible non-volatile recording media.

[0034] The memory module (11) can store various processing programs, parameters for processing programs, or processing result data. For example, the memory module (11) can store programs for performing diagnostics described later, parameters, and data obtained by performing such programs. The memory module (11) can also store various machine learning models described later.

[0035] Although not shown in the diagram, the diagnostic device (10) may further include an input module. The input module can acquire user input. For example, the input module can acquire user input. The input module can also acquire the subject's physical information as described herein (at least one of height, weight, age, sex, race, smoking status, blood pressure (e.g., blood pressure value, presence or absence of hypertension), presence or absence of diabetes (or blood glucose value), cholesterol value), etc. The input module can also acquire disease type-specific factors as described later. The processor (12) can acquire the information input through the input module.

[0036] The communication module (13) can communicate with various devices (e.g., learning devices, client devices, etc.). For example, the diagnostic device (10) can be implemented in a server configuration that communicates with a client device. Alternatively, the diagnostic device (10) can be implemented in a client device configuration.

[0037] <Example 1 - Prediction of stroke, myocardial infarction, and atrial fibrillation risk>

[0038] Figure 2 illustrates how cardiovascular risk (Reti-CVD) can serve as a biomarker for predicting the individual risks of cardiovascular stroke, myocardial infarction, and atrial fibrillation. (a) shows the incidence rates of stroke, myocardial infarction, and atrial fibrillation in low-risk, intermediate-risk, and high-risk groups according to the Reti-CVD score. (b) shows the adjusted hazard ratio (HR), HR trend, and C-index for each risk group. Here, CI represents the confidence interval.

[0039] (1) Stroke: The Reti-CVD score showed a significant correlation with the risk of stroke. Analysis revealed that the risk of stroke increased as the Reti-CVD score increased (adjusted HR trend, 1.50; 95% CI, 1.16-1.95; p=0.002). This suggests that the Reti-CVD score can be used as a useful biomarker to identify high-risk patient groups requiring early preventive intervention.

[0040] (2) Myocardial Infarction (MI): The Reti-CVD score showed a significant correlation with the risk of myocardial infarction (MI). As the Reti-CVD score increased, the risk of MI also increased (adjusted HR trend, 1.22; 95% CI, 1.01-1.47; p=0.036). This score can be used to identify individuals at high risk of myocardial infarction early and to strengthen preventive measures.

[0041] (3) Atrial Fibrillation (AF): The Reti-CVD score was closely associated with the risk of developing atrial fibrillation (AF) (adjusted HR trend, 1.32; 95% CI, 1.15-1.51; p<0.001). This suggests that the Reti-CVD score may be useful in assessing the risk of developing atrial fibrillation, and that Reti-CVD could be an important tool for early risk assessment and the establishment of preventive management strategies for atrial fibrillation.

[0042] <Example 2 - Prediction of risk for 14 cardiovascular disease types, including myocardial infarction and peripheral artery disease>

[0043] This study analyzed the correlation between the Reti-CVD score and 13 different cardiovascular diseases (CVD) and arterial hypertension. A cross-sectional analysis was conducted using 45,980 participants in the UK Biobank. Logistic regression analysis was used to identify differential correlations with various cardiovascular diseases, and factors such as hypertension, diabetes, dyslipidemia, and smoking were adjusted for during the analysis.

[0044] Figures 3 and 4 illustrate the correlation between Reti-CVD and various cardiovascular disease types. Referring to Figures 3 and 4, (a) and (b) are tables illustrating the relationship between Reti-CVD scores and the risk of different cardiovascular disease types. In (a) and (b), OR represents the odds ratio, LCL represents the lower control limit, and UCL represents the upper control limit. In these tables, a significant association was confirmed between the high-risk group due to Reti-CVD and the low-risk group regarding specific cardiovascular diseases.

[0045] (1) Coronary artery disease or myocardial infarction (MI): The higher the Reti-CVD score, the significantly increased the risk of developing coronary artery disease. In the high-risk group, the odds ratio (OR) was 10.37 (95% CI, 7.58-14.18), and a very significant increase was observed in the overall trend.

[0046] (2) Peripheral vascular disease or peripheral artery disease (PAD): As the Reti-CVD score increased, the risk of developing peripheral vascular disease also increased. The odds ratio for the high-risk group was 9.65 (95% CI, 2.94-31.64), indicating a statistically significant correlation.

[0047] (3) Atrial fibrillation (AF): A strong correlation was observed between the Reti-CVD score and the risk of developing atrial fibrillation, with the odds ratio for the high-risk group being 9.36 (95% CI, 6.51-13.45).

[0048] (4) Aortic valve stenosis (AVS): The higher the Reti-CVD score, the greater the risk of developing aortic valve stenosis, and the odds ratio for the high-risk group was shown to be 8.13 (95% CI, 1.87-35.35).

[0049] (5) Other diseases: Significant correlations with the Reti-CVD score were also confirmed in various cardiovascular diseases such as stroke, heart failure (HF), and pulmonary embolism (PE).

[0050] Examples 1 and 2 focused on analyzing the correlation between various cardiovascular disease types and associated conditions using the Reti-CVD score, and predicting the risk for each type. Such results can be used to detect the potential presence of cardiovascular disease that patients are currently unaware of and to assess its risk. Specifically, based on the correlation data between the Reti-CVD score and various CVDs, the diagnostic device can detect cardiovascular disease that may be present but undiagnosed by the patient. This is useful for discovering conditions where patients do not clearly experience symptoms or where the disease was not diagnosed in its early stages. Furthermore, based on the research findings, the diagnostic device can analyze the Reti-CVD score with the patient's existing health data and recommend the most likely cardiovascular disease for the patient as a priority. This system can guide patients and healthcare professionals to consider additional tests for potential risk conditions, promoting early diagnosis and preventive treatment. For example, if a patient shows a high risk associated with atrial fibrillation (AF) according to their Reti-CVD score, it is highly likely that the patient has undiagnosed atrial fibrillation, and the diagnostic device can recommend additional electrocardiograms or monitoring. The diagnostic device can guide other patients who show a moderate risk associated with peripheral vascular disease (PVD) or peripheral artery disease (PAD) to undergo more intensive testing for this condition, which they are more likely to develop.

