Comprehensive health management data processing system for coronary heart disease patients

Abstract

The invention belongs to the technical field of medical health, and particularly relates to a comprehensive health management data processing system for coronary heart disease patients. By introducing a multi-disease combined management model layer and a graded early warning and response module, management of multiple complications of coronary heart disease patients can be efficiently monitored and diagnosed, and the causal relationship, concurrent mode and treatment association between coronary heart disease and various diseases such as hypertension and diabetes mellitus can be clearly displayed; doctors can comprehensively understand complex relations among various diseases from a macroscopic angle, conflict of treatment schemes of all departments is avoided, limitation of single disease treatment is avoided, and the comprehensive treatment effect is improved; in addition, real-time assessment risks are output based on the model through a grading early warning and response module and are graded, the health risk of the patient can be found in time, and the occurrence risk of disease deterioration and complications is reduced. And medical resources can be reasonably allocated according to the risk level of the patient.

Description

Technical Field

[0001] The present invention belongs to the field of medical health technology, and in particular relates to a comprehensive health management data processing system for patients with coronary heart disease. Background Art

[0002] Coronary heart disease (CHD), a serious cardiovascular disease that threatens human health, has become a common medical phenomenon as more and more patients suffer from multiple conditions due to changes in modern living environments and lifestyles. For example, an elderly patient may suffer from coronary heart disease, hypertension, and diabetes, among other chronic conditions. Because these conditions fall under different medical specialties, patients often have to travel between multiple departments, including cardiology, endocrinology, and cardiovascular medicine.

[0003] However, there are serious obstacles to information communication between departments. On the one hand, there are differences in the diagnosis and treatment standards and drug use adopted by different departments. When diagnosing and treating the same patient, there is a lack of effective communication mechanisms to share information, resulting in difficulties in coordinating the treatment plans formulated by various departments. For example, the drugs prescribed by the cardiology department for the treatment of coronary heart disease may conflict with the treatment plan formulated by the endocrinology department for the control of diabetes in terms of drug interactions. Not only will the expected treatment effect fail to be achieved, but it may also bring additional health risks to the patient. On the other hand, patients need to repeat some examination items, which not only wastes a lot of time and money, but also increases the physical burden and psychological pressure on patients, further increasing the burden of medical treatment for patients.

[0004] Furthermore, most current early warning systems are designed independently for a single disease, focusing solely on changes in specific disease indicators while ignoring the patient's overall health. This single-disease independent early warning model hinders accurate and efficient diagnosis and treatment during medical intervention. Lacking comprehensive information, doctors are unable to promptly identify potential health risks and struggle to quickly develop targeted and effective treatment strategies when faced with complex situations involving multiple concurrent illnesses. Summary of the Invention

[0005] In response to the problems existing in the existing technology, the present invention proposes a comprehensive health management data processing system for patients with coronary heart disease. The purpose is to efficiently monitor and diagnose multiple complications of patients with coronary heart disease by introducing a multi-disease joint management model layer and a hierarchical early warning and response module, thereby solving the problems in the current medical system where disease management is limited to a single disease, information communication between departments is poor, and the existing early warning system is unable to conduct comprehensive assessments, resulting in heavy medical burdens and increased medical risks for patients.

[0006] The technical solution of the comprehensive health management data processing system for coronary heart disease patients of the present invention is as follows, comprising: The data collection module includes a basic health data collection unit, a simple cardiopulmonary function data collection unit, a daily behavior data linkage collection unit, and a comorbidity-related data collection unit, which are used to obtain the patient's basic health data, cardiopulmonary function data, daily behavior data, and comorbidity-related data; The data processing module is used to preprocess multi-source data, extract and fuse features, optimize and select features, and generate core feature data; Multi-disease joint management model layer: Based on disease-related knowledge graphs and multi-task learning models, it is used to achieve comprehensive risk assessment and intervention for coronary heart disease and complications; Grading warning and response module: triggers graded warnings through real-time data stream analysis, and coordinates resources with community hospitals and higher-level hospitals based on risk levels; User interaction module: including patient side and doctor side, the doctor side realizes data sharing and remote collaborative management with the superior hospital; The above modules form a closed-loop management chain of "data collection → data processing → multi-disease assessment → graded warning and response → user interaction" in sequence.

