A method for monitoring cardiovascular data in cardiology

Abstract

This invention discloses a method for monitoring cardiovascular data in cardiology, comprising the following steps: S1, Non-invasive collaborative acquisition of multimodal cardiovascular data: Constructing a multimodal non-invasive sensor array, fitting sensitive points on the wrist, chest, and behind the ear, to simultaneously acquire photoplethysmography (PPG) pulse wave signals, multi-lead electrocardiogram (ECG) signals, and bioimpedance signals, completing preliminary data encapsulation and synchronous transmission. This invention employs multimodal non-invasive sensor collaborative acquisition technology to simultaneously acquire multi-dimensional cardiovascular data such as ECG, pulse wave, and bioimpedance, breaking the limitations of single-dimensional monitoring and comprehensively covering core monitoring indicators such as cardiac electrophysiology, peripheral vascular function, and cardiac function, providing complete data support for disease assessment and effectively avoiding missed diagnoses of early cardiovascular abnormalities.

Description

Technical Field

[0001] This invention belongs to the field of cardiovascular medical monitoring technology, and in particular relates to a method for monitoring cardiovascular data in cardiovascular medicine. Background Technology

[0002] Cardiovascular disease is one of the leading causes of disease worldwide, and its incidence continues to rise due to population aging and lifestyle changes. Therefore, early warning, dynamic monitoring, and risk assessment of cardiovascular disease have become important research directions in the field of clinical cardiology.

[0003] Currently, the commonly used cardiovascular data monitoring methods in clinical and civilian fields are mainly divided into two categories: invasive monitoring and non-invasive monitoring. Among them, non-invasive monitoring has become the mainstream monitoring method due to its high safety and good compliance. However, existing technologies still have many shortcomings and cannot meet the dual needs of refined clinical monitoring and long-term home monitoring.

[0004] Existing non-invasive cardiovascular monitoring technologies mostly use a single sensor modality to acquire data, which can only obtain basic indicators such as heart rate, blood pressure, and blood oxygen. They cannot comprehensively capture multi-dimensional cardiovascular parameters such as electrocardiogram waveforms, pulse wave characteristics, vascular elasticity, and cardiac output. The single data dimension leads to one-sided disease assessment and easily misses early abnormal signals such as paroxysmal arrhythmias and occult myocardial ischemia. Although some multi-parameter monitoring devices integrate multiple sensor modules, they lack an appropriate signal collaborative processing mechanism and are easily affected by factors such as human movement, respiratory fluctuations, environmental electromagnetic interference, and poor skin contact. The signal noise is large and the data distortion is serious, and the monitoring accuracy cannot meet the clinical diagnosis and treatment standards.

[0005] Meanwhile, existing monitoring methods mostly use generalized data analysis models that do not incorporate personalized physiological information such as patient age, gender, underlying medical history, and body mass index. This results in poor adaptability to different populations and a high risk of misjudgment or missed diagnosis. Furthermore, they can only achieve real-time data display and simple threshold alarms, lacking hierarchical risk assessment and dynamic early warning mechanisms. They cannot deeply mine the trend changes and abnormal correlation characteristics of cardiovascular data, making it difficult to predict cardiovascular risks in advance and hindering timely clinical intervention and patient self-health management.

[0006] In addition, existing monitoring technologies have shortcomings in data transmission and storage. Bedside monitoring devices have complex wiring and poor mobility, while home monitoring devices have high data transmission latency and weak compatibility, making it impossible to achieve multi-terminal data synchronization and sharing. This is not conducive to medical staff remotely and in real time grasping patients' conditions, and it is also difficult to build complete cardiovascular health records, thus restricting the implementation of full-cycle management of cardiovascular diseases. Summary of the Invention

[0007] The purpose of this invention is to provide a method for monitoring cardiovascular data in cardiovascular medicine, so as to solve the technical problems mentioned in the background art.