[0051] 2. Cardiovascular prediction biomarkers for patients with pre-existing diseases

[0052] Reti-CVD can be a cardiovascular risk index that can more accurately predict the risk of secondary cardiovascular-related diseases by taking into account the patient's sex, race, and pre-existing disease status. This score can be derived through a model trained on fundus images and coronary calcium (CAC) data. While the Reti-CVD score basically assesses the risk of cardiovascular disease by analyzing the patient's fundus images, this can be extended to develop into a predictive tool that comprehensively considers the patient's pre-existing disease status. Through this, the risk of secondary cardiovascular diseases can be more accurately predicted in high-risk patients, such as those with chronic diseases like diabetes, hypertension, and hyperlipidemia, or those who have already experienced cardiovascular disease (heart failure, coronary artery disease, stroke, etc.). For example, the pattern of cardiovascular disease incidence may differ depending on sex (male / female) and race. By additionally reflecting such demographic variables in the Reti-CVD score, cardiovascular risk can be individually predicted for specific sexes and races. Also, patients with chronic diseases such as hypertension, hyperlipidemia, and diabetes may have an increased risk of additional cardiovascular diseases. The Reti-CVD-based risk prediction model according to the present invention can predict the risk of patients developing secondary cardiovascular diseases (e.g., heart failure, coronary artery disease, stroke, etc.) by taking into account such pre-existing conditions. Patients who already have cardiovascular disease (heart failure, coronary artery disease, stroke, etc.) have an additional risk of developing new types of cardiovascular disease. The Reti-CVD score may be useful in more accurately assessing the risk of developing secondary cardiovascular diseases in such patient groups.

[0053] <Example 1 - Patients with prediabetes and diabetes>

[0054] In this study, using data from the UK Biobank on prediabetes and diabetes, patients were classified into three groups—low-risk, intermediate-risk, and high-risk—based on their Reti-CVD score. Fatal and non-fatal cardiovascular disease types (coronary artery disease, ischemic stroke, and transient ischemic attack) were evaluated while the patients were followed up.

[0055] Figures 5 and 6 illustrate the cardiovascular disease prediction performance for diabetic patients in this embodiment. These figures show the results of classifying patients into low-risk (n=550), medium-risk (n=276), and high-risk (n=275) groups in a 2:1:1 ratio based on the Reti-CVD score, according to the 50th and 75th percentiles. This cardiovascular disease prediction score was derived based on retinal images of prediabetic and diabetic patients. To evaluate how well the Reti-CVD score predicts fatal and non-fatal cardiovascular events, survival analysis was performed using longitudinal data from the UK Biobank, utilizing a Cox proportional hazards model and hazard ratios (HRs). As shown in Figures 5 and 6, 138 out of 1101 prediabetic and diabetic patients (12.5%) experienced a cardiovascular event. During the 11-year interim follow-up period, such events occurred in 8.2% (45 / 550) of the low-risk group, 15.2% (42 / 276) of the intermediate-risk group, and 18.5% (51 / 275) of the high-risk group. Even after adjusting for factors such as age, sex, use of antihypertensive drugs, use of statins, and smoking history, a significant association was observed between the cardiovascular disease prediction score and the incidence of cardiovascular events. In particular, the risk of cardiovascular events was significantly increased in the intermediate-risk group (hazard ratio 1.57, 95% CI, 1.00-2.47) and the high-risk group (hazard ratio 1.88, 95% CI, 1.19-2.98) compared to the low-risk group. Furthermore, a gradual increase in the hazard ratio (hazard ratio trend 1.36, 95% CI, 1.09-1.70) was observed. These results suggest that the Reti-CVD score can be used as a useful tool for risk stratification among prediabetic and diabetic patients, demonstrating its significant potential in managing high-risk patients.

[0056] 3. Methods for predicting the risk of cardiovascular disease by type.

[0057] 3.1 Direct risk prediction models (classification models) for different types of cardiovascular disease

[0058] The diagnostic device can perform the following actions to directly predict the risk of different types of cardiovascular disease. The machine learning (deep learning) model used in the diagnostic device can learn the input fundus image and the presence or absence of cardiovascular disease types such as atrial fibrillation (AF), coronary artery disease (CAD), peripheral artery disease (PAD), heart failure (HF), and stroke as labels. Through this learning process, features associated with cardiovascular disease are learned and recognized. Based on the analysis results, the diagnostic device's processor can predict the likelihood of developing a particular cardiovascular disease type in the subject of the fundus image. Deep learning algorithms that predict the risk of multiple cardiovascular disease types from fundus images can be designed in various ways. These include single-model, parallel, and multi-task models. The single-model approach is a method of simultaneously predicting multiple cardiovascular disease types (e.g., atrial fibrillation (AF), coronary artery disease (CAD), peripheral artery disease (PAD), heart failure (HF), and stroke) using a single deep learning model. This model has output nodes that take fundus images as input and individually predict the probability of various cardiovascular disease types occurring. The final layer of the model has multiple output nodes, each node returning the probability of a specific cardiovascular disease type occurring. In the parallel model approach, separate deep learning models are built for each cardiovascular disease type. For example, there are models that predict atrial fibrillation (AF), coronary artery disease (CAD), and peripheral artery disease (PAD), each existing independently. Each model takes fundus images as input and predicts only the cardiovascular disease type it targets. The multi-task model approach is similar to the single model but enhances the learning of common features for predicting multiple cardiovascular disease types. One model learns common features and uses them to make individual predictions for each cardiovascular disease type. This model is designed so that after learning features related to multiple cardiovascular disease types in a common layer, it branches out for each cardiovascular disease type. Each branched path makes detailed predictions tailored to that cardiovascular disease type.

[0059] Figure 7 is a diagram illustrating a method for predicting the risk of different types of cardiovascular disease according to one embodiment.

[0060] Referring to Figure 7, the processor of the diagnostic device may include a step of acquiring fundus images (S100) and a step of acquiring cardiovascular disease diagnostic information (S200).