[0007] Preferably, the basic health data collection unit is connected to the medical information platform through the community hospital electronic health record system to automatically obtain the patient's age, gender, height, weight, smoking history, family medical history, electrocardiogram and echocardiogram data. The basic health data collection unit is also provided with a patient information entry interface.

[0008] Preferably, the simple cardiopulmonary function data acquisition unit adopts a portable sports bracelet and a portable pulmonary function meter. The sports bracelet integrates a heart rate sensor and an acceleration sensor. The pulmonary function meter measures vital capacity and forced expiratory volume in one second. The device has a built-in Bluetooth module to transmit data in real time.

[0009] Preferably, the daily behavior data linkage collection unit collects the patient's daily activity steps, exercise intensity and sleep data through smart wearable devices or mobile health applications, and develops a data interface to connect with the data fusion and intelligent processing module.

[0010] Preferably, the comorbidity-related data collection unit collects blood glucose data, 24-hour blood pressure fluctuation data, blood lipid data and medication compliance data through a portable blood glucose meter, a dynamic blood pressure monitor, a rapid blood lipid detection device and a smart medicine box.

[0011] Preferably, the data processing module includes: Data preprocessing uses adaptive filtering algorithms to remove noise, spatiotemporal interpolation algorithms to fill missing data, and standardizes and normalizes multi-source data; Feature extraction and fusion: Using the attention mechanism to extract dynamic features of heart rate variability and exercise energy expenditure from time series data, extract ventilation function features from lung function data, extract cardiac structure features from image data, extract blood sugar fluctuation trends and blood pressure circadian rhythms from comorbidity data, and construct feature vectors through multimodal fusion algorithms; Feature optimization selection combines genetic algorithms and gradient boosting decision trees to screen core feature data. The genetic algorithm performs global feature space search, and the gradient boosting decision tree performs local optimization based on feature contribution to remove redundant feature data.

[0012] Preferably, the disease-related knowledge graph is constructed by collecting medical knowledge and clinical data of a large number of patients with coronary heart disease and their complications, mining the causal relationship, concurrent patterns and treatment association information between diseases, and constructing an intuitive knowledge graph. The knowledge graph is regularly updated and optimized based on new clinical data, clinical guidelines, historical patient data, and the latest literature scanned through NLP technology, and after manual review.

[0013] Preferably, the multi-task learning model is guided by a disease-related knowledge graph, and the model architecture includes a shared feature layer and task branches. The shared feature layer uses a Transformer encoder to extract common features, and the task branches include prediction of cardiopulmonary fitness level of coronary heart disease, evaluation of the probability of achieving hypertension control targets, and identification of the risk of diabetes-related cardiovascular complications. At the same time, it learns multiple tasks such as diagnosis of coronary heart disease and its complications, disease prediction, and treatment recommendation, captures the correlation between different disease tasks, and provides decision support for personalized treatment.

[0014] Preferably, the graded warning and response module uses the Apache Flink engine to analyze heart rate, blood pressure, and blood oxygen data streams in real time, and classifies warning levels into low risk, medium risk, high risk, and emergency risk; a. If the indicators are normal, the risk is low and the indicator light is green. Only regular follow-up and health management recommendations are required. b. Detection of occasional ventricular premature beats or transient blood pressure elevation indicates medium risk, with the indicator light showing yellow, triggering a community review, adjustment of treatment plans, and increased follow-up frequency; c. Sustained ST-segment depression or blood oxygen <90% indicates high risk, and the indicator light turns red. The community hospital and higher-level hospital will initiate a collaborative mechanism, providing remote guidance for treatment or arranging referrals. d. When various indicators are closer to the preset risk values, it is an emergency risk and the indicator light will flash red. The emergency process will be immediately initiated and medical resources will be deployed for emergency treatment.