[0008] To achieve the above objectives, the specific technical solution of the present invention is as follows: A method for monitoring cardiovascular data in cardiovascular medicine, comprising the following steps: S1. Non-invasive Collaborative Acquisition of Multimodal Cardiovascular Data: A multimodal non-invasive sensor array is constructed, fitting three cardiovascular signal-sensitive points on the wrist, chest, and behind the ear to simultaneously acquire multi-dimensional raw cardiovascular data, avoiding the limitations of single-point acquisition. Specifically, a photoelectric sensor module is deployed at the wrist to acquire photoplethysmography (PPG) signals, extracting peripheral vascular data such as pulse waveform, pulse rate, blood oxygen saturation, and pulse wave velocity. A flexible electrocardiogram (ECG) sensor module is deployed at the chest to acquire multi-lead ECG signals, extracting cardiac electrophysiological data such as heart rate, ECG waveform morphology, ST segment, and QRS complex. An impedance sensor module is deployed behind the ear to acquire bioimpedance signals, extracting data related to vascular elasticity, blood flow velocity, and cardiac output. A fixed sampling frequency is set during acquisition to ensure strict synchronization of the time axis of each modality's data, synchronously completing the initial data encapsulation and transmitting it to the data processing terminal. S2. Layered Intelligent Noise Reduction and Precise Data Purification: After receiving the raw multimodal data, the data processing terminal initiates layered intelligent noise reduction processing, removing interference signals layer by layer and retaining effective physiological data to solve data distortion problems caused by motion artifacts, respiratory interference, electromagnetic noise, etc. The first layer is environmental noise filtering, which uses an adaptive bandpass filtering algorithm to remove high-frequency electromagnetic interference and low-frequency baseline drift noise, while retaining the effective frequency bands corresponding to cardiovascular physiological signals. The second layer is physiological artifact removal, which constructs a motion interference recognition model and combines it with human posture data collected by the acceleration sensor module to identify and remove motion artifact signals such as walking, raising arms, and turning over, while separating respiratory fluctuation interference to purify pure cardiovascular physiological signals. The third layer is abnormal data calibration, which uses an interpolation fitting algorithm to calibrate and complete abnormal data points with missing signals or abrupt amplitude changes, while completing data normalization processing to unify the dimensions of data from each modality and form a standardized multimodal cardiovascular dataset. S3. Personalized Cardiovascular Feature Modeling and Deep Analysis: A personalized cardiovascular feature analysis model is constructed, combining patient basic physiological information with a standardized multimodal dataset to achieve in-depth data mining and accurate assessment. First, personalized parameters of the monitored subjects are entered, including age, gender, height, weight, basic medical history, and previous cardiovascular examination results, completing the initial parameter configuration of the model and generating a dedicated feature analysis benchmark. Second, core feature parameters are extracted from the standardized dataset, including abnormal ECG waveform features, pulse wave velocity, vascular elasticity coefficient, heart rate variability, and blood oxygen fluctuation threshold, constructing a multi-dimensional feature vector. Finally, the feature vector is analyzed using deep learning algorithms, compared with the dedicated analysis benchmark, and parameter anomalies, trend changes, and multi-parameter correlation features are identified. This distinguishes between physiological fluctuations and pathological abnormalities, improving the accuracy of disease assessment and avoiding misjudgment errors common in general models. S4. Tiered Dynamic Cardiovascular Risk Early Warning: Based on personalized analysis results, a three-tiered dynamic risk early warning mechanism is established to achieve tiered early warning and precise alerts for cardiovascular risks, balancing the timeliness of early warnings with the rationality of alarms. The first-level warning is for low-risk situations, where a single parameter exhibits a slight abnormality without clinical or pathological significance. An abnormality is only displayed on the monitoring terminal without requiring an active alarm, reminding the monitored individual to adjust their status and retest. The second-level warning is for medium-risk situations, where multiple parameters are borderline abnormal and there is a potential pathological risk. Local audio-visual alerts are activated, and a warning message is sent to the linked mobile device, prompting the monitored individual to seek medical attention promptly. The third-level warning is for high-risk situations, such as malignant abnormalities in ECG waveforms, sudden increases or decreases in blood pressure, or severe arrhythmias. An emergency alarm is immediately triggered, and high-risk warning information and the monitored individual's location are simultaneously pushed to medical staff terminals and family members' mobile devices, facilitating rapid clinical intervention. S5. End-to-End Data Interoperability and Health Record Construction: Through a wireless communication module, end-to-end data interoperability is achieved. Standardized cardiovascular datasets, personalized analysis results, and early warning records are simultaneously uploaded to the cloud server and pushed to bedside monitoring terminals, medical workstations, patient mobile devices, and family terminals, enabling real-time data sharing across multiple terminals. The cloud server categorizes and stores the data, organizes it chronologically, and constructs a personalized cardiovascular health record for each monitored individual. This record documents full-cycle monitoring data, abnormalities, early warning information, and treatment intervention records, supporting data retrospective querying, trend analysis, and report export. It provides comprehensive data support for clinical diagnosis and treatment, disease follow-up, and health management. Simultaneously, encryption algorithms ensure data transmission and storage security and protect patient privacy.

[0009] Preferably, the sampling frequency in step S1 is set to 200-500Hz, which balances data acquisition accuracy and device power consumption, and is suitable for long-term continuous monitoring needs. The multimodal sensor acquisition array adopts a flexible wearable design, and a breathable and non-slip gel pad is set at the contact point to improve wearing comfort and signal acquisition stability.