[0061] In step S100, the diagnostic device's processor can acquire fundus images. Furthermore, depending on the embodiment, the diagnostic device's processor can perform preprocessing, expansion, serialization, etc., on the acquired fundus images. Various methods can be applied to these processes, so a detailed explanation is omitted.

[0062] Furthermore, in step S200, the diagnostic device's processor can acquire cardiovascular disease diagnostic information. In this specification, diagnostic information may be expressed as diagnostic support information. Cardiovascular disease diagnostic information may include risk prediction information for different types of cardiovascular diseases. In this specification, risk prediction information for different types of cardiovascular diseases may include risk prediction information for cardiovascular disease types such as stroke, myocardial infarction (MI), atrial fibrillation (AF), heart failure (HF), and peripheral artery disease (PAD). For example, risk prediction information for different types of cardiovascular diseases may include information on the probability value (score) and / or grade of the occurrence of the cardiovascular disease type in the subject of the fundus image. In addition, risk prediction information for different types of cardiovascular diseases may include information on the probability value (score) and / or grade of the occurrence of the cardiovascular disease type in the subject of the fundus image within a predetermined period (e.g., within 10 years, within 5 years). Furthermore, in stage S200, secondary auxiliary information, which will be described later, can be obtained. Specific implementation examples are explained below.

[0063] <Example 1 - Reti-AF Single Model>

[0064] This embodiment demonstrates the conceptual operation of the Reti-AF algorithm for predicting atrial fibrillation (AF) risk. The Reti-AF algorithm uses a patient's fundus image (retinal image) as input and learns the features of the retinal image to recognize patterns associated with the presence or absence of atrial fibrillation (AF). Through the analysis results, the algorithm learns to predict whether the input image is associated with atrial fibrillation (AF) or not. The Reti-AF algorithm can be implemented by the processor of a diagnostic device.

[0065] To evaluate the performance of Reti-AF, a deep learning algorithm related to this embodiment, univariable Cox regression analysis and multivariable Cox regression analysis were performed. Figure 8 is a diagram illustrating the predictive performance of atrial fibrillation risk according to one embodiment. Referring to Figure 8, (a) is a table summarizing the content related to atrial fibrillation risk (Reti-AF) through univariable Cox regression analysis, and (b) is a table summarizing the content related to atrial fibrillation risk (Reti-AF) through multivariable Cox regression analysis.

[0066] Univariate Cox regression analysis showed that a higher Reti-AF score (without adjusting for variables such as age and gender) was associated with a significantly increased incidence of AF. The hazard ratio (HR) for the high-risk group was 5.50 (95% CI, 4.08-7.41), demonstrating the effectiveness of Reti-AF in predicting AF risk. Multivariate Cox regression analysis showed that Reti-AF maintained a significant association with AF incidence even after adjusting for variables such as age, sex, smoking status, diabetes, hyperlipidemia, and hypertension. Age, in particular, acted as an independently significant variable for AF incidence, enhancing the predictive performance of Reti-AF. The C-index of the multivariate model was 0.76, demonstrating improved performance compared to the univariate analysis. In conclusion, the Reti-AF algorithm demonstrated significant performance in predicting the risk of AF incidence.

[0067] 3.2 Cardiovascular disease type-specific risk prediction models based on cardiovascular biomarkers (Reti-CVD)

[0068] 3.2.1 Basic Algorithm Structure

[0069] Patients who already have cardiovascular disease (heart failure, coronary artery disease, stroke, etc.) have an additional risk of developing new specific cardiovascular diseases. Therefore, it is necessary to predict the risk of developing each specific cardiovascular disease.

[0070] Figure 9 illustrates a method for predicting the risk of different cardiovascular diseases according to one embodiment. Figure 9 shows the process of evaluating the overall risk of cardiovascular disease in a patient through a Reti-CVD prediction model (diagnostic model (1000)) (first model (1100)), and predicting the risk of different cardiovascular diseases through an analysis that includes disease type-specific factors (second model (1200)). The first model (1100) and the second model (1200) may be included in the diagnostic model (1000).

[0071] Cardiovascular Risk Prediction Model ("Model 1"): At this stage, the patient's overall cardiovascular risk is assessed through Model 1 (1100). Model 1 (1100) learns the patient's fundus images and coronary calcium (CAC) score and can predict the cardiovascular risk score, Reti-CVD, based on the input fundus images. In addition to the Reti-CVD score, Model 1 (1100) can also predict various disease type-specific factors. For example, disease type-specific factors may include at least one of the parameters representing the subject's physical information, such as biological age, sex, height, weight, BMI index, and information on obesity (obesity index, BMI, body mass, body fat percentage, body muscle mass, presence or absence of obesity, etc.).

[0072] Disease type-specific factors (Laboratory markers) include hematocrit, red blood cell count, white blood cell count, hemoglobin, platelet count, total iron-binding capacity (TIBC), iron content, ferritin (storage iron protein), total protein, albumin, aspartate transaminase (AST), alanine aminotransferase (ALT), γ-GT, alkaline phosphatase (ALP), globulin, hepatitis antigen, hepatitis antibody, blood glucose, glycated hemoglobin (HbA1c), blood urea nitrogen (BUN), creatinine (Cr), uric acid, total cholesterol, high-density lipoprotein (HDL Cholesterol), low-density lipoprotein (LDL Cholesterol), triglycerides (TG), and apolipoprotein B. The diagnostic parameters may include at least one of the following: B, ApoB), serum lipoprotein(a) (Liproprotein(a), Lp(a)), C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), bicarbonate levels, pH levels, hematuria, occult blood levels in urine, protein levels in urine, glucose levels in urine, pH, bilirubin levels in urine, ketone levels in urine, urobilinogen levels in urine, leukocyte count in urine, creatinine levels in urine, albumin levels in urine, and albumin / creatinine ratio in urine. Disease type-specific factors may include at least one of the following: electrocardiogram (ECG), systolic blood pressure (SBP), diastolic blood pressure (DBP), family history information, or patient clinical history information.

[0073] On the other hand, the first model (1100) learns cardiovascular plaque types, i.e., the presence or absence of soft and hard plaques, along with fundus images, and can predict the cardiovascular plaque type according to the input fundus image. The details related to the prediction of cardiovascular plaque types will be explained in detail in section 3.2.4, Algorithm Reflecting Cardiovascular Plaque Types.