[0015] Preferably, the user interaction module includes: Patients can view assessment results, exercise prescriptions, and health recommendations through mobile apps, receive smart exercise reminders, and communicate with medical staff through online consultation windows. Doctors can use the visual assessment management platform to view detailed patient information, conduct group management and develop follow-up plans, and use data analysis to understand the overall status of community patients. When encountering complex cases, they can consult remotely with experts from higher-level hospitals through the platform. The doctor side has established a data sharing mechanism. On the one hand, it synchronizes patient data to the superior hospital to realize the full-process tracking and management of patient health data; on the other hand, it realizes data sharing and collaborative management between community hospitals and superior hospitals, and promotes the integration and utilization of medical resources.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention not only collects cardiopulmonary fitness data (basic health data, dynamic physiological data, and daily behavior data), but also comprehensively assesses the patient's health status through a comorbidity-related data collection unit and a multi-disease joint management model. It can clearly demonstrate the causal relationship, concurrent patterns, and treatment associations between coronary heart disease and multiple diseases such as hypertension and diabetes. Doctors can fully understand the complex relationships between multiple diseases from a macro perspective, avoid conflicts in treatment plans among different departments, avoid the limitations of single disease treatment, and improve the overall treatment effect.

[0017] 2. The multi-disease joint management model of the present invention integrates clinical data, provides comprehensive support for medical decision-making, and greatly reduces unnecessary repeated examinations; it not only saves patients' time and money, reduces physical and psychological burdens, but also optimizes the allocation of medical resources; at the same time, through the application and updating of the model, it also promotes information sharing and collaboration between different departments, which helps to improve the overall level of medical services.

[0018] 3. The hierarchical warning and response module of the present invention assesses and grades risks in real time based on model output, enabling timely identification of patients' health risks and reducing the risk of disease progression and complications. Furthermore, medical resources are rationally allocated based on the patient's risk level. Low-risk patients receive routine management in community hospitals, while medium- and high-risk patients receive coordinated attention from community hospitals and higher-level hospitals. Patients at critical risk can quickly receive emergency resource support, avoiding waste and over-concentration of medical resources and improving their utilization efficiency. Medical resources are rationally allocated based on the patient's risk level.

[0019] 4. Through the user interaction module, the present invention enables information sharing and collaborative cooperation among patients, community hospital medical staff, and experts from higher-level hospitals. Patients can obtain more comprehensive and continuous health services; community hospital medical staff can improve their diagnosis and treatment capabilities with the professional support of higher-level hospitals; and higher-level hospitals can also obtain patients' daily health data through community hospitals to achieve remote management and treatment plan optimization. This invention strengthens the integration and utilization of medical resources at all levels, narrows the urban-rural medical gap, and greatly improves the efficiency of doctor-patient collaboration and full-process management. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a flow chart of a comprehensive health management data processing system for coronary heart disease patients according to the present invention. DETAILED DESCRIPTION

[0021] In order to better understand the content of the present invention, the present invention will be further described below with reference to specific examples. The following examples are based on the technology of the present invention and provide detailed implementation methods and operating steps, but the scope of protection of the present invention is not limited to the following examples. That is, based on the examples of the present invention, all other examples obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0022] The present invention provides a comprehensive health management data processing system for patients with coronary heart disease, which is composed of a data acquisition module, a data processing module, a multi-disease joint management model layer, a hierarchical warning and response module, and a user interaction module. The data acquisition module includes a basic health data acquisition unit, a simple cardiopulmonary function data acquisition unit, a daily behavior data linkage acquisition unit, and a comorbidity-related data acquisition unit. The basic health data collection unit develops a dedicated electronic health record system for community hospitals and connects it with the medical information platform to automatically obtain the patient's basic health data such as age, gender, height, weight, smoking history, family medical history, previous electrocardiogram, echocardiogram, etc. at the superior hospital; at the same time, a simple patient information entry interface is designed to facilitate community medical staff to quickly supplement and collect information such as the patient's recent symptoms, medication status, etc., to ensure the comprehensiveness and timeliness of the data.

