Cardiovascular disease layered early warning system

Abstract

The invention relates to the technical field of health status assessment, in particular to a cardiovascular disease layered early warning system, which is characterized in that blood vessel compliance and pulse wave velocity are extracted, arterial dilation capacity is measured, blood vessel elasticity is calculated, blood perfusion distribution is detected, and abnormal vasoconstriction is classified on the basis of a hemodynamic state and a heart electrical activity state; in combination with the thickness of the blood vessel wall and the blood vessel stenosis rate, the plaque deposition condition is analyzed, the arteriosclerosis score is calculated, the thrombosis tendency is detected, the local blood flow blocking degree is quantified, the isolation forest algorithm is adopted, the blood flow dynamic change is calculated, the heart rhythm abnormality is analyzed, the acute vascular occlusion probability is measured, and the sudden heart rhythm abnormality risk is evaluated. And based on a gradient boosting tree algorithm, screening high-risk cases, matching medical intervention measures, setting a risk buffer mechanism, adjusting an individual health management strategy, and optimizing a cardiovascular risk hierarchical early warning system.

Description

Technical Field

[0001] The present invention relates to the technical field of health status assessment, and in particular to a stratified early warning system for cardiovascular diseases. Background Art

[0002] The field of health status assessment technology aims to achieve quantitative assessment, risk prediction and early warning of individual health status through physiological parameter monitoring, data analysis and intelligent algorithms, improve the early detection rate of diseases, optimize health management strategies, reduce the risk of sudden illness, and improve the utilization efficiency of medical resources.

[0003] The cardiovascular disease stratified warning system is designed to assess an individual's cardiovascular health status and provide stratified warnings based on risk levels. The purpose is to identify potential cardiovascular disease risks through real-time and periodic monitoring of physiological indicators such as heart rate, blood pressure, and electrocardiogram signals, combined with data analysis models, and provide personalized health management recommendations or medical intervention plans.

[0004] Existing technologies lack precise calculations of vascular resistance, cardiac output, and blood flow patterns in hemodynamic assessment, resulting in limited early identification capabilities of hemodynamic abnormalities. Cardiac electrical activity monitoring identifies abnormalities based on changes in ECG signals, but does not cover detailed analysis such as rhythm conduction delay calculation and atrial fibrillation pattern matching, resulting in reduced recognition accuracy of some complex arrhythmias. In addition, existing technologies lack systematic analysis of vascular elasticity, arterial dilatation capacity, and blood perfusion distribution, and do not integrate and calculate indicators such as vascular wall thickness and thrombosis risk, affecting the dynamic monitoring capability of early arteriosclerosis, resulting in insufficient identification of high-risk individuals or excessive warning of low-risk individuals, affecting the effectiveness of health management programs. Summary of the Invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose a stratified early warning system for cardiovascular disease.

[0006] In order to achieve the above-mentioned object, the present invention adopts the following technical solution: A stratified early warning system for cardiovascular disease comprises:

[0007] Obtain hemodynamic health status through individual hemodynamic modeling module;

[0008] Obtain the health status of cardiac electrical activity through the cardiac electrical activity monitoring module;

[0009] Based on the individual hemodynamic modeling module and the cardiac electrical activity monitoring module, the vascular function assessment module obtains the vascular health status;

[0010] Obtain arteriosclerosis health status through arteriosclerosis progression detection module;

[0011] Obtain hemodynamic health status through individual hemodynamic modeling module;

[0012] Based on the individual hemodynamic modeling module, the cardiac electrical activity monitoring module, the vascular function assessment module and the arteriosclerosis progression detection module, a cardiovascular event prediction module is used to obtain a cardiovascular event risk assessment result;

[0013] Obtain cardiovascular stratified early warning strategies through the stratified early warning response module;

[0014] Get access to long-term cardiovascular state optimization plans through the Long-term Cardiovascular State Maintenance module.

[0015] As a further embodiment of the present invention, the system includes:

[0016] Individual hemodynamic modeling module: This module extracts blood pressure, heart rate, and blood flow velocity through sensors to analyze blood pressure fluctuations, calculate heart rate stability, decompose pulse waveforms, measure cardiac output, calculate vascular resistance, and analyze blood flow patterns to determine hemodynamic health status.

[0017] Cardiac electrical activity monitoring module: This module extracts ECG signals, RR intervals, and P-QRS-T durations through sensors to perform cardiac rhythm assessment, arrhythmia screening, conduction delay calculation, atrial fibrillation pattern matching, and cardiac arrest precursor detection to determine the health status of cardiac electrical activity.