[0074] Cardiovascular disease type-specific risk prediction model ("second model"): At this stage, data derived from the cardiovascular risk prediction model is used to predict the risk of each type of cardiovascular disease (e.g., atrial fibrillation, coronary artery disease, stroke, etc.). The risk for each disease type is evaluated by comprehensively combining the Reti-CVD score obtained from the first model (1100) and disease type-specific factors. The data input to the second model (1200) may be the feature vector of the fundus image extracted by the first model (1100) instead of the Reti-CVD score. On the other hand, disease type-specific factors may be predicted by the first model as described above, and in this case, the prediction result (output result) of the disease type-specific factors may be input to the second model (1200).

[0075] The second model (1200) may further include an input module. The input module can obtain disease type-specific factors through user input. As shown in Figure 9, if separate patient data (e.g., sex, age, blood test results, smoking status, obesity status, etc.) is available, this patient data can be input into the cardiovascular disease type-specific risk prediction model. This model individually analyzes the likelihood of each specific cardiovascular disease occurring and predicts which cardiovascular diseases a patient is vulnerable to.

[0076] Output of Cardiovascular Disease Diagnostic Information: In the final stage, cardiovascular disease diagnostic information can be obtained and output by combining the overall cardiovascular risk (integrated score) and the risk for each type. This diagnostic information may include the current patient status, the risk for each type, and related explanatory information, and additionally, it may provide supplementary information (e.g., secondary supplementary information described later) for establishing personalized prevention and treatment plans.

[0077] <Example - Specific Patient Case>

[0078] The cardiovascular disease type-specific risk prediction model independently assesses the type-specific risk for each patient, taking into account the Reti-CVD score and type-specific covariates.

[0079] 1. Patient A's Case [Table 1]

[0080] 2. Patient B Case [Table 2]

[0081] 3.2.2 Detailed Algorithm of the Cardiovascular Disease Type-Specific Risk Prediction Model (Second Model)

[0082] For each type of cardiovascular disease (stroke, myocardial infarction (MI), atrial fibrillation (AF), heart failure (HF), peripheral artery disease (PAD)), the model can consider the Reti-CVD score and specific covariates. Specific covariates include, for example, blood pressure, history of transient ischemic attack (TIA), atrial fibrillation, and smoking status in the case of stroke; LDL cholesterol, family history of heart disease, smoking, and hypertension in the case of myocardial infarction (MI); and age, history of heart disease, obesity, and hypertension in the case of atrial fibrillation (AF). Furthermore, in the case of heart failure (HF), age, history of heart disease, and smoking may be included, and in the case of peripheral artery disease (PAD), smoking, diabetes, hypertension, hyperlipidemia, and age may be included.

[0083] The diagnostic device processor may use a multivariate Cox proportional hazards model or other appropriate statistical or machine learning model for each detailed disease type. The following models may be included in the second model (1200). The following are examples of risk formulas for each disease type performed by each model.

[0084] (1) Stroke model:

[0085] Stroke risk = exp(β1 × Reti-CVD + β2 × Blood Pressure + β3 × TIA History + β4 × Atrial Fibrillation Status + β5 × Smoking + …)

[0086] (2) Myocardial infarction (MI) model:

[0087] Myocardial infarction risk = exp(β1 × Reti-CVD + β2 × LDL cholesterol + β3 × Family history of heart disease + β4 × Smoking + β5 × Hypertension + …)

[0088] (3) Atrial fibrillation (AF) model:

[0089] Atrial fibrillation risk = exp(β1 × Reti-CVD + β2 × Age + β3 × History of Heart Disease + β4 × Obesity + β5 × Hypertension + …)

[0090] (4) Heart failure (HF) model:

[0091] Heart failure risk = exp(β1 × Reti-CVD + β2 × Age + β3 × Smoking + β4 × History of Heart Disease + …)

[0092] (5) Peripheral artery disease (PAD) model:

[0093] Peripheral artery disease risk = exp(β1 × Reti-CVD + β2 × Smoking + β3 × Diabetes + β4 × Hypertension + β5 × Hyperlipidemia + β6 × Age + …)

[0094] In the above formula, β1, β2, etc., represent coefficients for each covariate, and each coefficient reflects the impact of the covariate on the risk. Each model will output risk estimates (e.g., hazard ratios or probabilities) for specific disease types (stroke, myocardial infarction (MI), atrial fibrillation (AF), heart failure (HF), peripheral artery disease (PAD)). These risk estimates are calculated independently of each other, and although the Reti-CVD score is used as a common input for each model, the resulting values ​​can be calculated individually. Furthermore, these risk estimates are not added together because they represent individual probabilities or hazard ratios for different outcomes. Each risk can be interpreted independently based on the combination of the Reti-CVD score and disease type-specific factors. Each model can operate independently. In other words, based on the Reti-CVD score and type-specific covariates, the risk for each cardiovascular type (stroke, myocardial infarction, atrial fibrillation, heart failure, peripheral artery disease) can be calculated, and such models do not influence each other, allowing for independent risk prediction for each type.

[0095] 3.2.3 Algorithms including recommended test decision models

[0096] Figure 10 is a diagram illustrating a method for predicting the risk of different types of cardiovascular diseases according to another embodiment.

[0097] Referring to Figure 10, a recommended test decision model can be added to the basic algorithm configuration described in 3.2.1 above. The content described above can be applied to the first model (cardiovascular risk prediction model, 1100) and the second model (cardiovascular disease type-specific risk prediction model, 1200), so a detailed explanation of the content will be omitted. The third model (1300) can also be included in the diagnostic model (1000).

[0098] Recommended Test Decision Model ("Model 3"): Model 3 (1300) determines what additional clinical tests are necessary for a patient based on the risk of cardiovascular disease type, and the diagnostic device processor can provide information on the additional clinical tests. Here, additional clinical tests may include electrocardiogram (ECG), echocardiogram, ankle-brachial index (ABI), ankle-brachial pressure index (ABPI), blood test, urinalysis, ultrasound, etc.