[0023] The simplified cardiopulmonary function data collection unit introduces portable devices, such as a portable sports bracelet (with integrated heart rate and acceleration sensors) and a portable spirometer (which can measure vital capacity, forced expiratory volume in one second, and other indicators). Patients can complete cardiopulmonary function data collection in community hospitals through a short period of simple exercise (such as marching in place, walking up and down stairs) or breathing tests. The device has a built-in Bluetooth module that can transmit the collected data in real time to the data fusion and intelligent processing module. The daily behavior data linkage collection unit uses smart wearable devices (smart watches) or mobile health applications to record daily activity steps, exercise intensity, sleep duration and other data; by developing a data interface, it realizes data fusion and docking of the intelligent processing module with smart wearable devices and mobile health applications, automatically synchronizes the patient's daily behavior data, and supplements the cardiopulmonary fitness assessment information from the dimension of daily life.

[0024] The comorbidity-related data collection unit utilizes multi-parameter detection equipment deployed in community hospitals. It is a portable terminal that integrates blood glucose, blood pressure, and blood lipid detection. After the patient completes the cardiopulmonary fitness assessment, comorbidity screening is performed simultaneously; it mainly collects blood glucose data, blood pressure data, blood lipid data, and medication compliance data; for blood glucose data, fasting blood glucose, postprandial blood glucose, and glycosylated hemoglobin indicators are obtained through a portable blood glucose meter or continuous blood glucose monitoring device; for blood pressure data, a dynamic blood pressure monitor is introduced to record 24-hour blood pressure fluctuations and circadian rhythm characteristics; for blood lipid data, low-density lipoprotein cholesterol, triglycerides and other indicators are obtained through rapid detection equipment in community hospitals; for medication compliance data, the patient's medication time, dosage, and missed doses are recorded through a smart medicine box. The smart medicine box uploads medication data via Bluetooth and automatically associates with the electronic health record system.

[0025] As an embodiment of the present invention, patients with limited mobility can rent portable devices to complete data collection at home. The devices automatically link to electronic health records to ensure data continuity. As the portable sports bracelet and portable pulmonary function meter are easy to use, patients can rent them according to the guidance of medical staff to facilitate cardiopulmonary data collection for special groups with difficulty walking.

[0026] Basic health data provides static characteristics, portable device data provides dynamic physiological response characteristics, and daily behavior data reflects long-term trends. The three form a multi-dimensional data complementarity in time and space; more importantly, the present invention also has a comorbidity-related data collection unit, which not only collects coronary heart disease-related data, but also collects functional data of other common diseases of patients (such as hypertension, diabetes, etc.), which can comprehensively reflect the patient's health status.

[0027] The data acquisition module transmits the collected multi-source data to the data processing module, which performs data preprocessing, feature extraction and fusion, and feature optimization and selection in sequence.

[0028] First, we preprocess the collected multi-source data, using an adaptive filtering algorithm to remove noise and a spatiotemporal interpolation algorithm to fill in missing data. We then standardize and normalize the data based on the data type, unifying the data format and eliminating differences in data dimensions, providing a high-quality data foundation for subsequent analysis.

[0029] Secondly, feature extraction and fusion use the attention mechanism in deep learning to extract features from basic health data, simple cardiopulmonary function data, daily behavior data and comorbidity-related data; extract dynamic features such as heart rate variability and exercise energy consumption from time series data; extract ventilation function-related features from pulmonary function data; extract cardiac structure features from image data (such as echocardiography), and extract blood sugar fluctuation trends and blood pressure circadian rhythms from comorbidity data; then, a multimodal feature fusion algorithm is used to fuse the features of different types of data to construct a feature vector containing the patient's comprehensive health information.

[0030] Finally, feature optimization selection combines genetic algorithms and gradient boosting decision trees to screen the fused features; the genetic algorithm performs a global search in the feature space to find potential excellent feature combinations; the gradient boosting decision tree performs local optimization based on the contribution of the features to the evaluation results. The two work together to remove redundant and irrelevant features and retain the core feature data that is most influential for cardiopulmonary fitness assessment, significantly reducing the computing resource consumption of community hospitals, providing accurate data support for subsequent evaluation models, and enhancing the depth and effectiveness of data processing. The core feature data processed by the data processing module becomes the input of the multi-disease joint management model layer; the multi-disease joint management model layer includes the construction of disease-related knowledge graphs and multi-task learning models; The disease-related knowledge graph construction collects medical knowledge and clinical data from a large number of patients with coronary heart disease and its comorbidities (such as hypertension and diabetes). It explores the causal relationships, concurrent patterns, and treatment associations between diseases, constructing an intuitive and interpretable disease-related knowledge graph to clarify how hypertension affects the risk of coronary heart disease and the interaction between diabetes and coronary heart disease in drug treatment. This enables doctors to fully understand the complex relationships between multiple diseases from a macro perspective, providing a strong basis for formulating comprehensive treatment plans.