[0018] Vascular function assessment module: Based on the hemodynamic health status and cardiac electrical activity health status, it extracts vascular compliance and pulse wave velocity, measures arterial dilatation capacity, calculates vascular elasticity, and detects blood perfusion distribution. It also performs vascular contraction abnormality classification and blood flow pattern analysis to obtain vascular health status;

[0019] Arteriosclerosis progression detection module: Based on the vascular health status, it extracts the vascular wall thickness and vascular stenosis rate, performs plaque deposition analysis, calculates the vascular sclerosis score, detects thrombosis tendency, and calculates the degree of local blood flow obstruction to obtain the vascular sclerosis health status;

[0020] Cardiovascular event prediction module: Based on the hemodynamic health status, cardiac electrical activity health status, vascular health status, and arteriosclerosis health status, the module uses the isolation forest to calculate hemodynamic changes, analyze abnormal heart rhythms, and determine the probability of acute vascular occlusion. It also performs a risk assessment of sudden abnormal heart rhythms to obtain cardiovascular event risk assessment results.

[0021] Hierarchical early warning response module: Based on the cardiovascular event risk assessment results, a gradient boosting tree is used to screen high-risk cases, match medical interventions, set risk buffer mechanisms, and adjust individual health management strategies to obtain a cardiovascular stratified early warning strategy;

[0022] Long-term cardiovascular status maintenance module: Based on the cardiovascular stratified early warning strategy, a long-term short-term memory network is used to perform long-term hemodynamic trend analysis, chronic lesion monitoring, and individual health baseline adjustment. In addition, the long-term monitoring plan is optimized to obtain a long-term cardiovascular status optimization plan.

[0023] As a further embodiment of the present invention, the individual hemodynamic modeling module includes:

[0024] Physiological signal acquisition submodule: This module extracts blood pressure, heart rate, and blood flow velocity through sensors, segments the blood pressure curve into segments and calculates peak values, removes high-frequency noise signals, completes missing data, and performs short-term mean analysis to evaluate signal stability and obtain basic blood flow signal data.

[0025] Blood flow characteristic calculation submodule: Based on the blood flow basic signal data, it measures the blood pressure change rate, calculates the heart beat interval, analyzes blood flow velocity abnormalities, and performs vascular compliance measurement and blood flow velocity fluctuation matching to obtain blood flow dynamic characteristic parameters;

[0026] Hemodynamic assessment submodule: Based on the dynamic characteristic parameters of blood flow, calculate the heart's blood output per beat, measure the vascular resistance distribution, classify the blood flow pattern, and perform turbulent interval identification and abnormal flow velocity screening to obtain the hemodynamic health status.

[0027] As a further solution of the present invention, the cardiac electrical activity monitoring module includes:

[0028] ECG signal acquisition submodule: It extracts ECG signals, RR intervals, and P-QRS-T durations through sensors, extracts cardiac waveform peaks, corrects low-frequency drift, removes interference signals, and performs cardiac cycle segmentation and abnormal band elimination to obtain ECG basic signal data.

[0029] Rhythm feature calculation submodule: based on the basic ECG signal data, measures the heartbeat interval, calculates the rhythm stability, classifies the waveform morphology, and performs atrioventricular conduction time measurement and heartbeat rhythm classification to obtain characteristic parameters of cardiac electrical activity;

[0030] Abnormal signal evaluation submodule: Based on the characteristic parameters of cardiac electrical activity, analyze arrhythmias, screen cardiac arrest, calculate electrical axis deviation, and perform abnormal heartbeat signal screening and conduction abnormality detection to obtain the health status of cardiac electrical activity.

[0031] As a further embodiment of the present invention, the vascular function assessment module includes:

[0032] Vascular compliance measurement submodule: Based on the hemodynamic health status and cardiac electrical activity health status, it extracts pulse wave peak value, calculates vascular diameter change, measures pressure response, compares elastic recoil rate, and calculates diastolic and systolic compliance to obtain vascular compliance parameters;

[0033] Blood flow distribution analysis submodule: Based on the vascular compliance parameters, it measures the pulse wave propagation time, calculates the blood flow velocity gradient, evaluates the local resistance, and performs flow attenuation ratio measurement and blood flow redistribution comparison to obtain blood perfusion distribution characteristics;

[0034] Vascular abnormality identification submodule: Based on the blood perfusion distribution characteristics, it classifies vascular contraction abnormalities, evaluates blood flow adaptability, calculates pulse wave phase difference, identifies turbulence and recirculation areas, detects abnormal blood flow morphology, and obtains vascular health status.