[0099] For example, if a high risk of stroke is predicted, additional blood pressure monitoring or brain imaging may be recommended. If there is a high risk of myocardial infarction, additional blood tests (e.g., checking LDL levels) and cardiac function tests may be necessary. If there is a high risk of heart failure, it is necessary to evaluate cardiac function in more detail through echocardiography or BNP (B-type natriuretic peptide) testing. If there is a high risk of peripheral artery disease (PAD), it may be recommended to evaluate blood flow status through lower extremity Doppler ultrasound, ankle-brachial index (ABI), or other vascular tests. Also, if there is a high predicted risk of atrial fibrillation (AF), ECG monitoring (e.g., 24-hour Holter monitoring) may be recommended to detect paroxysmal AF. In cases of heart failure, peripheral artery disease, and atrial fibrillation, it may be recommended that patients measure their own heart rate and pain levels through relevant mobile service apps such as heart rate monitoring apps and pain tracking apps.

[0100] <Example of Recommended Test Output for Patients A and B>

[0101] This embodiment demonstrates the process of determining whether additional clinical tests are necessary based on the predicted risk of different cardiovascular disease types for patients A and B. Depending on each patient's condition, appropriate additional tests are recommended for disease types predicted to be at high risk, thereby enabling precise evaluation and management of the patients' cardiovascular health.

[0102] 1. Patient A's Case [Table 3]

[0103] 2. Patient B Case [Table 4]

[0104] When the diagnostic device's processor acquires new data regarding recommended tests, it can re-input the patient's data into the second model (1200), thereby re-evaluating the risk of cardiovascular disease type based on this data, as shown in Figure 10.

[0105] 3.2.4 Algorithm reflecting cardiovascular plaque type

[0106] The algorithm reflecting cardiovascular plaque types can be combined with the basic algorithm configuration described in 3.2.1 above, or with an algorithm that further includes the recommended examination decision model described in 3.2.3. The above-mentioned content can be applied to the first model (cardiovascular risk prediction model, 1100) and the second model (cardiovascular disease type-specific risk prediction model, 1200), so a detailed explanation will be omitted. As briefly mentioned earlier, the first model (1100) can learn the correlation between input fundus images and cardiovascular plaque types. Cardiovascular plaque types can be obtained from CT coronary angiography (CT-CA), cardiovascular magnetic resonance (CMR), 18F-FDG PET (Positron Emission Tomography), 18F-NaF PET (Positron Emission Tomography), coronary angiography, intravascular ultrasound (IVUS), intravascular ultrasound-radiofrequency analysis (IVUS-RF Analysis), optical coherence tomography (OCT), optical frequency domain imaging (OFDI), intravascular thermal imaging (Intracoronary Thermography), Raman spectroscopy, near-infrared spectroscopy (NIRS), and other methods. Specifically, a cardiovascular risk prediction model can be trained using labels indicating soft plaques and corresponding retinal images, and labels indicating hard plaques and corresponding retinal images. This allows the cardiovascular risk prediction model to obtain probability values ​​(and / or grades) for the probability that a subject in a retinal image has soft plaques, or probability values ​​(and / or grades) for the probability that a subject in a retinal image has hard plaques. The first model can predict plaque type as a probability value, along with Reti-CVD and disease type-specific factors, and the probability value can be expressed as a score.

[0107] The plaque type predicted from the first model (1100) can be used to predict the risk of different types of cardiovascular disease. For example, soft plaques may be associated with a higher risk of acute events, particularly stroke, due to the possibility of rupture and embolism. Soft plaques also increase the risk of heart failure and peripheral artery disease, as the possibility of embolus in soft plaques increases the risk of acute limb ischemia. In contrast, hard plaques are associated with progressive occlusion and can lead to myocardial infarction over time. Hard plaques also increase the risk of heart failure and peripheral artery disease, as the progressive narrowing of blood vessels increases the risk of chronic limb ischemia and intermittent claudication.

[0108] As a result, the second model (1200) can consider the plaque type predicted from the first model (1100) as a detailed disease type covariate to predict the risk of various cardiovascular diseases such as stroke, myocardial infarction, heart failure, and peripheral artery disease. Furthermore, the third model (1300) can consider such plaque types to determine what clinical tests are recommended for the patient. For example, if the second model predicts a high risk of stroke and / or the third model predicts the presence of soft or hard plaque, the third model (1300) can generate necessary information such as blood biomarkers (e.g., blood coagulation tests, lipid profiles), blood pressure information, carotid artery status information, and information necessary for diagnosing atrial fibrillation (age, blood pressure information, etc.), and recommend clinical tests to obtain this information. Furthermore, if a high risk of myocardial infarction is confirmed and / or hard plaque is present, the third model (1300) can generate necessary information such as blood biomarkers (e.g., lipid profile) and family history, and recommends clinical tests to obtain this information. Also, based on the diagnostic information for heart failure, if a high risk of heart failure is confirmed and / or soft or hard plaque is present, the third model (1300) can generate necessary information such as blood biomarkers (e.g., BNP levels), a history of previous myocardial infarction, and overall cardiac function indicators, and recommends clinical tests to obtain this information. Also, based on the diagnostic information for peripheral artery disease, if a high risk of peripheral artery disease is confirmed and / or soft or hard plaque is present, the third model can generate necessary information such as smoking information, diabetes information, and blood pressure information, and recommends clinical tests to obtain this information.

[0109] 3.2.5 Detailed subtype risk prediction algorithm for cardiovascular disease types

[0110] Various forms (subtypes) exist for each type of cardiovascular disease. Since each subtype differs in its pathological mechanism, symptoms, prognosis, and treatment methods, it is important to clearly distinguish between them. This invention provides an algorithm that predicts the risk of cardiovascular disease by type, going beyond simply predicting existing integrated cardiovascular risk through machine learning models based on fundus images. The processor of a diagnostic device can predict the risk of detailed subtypes of cardiovascular disease based on the algorithm according to this invention. The basic model configuration can be applied as described in Figure 9, so a detailed explanation is omitted.