[0031] The multi-task learning model is guided by a disease-related knowledge graph and simultaneously learns multiple tasks, including the diagnosis, disease prediction, and treatment recommendations for coronary heart disease and its complications. The model architecture includes a shared feature layer and task branches. The shared feature layer uses a Transformer encoder to extract common features (such as inflammatory factor levels and autonomic nervous system function status), which are then divided into three branches: prediction of cardiopulmonary fitness level for coronary heart disease, assessment of the probability of achieving hypertension control targets, and identification of the risk of diabetes-related cardiovascular complications. The multi-task learning model can effectively capture the correlation between different disease tasks and improve the learning effect of each task. When predicting the progression of coronary heart disease, the impact of complications such as diabetes and hypertension is considered, thereby more accurately assessing the patient's health status and providing precise decision support for personalized treatment.

[0032] As an embodiment of the present invention, the disease-related knowledge graph is regularly updated and optimized based on new clinical data. The update is completed after manual review by integrating clinical guidelines and historical patient data and scanning the latest literature through NLP technology.

[0033] The hierarchical warning and response module will conduct real-time risk assessment and hierarchical response strategies for the patient's health status based on the output results of the multi-disease joint management model layer, combined with preset risk thresholds and indicator weights.

[0034] (1) Real-time risk assessment: Using the Apache Flink engine, the system analyzes heart rate, blood pressure, and blood oxygen data streams in real time. Based on the streaming data, the system categorizes warning levels into low risk, medium risk, high risk, and emergency risk. When a patient's coronary heart disease worsens and comorbidity indicators fluctuate abnormally, the system automatically issues a warning signal of the corresponding level, alerting medical staff to the patient's condition.

[0035] (2) Gradual response strategy: formulate corresponding graded response strategies for different warning levels; a. If the indicators are normal, indicating low risk, the indicator light will be green. Only regular follow-up is required. The user interaction module will provide health management suggestions to patients, reminding them to maintain a healthy lifestyle and take medication on time. b. Detection of occasional ventricular premature beats or transient blood pressure elevation indicates medium risk, with the indicator light showing yellow. This triggers a community review, prompting community medical staff to proactively contact the patient, adjust the treatment plan, and increase the frequency of follow-up visits. c. Sustained ST-segment depression or blood oxygen <90% indicates high risk, and the indicator light turns red. The collaborative mechanism between the community hospital and the superior hospital is activated, and experts from the superior hospital provide remote guidance on treatment and arrange for patient referral if necessary. d. When various indicators are closer to the preset risk values, it is an emergency risk and the indicator light will flash red. The emergency process will be immediately initiated and medical resources will be deployed for emergency treatment.

[0036] The user interaction module includes both a patient and doctor side. The patient side is used for feedback and interaction. Through a dedicated mobile application, the application presents cardiopulmonary fitness assessment results, exercise prescriptions, and health recommendations to patients in the form of intuitive charts and text. The mobile application also features intelligent reminders, providing regular reminders for patients to exercise, take medication, and receive follow-up examinations. It also provides an online consultation window, allowing patients to consult community medical staff at any time regarding issues encountered during exercise. Medical staff will provide timely answers and guidance, enhancing patients' self-management skills and treatment compliance.

[0037] The doctor's side is used for decision support, providing community hospital medical staff with a visual assessment and management platform that integrates detailed patient assessment reports, exercise prescription implementation status, health indicator change trends, and other information. Medical staff can use the platform to group and manage patients and develop personalized follow-up plans for different patient groups. Utilizing the platform's data analysis capabilities, they can quickly understand the overall cardiopulmonary fitness and disease management effectiveness of community coronary heart disease patients, providing data support for optimizing community coronary heart disease prevention and control strategies. At the same time, the platform supports remote consultations with experts from higher-level hospitals. When encountering complex cases, community medical staff can upload patient information to obtain professional guidance and advice from higher-level experts.