[0010] Preferably, in step S2, the adaptive bandpass filter frequency band is 0.5-40Hz, which accurately matches the frequency bands of cardiovascular physiological signals such as ECG, pulse wave, and bioimpedance, maximizing the elimination of invalid noise. The motion interference recognition model adopts a lightweight neural network with a recognition accuracy of ≥95% and a signal distortion rate of ≤2% after calibration and completion.

[0011] Preferably, the deep learning algorithm in step S3 adopts a fusion model of convolutional neural network and long short-term memory network, and is trained and optimized through massive clinical cardiovascular data. The accuracy of pathological abnormality identification is ≥98%, and it can accurately identify common cardiovascular abnormalities such as atrial fibrillation, premature beats, myocardial ischemia, and hypertensive crisis.

[0012] Preferably, in step S5, Bluetooth, Wi-Fi, and 5G dual-mode communication are used. Wi-Fi transmission is used in bedside monitoring scenarios, and Bluetooth + 5G transmission is used in home monitoring scenarios. The data transmission delay is ≤1s, and the cloud data is stored using national cryptographic encryption algorithms, so the risk of privacy leakage is zero.

[0013] The cardiovascular data monitoring method of the present invention has the following advantages: 1. This invention employs multimodal non-invasive sensing collaborative acquisition technology to simultaneously acquire multi-dimensional cardiovascular data such as electrocardiogram, pulse wave, and bioimpedance. It breaks through the limitations of a single monitoring dimension, comprehensively covers core monitoring indicators such as cardiac electrophysiology, peripheral blood vessels, and cardiac function, provides complete data support for disease assessment, and effectively avoids the missed diagnosis of early cardiovascular abnormalities.

[0014] 2. This invention uses layered intelligent noise reduction processing to remove environmental noise, motion artifacts, and physiological interference layer by layer, which greatly improves the purity and accuracy of cardiovascular data. It solves the pain points of traditional monitoring technology being susceptible to interference and data distortion. Even under slight human movement, it can still ensure the reliability of monitoring data and is suitable for continuous monitoring in daily activities.

[0015] 3. This invention constructs a personalized cardiovascular feature analysis model, combines the basic physiological information of the monitored subjects to customize the analysis benchmark, abandons the generalized analysis mode, greatly improves the monitoring adaptability of people of different ages, physical conditions and medical histories, accurately distinguishes between physiological fluctuations and pathological abnormalities, reduces the misjudgment rate, and meets the refined needs of clinical diagnosis and treatment in cardiovascular medicine.

[0016] 4. This invention sets up a three-level dynamic risk warning mechanism to achieve stratified and accurate early warning of cardiovascular risks. This avoids frequent invalid alarms that affect the user experience, and enables rapid response to high-risk abnormalities, thus achieving early prediction and timely intervention of cardiovascular risks and providing protection for patients' life safety.

[0017] 5. This invention enables end-to-end data interoperability and health record construction, and multi-terminal data synchronization and sharing, facilitating remote monitoring by medical staff, self-management by patients, and real-time awareness by family members. It constructs a full-cycle monitoring and management system for cardiovascular diseases, combining clinical practicality with home convenience, and has a wide range of applicable scenarios and extremely high promotional value. Detailed Implementation

[0018] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the embodiments of the invention. Therefore, the description is considered to be exemplary in nature and not restrictive.

[0019] In the description of the embodiments of the present invention, it should be understood that the terms "length", "vertical", "horizontal", "top", "bottom", etc., indicating the orientation or positional relationship are only for the convenience of describing the embodiments of the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation on the embodiments of the present invention.

[0020] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of embodiments of the present invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0021] In this embodiment of the invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection, an electrical connection, or a communication connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this embodiment of the invention according to the specific circumstances.

[0022] The following disclosure provides many different implementations or examples for carrying out different structures of the embodiments of the present invention. To simplify the disclosure of the embodiments of the present invention, specific examples of components and arrangements are described below. Of course, these are merely examples and are not intended to limit the embodiments of the present invention. Furthermore, reference numerals and / or reference letters may be repeated in different examples of the embodiments of the present invention; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various implementations and / or arrangements discussed.

[0023] To better understand the purpose, structure, and function of this invention, a cardiovascular data monitoring method for cardiovascular medicine is described in further detail below.