[0111] For example, in the case of heart failure (HF), there are subtypes such as heart failure with reduced left ventricular ejection fraction (HFrEF), heart failure with preserved left ventricular ejection fraction (HFpEF), and heart failure with mildly reduced left ventricular ejection fraction (HFmrEF). Heart failure with reduced left ventricular ejection fraction (HFrEF) is a condition in which the left ventricular ejection fraction (LVEF) is reduced to 40% or less, and it occurs when the contractile function of the myocardium is impaired and the heart's blood ejection capacity is greatly reduced. In this case, disease-specific factors include previous myocardial infarction (MI), chronic hypertension, valvular heart disease, diabetes, and smoking, and echocardiography, cardiac MRI, and BNP (B-type natriuretic peptide) testing may be recommended. Heart failure with preserved left ventricular ejection fraction (HFpEF) is a condition in which the left ventricular ejection fraction (LVEF) is preserved to 50% or more, and it is mainly due to problems with ventricular relaxation function, where the heart does not relax sufficiently and its ability to fill with blood is reduced. In this case, obesity, hypertension, diabetes, advanced age, and gender may be considered disease type-specific factors, and echocardiography, cardiac MRI, and BNP testing may be recommended. Furthermore, heart failure with mildly reduced left ventricular ejection fraction (HFmrEF) is heart failure where the left ventricular ejection fraction (LVEF) is between 40% and 49%, and is considered an intermediate form between HFrEF and HFpEF. Advanced age, hypertension, diabetes, obesity, and valvular heart disease may be considered disease type-specific factors, and echocardiography, cardiac MRI, and MBP testing may be recommended.

[0112] Similar to cases of heart failure, stroke can be classified into ischemic stroke, thrombotic stroke (cerebral thrombosis), cerebral embolism (embolic stroke), lacunar infarction, hemorrhagic stroke, intracerebral hemorrhage, subarachnoid hemorrhage, transient ischemic attack, brainstem stroke, and stroke of unknown cause. Atrial fibrillation includes detailed subtypes such as paroxysmal atrial fibrillation, persistent atrial fibrillation, long-term persistent atrial fibrillation, and permanent atrial fibrillation. Myocardial infarction can be classified into detailed subtypes such as type 1 (spontaneous myocardial infarction), type 2 (myocardial infarction due to ischemic disequilibrium), type 3 (myocardial infarction resulting in death when biomarker values ​​are unavailable), type 4a (myocardial infarction associated with percutaneous coronary intervention (PCI)), type 4b (myocardial infarction associated with stent thrombosis), and type 5 (myocardial infarction associated with coronary artery bypass grafting (CABG)).

[0113] As a result, the second model according to the present invention (cardiovascular disease risk prediction model) can predict subtype-specific risks for each disease type by reflecting detailed subtype-specific factors for each type in the cardiovascular risk. Furthermore, through the third model (recommended test judgment model), it is possible to recommend necessary clinical tests for the target patient's cardiovascular disease, taking into account the detailed subtypes of each disease type.

[0114] 3.2.6 Secondary auxiliary information

[0115] Diagnostic information may include the aforementioned cardiovascular risk score (Reti-CVD), plaque-based diagnostic information, disease-specific factors for the target patient, risk and related information for each detailed disease type, and recommended test information. Here, we will describe secondary supplementary information generated additionally based on the above diagnostic information. Specifically, secondary supplementary information includes secondary guide information obtained based on the above diagnostic information. Secondary guide information may include, as an example, prescription information, intervention information, and management information.

[0116] Prescription information can refer to information about medications (e.g., prescription drugs) recommended for a subject to maintain or improve their cardiovascular disease risk in accordance with their diagnostic information. Prescription information may also include information about the prescribed medication, timing of administration, and dosage. For example, prescription information may include information about the prescription of one or more of the following: HMG-CoA reductase inhibitors (statins, including various formulations such as simvastatin, atorvastatin, and rosuvastatin), PCSK9 inhibitors, fibrate derivative combination therapy, aspirin, bile acid sequestrants and nicotinic acid, omega-3 fatty acids, ezetimibe, and fibrates. Furthermore, if the diagnostic information includes the presence of soft and / or hard plaques in the subject, prescription information may include information about high-intensity statin therapy, or calcium channel blockers or ACE inhibitors for hypertension management.

[0117] Furthermore, action information may refer to information about future actions recommended for the subject to maintain or improve their cardiovascular disease risk in accordance with the diagnostic information. For example, additional screening information may include information about secondary diagnoses or medical procedures the subject may undergo. As an example, additional screening information may include information about additional tests required, hospitals / healthcare providers where additional tests can be performed, and recommended procedures / surgeries.

[0118] Furthermore, management information may include information on non-medical measures recommended to the subject to maintain or improve cardiovascular disease risk in accordance with the diagnostic information. For example, management information may include information on lifestyle habits, dietary habits, exercise, and non-specialized medications such as nutritional supplements to lower cardiovascular disease risk. Also, if the diagnostic information includes information that the subject has soft plaque, the management information may include information on strict lifestyle changes (e.g., diet, exercise, smoking cessation, etc.).

[0119] Furthermore, if the diagnostic information includes information indicating the presence of hard plaque in the subject, the management information may include information regarding long-term lifestyle improvements. In one embodiment, the diagnostic device may be linked to an external monitoring device. Here, the monitoring device may mean a device that monitors the subject's lifestyle or behavior. For example, the monitoring device may include a portable device, a wearable device, a wellness measuring device, etc. The monitoring device may also be the client device mentioned above. For example, the monitoring device may include an imaging unit and acquire images of the inside and outside of the eyeball by photographing the inside and outside of the eyeball through the imaging unit.

[0120] Furthermore, the monitoring device can monitor various information such as the subject's activity level, exercise method, exercise duration, food intake, intake amount, health supplement intake information, sleep duration, sleep habits, heart rate, blood pressure, blood glucose levels, body water content, oxygen level, body temperature, oxygen saturation, pulse wave, whether or not they have visited a hospital, whether or not they have undergone an examination, whether or not they have undergone a procedure / surgery, and eye images.

[0121] The diagnostic device's processor can communicate with the monitoring device via a communication module, either wired or wirelessly.