[0035] As a further embodiment of the present invention, the arteriosclerosis progression detection module includes:

[0036] Vascular wall thickness measurement submodule: Based on the vascular health status, separates the vascular wall layers, calculates the thickness change rate, compares the long-term trend, and measures the lumen cross-sectional area and calculates the degree of vascular contraction to obtain vascular structural parameters;

[0037] Plaque deposition analysis submodule: Based on the vascular structural parameters, it measures vascular wall proliferation, calculates plaque formation rate, and assesses calcification degree, and also calculates vascular elasticity impact and classifies sclerosis degree to obtain vascular plaque deposition characteristics;

[0038] Blood flow retardation assessment submodule: Based on the characteristics of vascular plaque deposition, the flow velocity reduction rate in the stenosis area is calculated, the blood flow stagnation area is identified, and high coagulation risk point screening and termination area comparison are performed to obtain the health status of arteriosclerosis.

[0039] As a further embodiment of the present invention, the cardiovascular event prediction module includes:

[0040] Hemodynamic change assessment submodule: Based on the hemodynamic health status, cardiac electrical activity health status, vascular health status, and arteriosclerosis health status, the module extracts blood flow acceleration, calculates pulsation intensity changes, and performs local blood flow dynamic deviation analysis. It also monitors cardiac output, measures changes in stroke volume over a short period of time, calculates pulse pressure fluctuation amplitude, and assesses circulating blood perfusion levels to obtain hemodynamic fluctuation characteristics.

[0041] Abnormal heart rhythm detection submodule: Based on the hemodynamic fluctuation characteristics, it calculates the heartbeat interval, extracts RR interval data, detects the difference in beat intervals, and measures instantaneous heart rate changes to classify abnormal heart rhythms. It screens abnormal heartbeat events by comparing ECG waveform patterns, measures the cumulative frequency of abnormal beats, and matches arrhythmia types to obtain abnormal heart rhythm identification results.

[0042] Vascular occlusion risk analysis submodule: Based on the results of abnormal heart rhythm identification, the impact of vascular stenosis is assessed. Isolation forest is used to calculate changes in blood flow resistance, determine areas of insufficient blood supply, and perform local blood flow rate reduction detection to calculate occlusion risk. By evaluating the stability of the vascular wall, determining the trend of plaque rupture, and screening the area of thrombosis, the risk assessment results of cardiovascular events are obtained.

[0043] As a further embodiment of the present invention, the isolation forest is according to the formula:

[0044]

[0045] Where: R f is the change value of blood flow resistance, η is blood viscosity, L is blood vessel length, r is blood vessel radius, ρ is blood density, v is local blood flow velocity, is the dynamic pressure, is the rate of change of vascular wall shear stress, S t To detect the cross-sectional area of blood vessels, S ref is the reference cross-sectional area, α is the dynamic pressure correction value weight, β is the shear stress weight, and γ is the vascular stenosis influence weight.

[0046] As a further solution of the present invention, the hierarchical early warning response module includes:

[0047] High-risk case screening submodule: Based on the cardiovascular event risk assessment results, a gradient boosting tree is used to calculate the risk level. By analyzing the degree of blood flow restriction and the abnormal heart rhythm index, individual risk levels are divided and screened. At the same time, emergency intervention assessment is performed. By measuring the short-term risk fluctuation rate, the proportion of high-risk individuals is calculated to obtain high-risk individual classification results;

[0048] Medical intervention matching submodule: Based on the high-risk individual classification results, intervention needs are determined, and by comparing medical history records, appropriate medical measures are screened, and emergency treatment priority is determined. Vascular lesion characteristics are analyzed to obtain a medical intervention implementation plan.

[0049] Individual health management adjustment submodule: Based on the medical intervention implementation plan, individualized health adjustments are made. By analyzing changes in health status, a lifestyle optimization plan is formulated. At the same time, dynamic health management corrections are made. By tracking changes in intervention effects, intervention details are optimized, and long-term follow-up plans are carried out to obtain cardiovascular stratified early warning strategies.