[0024] Example 1 Embodiment 1 of the present invention provides a method for monitoring cardiovascular data in the cardiovascular department, applicable to continuous bedside monitoring of hospitalized patients in the cardiology department, comprising the following steps: S1. Multimodal data acquisition: The flexible wearable sensor array is attached to the patient's wrist, chest, and behind the ear respectively. The acquisition function is turned on, the sampling frequency is set to 300Hz, and the photoplethysmography pulse wave signal, multi-lead electrocardiogram signal, and bioimpedance signal are acquired simultaneously and transmitted to the bedside monitoring terminal in real time. S2. Layered noise reduction and purification: After receiving the raw data, the terminal removes environmental noise through 0.5-40Hz adaptive bandpass filtering, removes motion artifacts such as turning over and raising hands by combining acceleration data, calibrates abnormal data points and completes normalization processing to obtain a standardized cardiovascular dataset. S3. Personalized Analysis: Input parameters such as patient age, gender, coronary heart disease history, and baseline blood pressure, configure model benchmarks, extract core features such as heart rate, ST segment, and pulse wave velocity, and identify potential features of myocardial ischemia by integrating deep learning models. S4. Risk Warning: If monitoring detects slight ST segment depression in a patient, it is determined to be of medium risk. A Level II warning is activated, with local audio-visual alerts and information pushed to the medical workstation. Medical staff can view the data in a timely manner and adjust the treatment plan accordingly. S5. Data Management: Monitoring data is synchronously uploaded to the cloud to build a patient-specific health record, recording monitoring data, early warning information and treatment intervention records, which facilitates disease follow-up and efficacy evaluation.

[0025] Example 2 A cardiovascular data monitoring method for cardiovascular medicine according to Embodiment 2 of the present invention is suitable for long-term home monitoring of elderly patients with hypertension, and includes the following steps: S1. Multimodal data acquisition: Patients wear a portable sensor array to automatically collect multi-dimensional cardiovascular data during daily home activities. The sampling frequency is 250Hz, and the system operates with low power consumption to achieve continuous monitoring for 24 hours. S2. Layered noise reduction and purification: Eliminate motion interference such as daily walking and sitting posture changes, as well as environmental electromagnetic noise, purify pure data, calibrate data anomalies, and ensure monitoring accuracy. S3. Personalized Analysis: Combining parameters such as patient age, history of hypertension, and body mass index, analyze the trends of blood pressure fluctuations, heart rate variability, and vascular elasticity changes to distinguish between physiological fluctuations and pathological abnormalities. S4. Risk Warning: If a patient's blood pressure rises suddenly at night and their heart rate becomes irregular, triggering a Level 3 high-risk warning, the warning information will be immediately pushed to the family's mobile phone and the community medical care terminal, and the family will accompany the patient to seek medical treatment in a timely manner. S5. Data Management: Cloud storage of full-cycle monitoring data, generating weekly / monthly monitoring reports. When patients return for follow-up visits, the reports can be exported for doctors' reference to optimize antihypertensive treatment plans.

[0026] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A cardiovascular data monitoring method for cardiovascular medicine, characterized by: Comprise the following steps: S1, multi-modal cardiovascular data non-invasive cooperative acquisition: build a multi-modal non-invasive sensing acquisition array, adhere to the sensitive points of human wrist, chest and behind ear, synchronously collect photoplethysmographic wave signal, multi-lead electrocardiogram signal and bioimpedance signal, complete preliminary encapsulation and synchronous transmission of data; S2, hierarchical intelligent noise reduction and data accurate purification: three-layer noise reduction processing is performed on the original multi-modal data, environmental noise, motion artifacts and physiological interference are sequentially removed, abnormal data is calibrated and normalized processing is completed, forming a standardized multi-modal cardiovascular data set; S3, personalized cardiovascular feature modeling and deep analysis: enter the individual physiological parameters of the monitoring object, configure the exclusive analysis benchmark, extract the core feature vector of the data, and analyze it through the deep learning algorithm to distinguish physiological fluctuations from pathological abnormalities; S4, hierarchical dynamic cardiovascular risk early warning: based on the analysis results, a three-level early warning mechanism is established, corresponding warning modes are triggered for different risk levels, and low-risk prompts, medium-risk reminders and high-risk emergency alarms are realized; S5, full-link data intercommunication and health record construction: through wireless communication, multi-terminal data sharing is realized, cloud storage of monitoring data is realized, exclusive cardiovascular health records are constructed, and data backtracking and trend analysis are supported.

2. The cardiovascular data monitoring method of claim 1, wherein: In step S1, the sampling frequency is set to 200-500Hz, the multi-modal sensing acquisition array adopts a flexible wearable design, and a breathable anti-slip gel pad is provided at the adhesion site.

3. The cardiovascular data monitoring method of claim 1, wherein: In step S2, the adaptive band-pass filter frequency band is 0.5-40Hz, the motion interference recognition model adopts a lightweight neural network, and the signal distortion rate is ≤2%.

4. The cardiovascular data monitoring method of claim 1, wherein: In step S3, the deep learning algorithm adopts a convolutional neural network and long short-term memory network fusion model, and the pathological abnormality recognition accuracy is ≥98%.

5. The cardiovascular data monitoring method of claim 1, wherein: In step S5, Bluetooth, Wi-Fi and 5G dual-mode communication are adopted, the data transmission delay is ≤1s, and the cloud data is stored using a national encryption algorithm.