[0122] The processor of the diagnostic device can provide guide information to the monitoring device. The monitoring device can then provide various information to the subject based on the guide information and the monitored information. For example, the monitoring device can acquire management information (e.g., lifestyle information, dietary information, exercise information) as guide information from the diagnostic device, compare the management information with the monitored information to determine whether the monitored information matches the management information, and provide the determination result and / or additional information accordingly.

[0123] For example, if the monitored exercise time is less than the exercise time in the management information, the monitoring device can provide the subject with information to exercise according to the management information. Also, if the monitored information on food intake corresponds to the food intake in the management information, the monitoring device can provide the subject with information that they are consuming food according to the management information.

[0124] Furthermore, the diagnostic device's processor can acquire information monitored from the monitoring device. Based on the guide information and the monitored information, the diagnostic device's processor can provide various information to the subject. For example, the diagnostic device's processor can compare the monitored information with the guide information to determine whether the monitored information matches the management information, and provide the determination result and / or additional information. The aforementioned examples of monitoring devices can be applied to the operation of the diagnostic device's processor.

[0125] Furthermore, the diagnostic device's processor can generate guide information by reflecting the monitoring information received from the monitoring device. For example, the diagnostic device's processor can acquire the subject's condition information (exercise status, lifestyle, eating habits, etc.) based on the monitoring information, and modify the guide information determined as diagnostic information to suit the subject based on the subject's condition information.

[0126] In one embodiment, the processor of the diagnostic device can provide guide information using a predetermined database. For example, the diagnostic device may include a database that matches diagnostic information scores and / or grades with guide information. For example, if the diagnostic information is expressed as three grades, the database may include guide information matched to the low-risk grade (e.g., prescription information - none, action information - information on the next consultation date, management information - provision of dietary information, provision of exercise information), guide information matched to the intermediate-risk grade (e.g., prescription information - none, action information - provision of additional test information, management information - provision of dietary information, provision of exercise information, provision of non-specialized drug information), and guide information matched to the high-risk grade (e.g., prescription information - provision of statin prescription information, action information - additional test information, provision of recommended treatment / surgery information, management information - provision of dietary information, provision of exercise information, provision of non-specialized drug information). Based on the database, the processor of the diagnostic device can provide guide information that matches the diagnostic information.

[0127] In another embodiment, the diagnostic device's processor may provide guide information using a machine learning model. For example, the diagnostic device may include a guide information model based on a machine learning model or a neural network model. The guide information model may be learned based on the score and / or grade of the diagnostic information and the guide information. Additionally, the guide information model may also be learned based on the subject's physical information (at least one of the following: height, weight, age, sex, race, smoking status, blood pressure (e.g., blood pressure value, presence or absence of hypertension), presence or absence of diabetes (or blood glucose value), and cholesterol level). This allows the diagnostic device's processor to input the score and / or grade of the diagnostic information and the subject's physical information into the guide information model to obtain guide information for the subject. Depending on the embodiment, the guide information model may be included in the diagnostic model described above, or it may be configured independently of the diagnostic model.

[0128] Various embodiments of this specification may be implemented by software that includes instruction words stored in a machine-readable storage medium (e.g., a computer). The machine is a device capable of calling instruction words stored in the storage medium and operating according to the called instruction words, and may include electronic devices according to the disclosed embodiments. When the instruction is executed by a processor, the processor may perform the function corresponding to the instruction directly or, under the control of the processor, using other components. The instruction may include code generated or executed by a compiler or interpreter. The machine-readable storage medium may be provided in the form of a non-transitory storage medium. Here, “non-transitory storage medium” means that it does not contain signals and is tangible, without distinguishing whether data is stored semi-permanently or temporarily on the storage medium. For example, “non-transitory storage medium” may include a buffer on which data is temporarily stored.

[0129] According to one embodiment, the methods relating to the various embodiments disclosed herein may be provided in a Computer Program Product. The Computer Program Product may be traded as a commodity between sellers and buyers. The Computer Program Product may be distributed in the form of a device-readable storage medium (e.g., Compact Disc Read Only Memory, CD-ROM) or online through an application store (e.g., Play Store™). In the case of online distribution, at least a portion of the Computer Program Product (e.g., a Downloadable App) may be at least temporarily stored or temporarily generated in a storage medium such as the memory of a manufacturer's server, an application store's server, or an intermediary server.

[0130] Although the embodiments have been described above with reference to the limited embodiments and drawings, various modifications and variations can be made from the above description by a person with ordinary skill in the art. For example, the described technique may be performed in a different order than described, and / or the components of the described system, structure, apparatus, circuit, etc. may be combined or assembled in a different manner than described, or substituted or replaced by other components or equivalents, and the appropriate results may still be achieved.

[0131] Therefore, other realizations, other embodiments, and those equivalent to the claims described below also fall within the scope of the claims.

Claims

1. A cardiovascular disease type-specific risk prediction and diagnostic device, Equipped with at least one processor, The aforementioned at least one processor predicts the risk of different types of cardiovascular disease using a diagnostic model. The aforementioned diagnostic model, The first model analyzes input fundus images to predict cardiovascular risk, A second model, which receives input of the cardiovascular risk score and disease type-specific factors predicted by the first model, predicts the risk of cardiovascular disease by type. Includes, The aforementioned at least one processor is The system is characterized by providing diagnostic information based on the cardiovascular risk predicted by the first model and the risk of different types of cardiovascular diseases predicted by the second model. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

2. The apparatus according to claim 1, The detailed classification of cardiovascular diseases is characterized by including one or more of the following: myocardial infarction (MI), peripheral artery disease (PAD), stroke, atrial fibrillation (AF), and heart failure (HF). A diagnostic device for predicting the risk of different types of cardiovascular diseases.