[0050] As a further solution of the present invention, the gradient boosting tree is according to the formula:

[0051]

[0052] Where: R c is the individual cardiovascular risk score, w j is the feature weight, g(y j ) is the gradient boosting tree prediction function, λ is the regularization coefficient, is the heart rate change rate, V var is the blood flow velocity variability, V base is the baseline blood flow velocity, is the blood flow variability ratio, Q lesion is the blood flow of the diseased vessel segment, Q healthy is the blood flow in the healthy blood vessel segment, is the blood flow difference ratio, θ1 is the heart rate change rate weight, θ2 is the blood flow variability ratio weight, and θ3 is the blood flow difference ratio weight.

[0053] As a further embodiment of the present invention, the long-term cardiovascular status maintenance module includes:

[0054] Hemodynamic trend analysis submodule: Based on the cardiovascular stratification warning strategy, blood pressure fluctuation extraction, cardiac cycle stability detection and pulse wave propagation analysis are performed, and arterial compliance comparison, blood flow pattern classification and local blood flow obstruction calibration are performed to obtain hemodynamic trend characteristics.

[0055] Chronic lesion monitoring submodule: Based on the hemodynamic trend characteristics, a long-short-term memory network is used to calculate the vascular wall thickening rate, evaluate the progression of vascular stenosis, and screen for blood supply restriction. It also performs perfusion abnormality area calibration, arteriosclerosis index analysis, and early identification of vascular occlusion to obtain chronic lesion monitoring results.

[0056] Health baseline adjustment submodule: Based on the chronic disease monitoring results, individual health baseline correction, vascular function matching optimization and long-term monitoring frequency setting are carried out, and dynamic follow-up adjustment, health intervention correction and lifestyle optimization matching are carried out to obtain a long-term cardiovascular status optimization plan.

[0057] Compared with the prior art, the advantages and positive effects of the present invention are:

[0058] 1. The present invention combines hemodynamic status with cardiac electrical activity status to extract vascular compliance and pulse wave velocity, measure arterial dilatation capacity, calculate vascular elasticity, detect blood perfusion distribution, and classify vascular contraction abnormalities to optimize vascular function assessment;

[0059] 2. In this invention, the thickness of the blood vessel wall and the rate of vascular stenosis are combined to analyze the plaque deposition, calculate the arteriosclerosis score, detect the tendency of thrombosis, quantify the degree of local blood flow obstruction, and accurately track the progression of arteriosclerosis;

[0060] 3. In this invention, the isolation forest algorithm is used to calculate hemodynamic changes, analyze arrhythmias, determine the probability of acute vascular occlusion, assess the risk of sudden arrhythmia, and achieve individualized prediction of cardiovascular events;

[0061] 4. In the present invention, the gradient boosting tree algorithm is used to screen high-risk cases, match medical intervention measures, set risk buffer mechanisms, adjust individual health management strategies, optimize the cardiovascular risk stratification warning system, and improve the stability and accuracy of cardiovascular health maintenance. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 is a system flow chart of the present invention;

[0063] Figure 2 is a flow chart of the present invention;

[0064] Figure 3 Schematic diagram of the system framework of the present invention. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0066] The present invention provides a technical solution: a cardiovascular disease stratified early warning system comprising:

[0067] Obtain hemodynamic health status through individual hemodynamic modeling module;

[0068] Obtain the health status of cardiac electrical activity through the cardiac electrical activity monitoring module;

[0069] Based on the individual hemodynamic modeling module and the cardiac electrical activity monitoring module, the vascular function assessment module obtains the vascular health status;

[0070] Obtain arteriosclerosis health status through arteriosclerosis progression detection module;

[0071] Obtain hemodynamic health status through individual hemodynamic modeling module;

[0072] Based on the individual hemodynamic modeling module, cardiac electrical activity monitoring module, vascular function assessment module and arteriosclerosis progression detection module, the cardiovascular event prediction module obtains cardiovascular event risk assessment results;

[0073] Obtain cardiovascular stratified early warning strategies through the stratified early warning response module;

[0074] Get access to long-term cardiovascular state optimization plans through the Long-term Cardiovascular State Maintenance module.

[0075] See also Figure 1 and Figure 2 , a stratified early warning system for cardiovascular disease includes:

[0076] Individual hemodynamic modeling module: This module extracts blood pressure, heart rate, and blood flow velocity through sensors to analyze blood pressure fluctuations, calculate heart rate stability, decompose pulse waveforms, measure cardiac output, calculate vascular resistance, and analyze blood flow patterns to determine hemodynamic health status.