[0038] As an implementation method of the present invention, the doctor side also establishes a data sharing mechanism between the community hospital and the superior hospital, synchronizing the patient's assessment data, exercise prescription execution status and other information in the community hospital to the superior hospital's medical information system, thereby realizing the full-process tracking and management of the patient's health data. Experts in the superior hospital can remotely adjust and optimize the patient's treatment plan based on the data uploaded by the community hospital; medical staff in the community hospital can also obtain the latest diagnosis and treatment guidelines and research results of the superior hospital in a timely manner, improve their own professional level and service capabilities, and form a health management model for coronary heart disease patients in which the community hospital and the superior hospital cooperate and promote the integration and utilization of medical resources.

[0039] The present invention realizes a comprehensive assessment of the patient's health status by comprehensively collecting data, constructing a multi-disease joint management model, a hierarchical early warning response, and a user interaction module. In terms of data collection and evaluation, not only cardiopulmonary fitness-related data are collected, but also with the help of comorbidity collection units and multi-disease joint management models, the patient's health is comprehensively assessed, the relationship between multiple diseases is clearly presented, conflicts in treatment plans are avoided, and the comprehensive treatment effect is improved. At the same time, the multi-disease joint management model layer integrates data to support medical decision-making, reduce repeated examinations, optimize the allocation of medical resources, and promote departmental collaboration. The hierarchical early warning and response module can assess risks in real time, allocate resources reasonably, and reduce disease risks. The user interaction module strengthens information sharing and collaboration between patients, community hospital medical staff, and experts from higher-level hospitals, improves the efficiency of doctor-patient collaboration and full-process management, and narrows the urban and rural medical gap.

[0040] The present invention can have other forms of embodiments according to the above method, which are not listed one by one. Therefore, any simple modification, equivalent change and modification made by any person skilled in the art to the above embodiment according to the technical essence of the present invention without departing from the scope of the technical solution of the present invention shall still fall within the scope of the technical solution of the present invention.

Claims

1. A comprehensive health management data processing system for patients with coronary heart disease, characterized in that: include: The data collection module includes a basic health data collection unit, a simple cardiopulmonary function data collection unit, a daily behavior data linkage collection unit, and a comorbidity-related data collection unit, which are used to obtain the patient's basic health data, cardiopulmonary function data, daily behavior data, and comorbidity-related data; The data processing module is used to preprocess multi-source data, extract and fuse features, optimize and select features, and generate core feature data; Multi-disease joint management model layer: Based on disease-related knowledge graphs and multi-task learning models, it is used to achieve comprehensive risk assessment and intervention for coronary heart disease and complications; Grading warning and response module: triggers graded warnings through real-time data stream analysis, and coordinates resources with community hospitals and higher-level hospitals based on risk levels; User interaction module: including patient side and doctor side, the doctor side realizes data sharing and remote collaborative management with the superior hospital; The above modules form a closed-loop management chain of "data collection → data processing → multi-disease assessment → graded warning and response → user interaction" in sequence.

2. A comprehensive health management data processing system for patients with coronary heart disease according to claim 1, characterized in that: The basic health data collection unit is connected to the medical information platform through the community hospital electronic health record system to automatically obtain the patient's age, gender, height, weight, smoking history, family medical history, electrocardiogram and echocardiogram data. The basic health data collection unit is also provided with a patient information entry interface.

3. A comprehensive health management data processing system for patients with coronary heart disease according to claim 1, characterized in that: The simple cardiopulmonary function data acquisition unit adopts a portable sports bracelet and a portable pulmonary function meter. The sports bracelet integrates a heart rate sensor and an acceleration sensor. The pulmonary function meter measures vital capacity and forced expiratory volume in one second. The device has a built-in Bluetooth module to transmit data in real time.

4. A comprehensive health management data processing system for patients with coronary heart disease as claimed in claim 1, characterized in that: The daily behavior data linkage collection unit collects the patient's daily activity steps, exercise intensity and sleep data through smart wearable devices or mobile health applications, and develops a data interface to connect with the data fusion and intelligent processing module.