3. The apparatus according to claim 1, The aforementioned disease type-specific factors are characterized by including one or more of the following: patient's age, sex, blood biomarkers, urine test values, smoking status, presence or absence of obesity, electrocardiogram (ECG), or patient's family history. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

4. The apparatus according to claim 1, The disease type-specific factors are characterized by being predicted by the first model or input from an external source. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

5. The apparatus according to claim 1, The diagnostic information is characterized by including risk grades and related information for each cardiovascular type. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

6. A cardiovascular disease type-specific risk prediction and diagnostic device, Equipped with at least one processor, The at least one processor provides cardiovascular diagnostic information using a diagnostic model. The aforementioned diagnostic model, The first model analyzes input fundus images to predict cardiovascular risk, A second model, which receives input of the cardiovascular risk score and disease type-specific factors predicted by the first model, predicts the risk of cardiovascular disease by type. A third model determines the clinical tests recommended for a patient based on the risk of different types of cardiovascular disease predicted by the second model, Includes, The aforementioned at least one processor is The present invention provides cardiovascular diagnostic information based on the cardiovascular risk predicted by the first model and the risk of different types of cardiovascular diseases predicted by the second model. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

7. The apparatus according to claim 6, The detailed classification of cardiovascular diseases is characterized by including one or more of the following: myocardial infarction (MI), peripheral artery disease (PAD), stroke, atrial fibrillation (AF), and heart failure (HF). A diagnostic device for predicting the risk of different types of cardiovascular diseases.

8. The apparatus according to claim 6, The aforementioned disease type-specific factors are characterized by including one or more of the following: patient's age, sex, blood biomarkers, smoking status, presence or absence of obesity, urinalysis values, electrocardiogram (ECG), or patient's family history. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

9. The apparatus according to claim 6, The disease type-specific factors are characterized by being predicted by the first model or input from an external source. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

10. The apparatus according to claim 6, The additional clinical tests are characterized by including at least one of the following: electrocardiogram (ECG), echocardiogram, ankle-brachial index (ABI), ankle-brachial pressure index (ABPI), blood test, urinalysis, and ultrasound. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

11. The apparatus according to claim 6, The cardiovascular diagnostic information is characterized by including risk grades and related information for each disease type, or information related to recommended tests. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

12. A cardiovascular disease type-specific risk prediction and diagnostic device, Equipped with at least one processor, The at least one processor provides cardiovascular diagnostic information using a diagnostic model. The aforementioned diagnostic model, A first model analyzes input fundus images to predict cardiovascular risk and cardiovascular plaque type, A second model, which receives the cardiovascular risk score and disease type-specific factors predicted by the first model, predicts the risk of cardiovascular disease by type. A third model determines the clinical tests recommended for a patient based on the risk of different types of cardiovascular disease predicted by the second model, Includes, The system is characterized by providing cardiovascular diagnostic information based on the results predicted from the first to third models. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

13. The apparatus according to claim 12, The detailed classification of cardiovascular diseases is characterized by including one or more of the following: myocardial infarction (MI), peripheral artery disease (PAD), stroke, atrial fibrillation (AF), and heart failure (HF). A diagnostic device for predicting the risk of different types of cardiovascular diseases.

14. The apparatus according to claim 12, The aforementioned disease type-specific factors are characterized by including one or more of the following: patient's age, sex, blood biomarkers, smoking status, presence or absence of obesity, urinalysis values, electrocardiogram (ECG), or patient's family history. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

15. The apparatus according to claim 12, The disease type-specific factors are characterized by being predicted by the first model or input from an external source. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

16. The apparatus according to claim 12, The additional clinical tests are characterized by including at least one of the following: electrocardiogram (ECG), echocardiogram, ankle-brachial index (ABI), ankle-brachial pressure index (ABPI), blood test, urinalysis, and ultrasound. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

17. The apparatus according to claim 12, The cardiovascular diagnostic information is characterized by including risk grades and related information for each type of cardiovascular disease, plaque types, or recommended test-related information. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

18. The apparatus according to claim 15, When disease type-specific factors are input to the second model from an external source, the risk of cardiovascular disease type is re-evaluated based on the disease type-specific factors obtained from the external source. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

19. A cardiovascular disease type-specific risk prediction and diagnostic device, Equipped with at least one processor, The aforementioned at least one processor predicts the risk of different types of cardiovascular disease using a diagnostic model. The aforementioned diagnostic model, The first model analyzes input fundus images to predict cardiovascular risk, A second model, which receives input of the cardiovascular risk score and detailed subtype-specific factors for each disease type predicted by the first model, predicts the detailed subtype risk for each disease type. Includes, The aforementioned at least one processor is The diagnostic information is provided based on the cardiovascular risk predicted by the first model and the detailed subtype risk for each type predicted by the second model. A diagnostic device for predicting the risk of different types of cardiovascular diseases.

20. The apparatus according to claim 19, The aforementioned cardiovascular disease category is characterized by including one or more of the following: myocardial infarction (MI), peripheral artery disease (PAD), stroke, atrial fibrillation (AF), and heart failure (HF). A diagnostic device for predicting the risk of different types of cardiovascular diseases.

21. The apparatus according to claim 20, The aforementioned heart failure includes one or more subtypes of heart failure with reduced left ventricular ejection fraction (HFrEF), heart failure with mildly reduced left ventricular ejection fraction (HFmrEF), or heart failure with preserved left ventricular ejection fraction (HFpEF). The aforementioned stroke includes one or more subtypes of ischemic stroke, thrombotic cerebral infarction (cerebral thrombosis), cerebral embolism (embolic cerebral infarction), lacunar infarction, hemorrhagic stroke, intracerebral hemorrhage, subarachnoid hemorrhage, transient ischemic attack, brainstem stroke, and stroke of unknown cause. The atrial fibrillation includes one or more subtypes of paroxysmal atrial fibrillation, persistent atrial fibrillation, long-term persistent atrial fibrillation, or permanent atrial fibrillation. The myocardial infarction is characterized by including one or more subtypes from among type 1 (spontaneous myocardial infarction), type 2 (myocardial infarction due to ischemic disequilibrium), type 3 (myocardial infarction resulting in death when biomarker values ​​are unavailable), type 4a (myocardial infarction associated with percutaneous coronary intervention (PCI)), type 4b (myocardial infarction associated with stent thrombosis), or type 5 (myocardial infarction associated with coronary artery bypass grafting (CABG)). A diagnostic device for predicting the risk of different types of cardiovascular diseases.

Images