5. A comprehensive health management data processing system for patients with coronary heart disease as claimed in claim 1, characterized in that: The comorbidity-related data collection unit collects blood glucose data, 24-hour blood pressure fluctuation data, blood lipid data and medication compliance data through a portable blood glucose meter, a dynamic blood pressure monitor, a rapid blood lipid detection device and a smart medicine box.

6. A comprehensive health management data processing system for patients with coronary heart disease as claimed in claim 1, characterized in that: The data processing module includes: Data preprocessing uses adaptive filtering algorithms to remove noise, spatiotemporal interpolation algorithms to fill missing data, and standardizes and normalizes multi-source data; Feature extraction and fusion: Using the attention mechanism to extract dynamic features of heart rate variability and exercise energy expenditure from time series data, extract ventilation function features from lung function data, extract cardiac structure features from image data, extract blood sugar fluctuation trends and blood pressure circadian rhythms from comorbidity data, and construct feature vectors through multimodal fusion algorithms; Feature optimization selection combines genetic algorithms and gradient boosting decision trees to screen core feature data. The genetic algorithm performs global feature space search, and the gradient boosting decision tree performs local optimization based on feature contribution to remove redundant feature data.

7. A comprehensive health management data processing system for patients with coronary heart disease as claimed in claim 1, characterized in that: The disease-related knowledge graph is constructed by collecting medical knowledge and clinical data of a large number of patients with coronary heart disease and their complications, mining the causal relationship, concurrent patterns and treatment-related information between diseases, and constructing an intuitive knowledge graph. The knowledge graph is regularly updated and optimized based on new clinical data, clinical guidelines, historical patient data, and the latest literature scanned through NLP technology, and after manual review.

8. A comprehensive health management data processing system for patients with coronary heart disease as claimed in claim 7, characterized in that: The multi-task learning model is guided by a disease-related knowledge graph. The model architecture includes a shared feature layer and task branches. The shared feature layer uses a Transformer encoder to extract common features. The task branches include predicting the cardiopulmonary fitness level of coronary heart disease, assessing the probability of achieving hypertension control targets, and identifying the risk of diabetes-related cardiovascular complications. At the same time, it learns multiple tasks such as the diagnosis, disease prediction, and treatment recommendation of coronary heart disease and its complications, capturing the correlation between different disease tasks and providing decision support for personalized treatment.

9. A comprehensive health management data processing system for patients with coronary heart disease as claimed in claim 1, characterized in that: The graded warning and response module uses the Apache Flink engine to analyze heart rate, blood pressure, and blood oxygen data streams in real time, and classifies warning levels into low risk, medium risk, high risk, and emergency risk. a. If the indicators are normal, the risk is low and the indicator light is green. Only regular follow-up and health management recommendations are required. b. Detection of occasional ventricular premature beats or transient blood pressure elevation indicates medium risk, with the indicator light showing yellow, triggering a community review, adjustment of treatment plans, and increased follow-up frequency; c. Sustained ST-segment depression or blood oxygen <90% indicates high risk, and the indicator light turns red. The community hospital and higher-level hospital will initiate a collaborative mechanism, providing remote guidance for treatment or arranging referrals. d. When various indicators are closer to the preset risk values, it is an emergency risk and the indicator light will flash red. The emergency process will be immediately initiated and medical resources will be deployed for emergency treatment.

10. A comprehensive health management data processing system for patients with coronary heart disease according to claim 9, characterized in that: The user interaction module includes: Patients can view assessment results, exercise prescriptions, and health recommendations through mobile apps, receive smart exercise reminders, and communicate with medical staff through online consultation windows. Doctors can use the visual assessment management platform to view detailed patient information, conduct group management and develop follow-up plans, and use data analysis to understand the overall status of community patients. When encountering complex cases, they can consult remotely with experts from higher-level hospitals through the platform. The doctor side has established a data sharing mechanism. On the one hand, it synchronizes patient data to the superior hospital to realize the full-process tracking and management of patient health data; on the other hand, it realizes data sharing and collaborative management between community hospitals and superior hospitals, and promotes the integration and utilization of medical resources.

Images