[0077] Cardiac electrical activity monitoring module: This module extracts ECG signals, RR intervals, and P-QRS-T durations through sensors to perform cardiac rhythm assessment, arrhythmia screening, conduction delay calculation, atrial fibrillation pattern matching, and cardiac arrest precursor detection to determine the health status of cardiac electrical activity.

[0078] Vascular function assessment module: Based on the hemodynamic health status and cardiac electrical activity health status, it extracts vascular compliance and pulse wave velocity, measures arterial dilatation capacity, calculates vascular elasticity, and detects blood perfusion distribution. It also classifies vascular contraction abnormalities and analyzes blood flow patterns to determine vascular health status.

[0079] Arteriosclerosis Progression Detection Module: Based on the vascular health status, it extracts the vascular wall thickness and vascular stenosis rate, performs plaque deposition analysis, calculates the vascular sclerosis score, detects thrombosis tendency, and calculates the degree of local blood flow obstruction to obtain the vascular sclerosis health status;

[0080] Cardiovascular event prediction module: Based on the hemodynamic health status, cardiac electrical activity health status, vascular health status, and arteriosclerosis health status, the module uses the isolation forest to calculate hemodynamic changes, analyze arrhythmias, and determine the probability of acute vascular occlusion. It also performs a risk assessment for sudden arrhythmias to obtain cardiovascular event risk assessment results.

[0081] Hierarchical early warning response module: Based on the cardiovascular event risk assessment results, a gradient boosting tree is used to screen high-risk cases, match medical interventions, set risk buffer mechanisms, and adjust individual health management strategies to obtain a cardiovascular stratified early warning strategy;

[0082] Long-term cardiovascular status maintenance module: Based on the cardiovascular stratified early warning strategy, it uses long-short-term memory networks to conduct long-term hemodynamic trend analysis, chronic disease monitoring, and individual health baseline adjustment. It also optimizes long-term monitoring plans to obtain long-term cardiovascular status optimization plans.

[0083] See also Figure 3 , individual hemodynamic modeling modules include:

[0084] Physiological signal acquisition submodule: This module extracts blood pressure, heart rate, and blood flow velocity through sensors, segments the blood pressure curve into segments and calculates peak values, removes high-frequency noise signals, completes missing data, and performs short-term mean analysis to evaluate signal stability and obtain basic blood flow signal data.

[0085] Blood flow characteristic calculation submodule: Based on the basic blood flow signal data, it measures the blood pressure change rate, calculates the heart beat interval, analyzes blood flow velocity abnormalities, and performs vascular compliance measurement and blood flow velocity fluctuation matching to obtain blood flow dynamic characteristic parameters;

[0086] Hemodynamic assessment submodule: Based on the dynamic characteristic parameters of blood flow, it calculates the cardiac output per beat, measures the distribution of vascular resistance, classifies blood flow patterns, and performs turbulent interval identification and abnormal flow velocity screening to obtain the hemodynamic health status;

[0087] Physiological signal acquisition submodule: Based on the blood pressure, heart rate, and blood flow velocity signals collected by the sensor, wavelet transform is used with the decomposition level set to 4. The peak value of the blood pressure curve is calculated in segments. Fast Fourier transform is used to convert the signal to the frequency domain. The cutoff frequency is set to 50Hz, and the components above this frequency are attenuated to remove high-frequency noise signals. Lagrange interpolation method is used with the interpolation order set to 3 to interpolate and fill in missing data points. Sliding window mean filtering is used with the window size set to 5 sampling points to calculate the short-term mean. The signal stability of continuous time periods is analyzed to generate basic blood flow signal data.

[0088] Blood flow feature calculation submodule: Based on the basic blood flow signal data, numerical differentiation is used, the forward difference formula is adopted, and the step size is set to 0.01 seconds. The blood pressure change rate is calculated, the adjacent heartbeat waveform peak time points are matched, and the autoregressive model is used with the order set to 2. The parameters are estimated based on the least squares method, the heartbeat interval is calculated, and Z-score normalization is used with the mean set to 0 and the standard deviation set to 1. The blood flow velocity data is standardized and a threshold is set to screen for abnormal blood flow velocity. Regression curve fitting is used with the least squares method. The fitting function is set to a third-order polynomial to measure vascular compliance. The blood flow velocity fluctuation is pattern matched. The dynamic time warping algorithm is used with the step size constraint set to 1 and the window width set to 10 sampling points. The minimum alignment distance between time series is calculated to generate blood flow dynamic feature parameters.

[0089] Hemodynamic assessment submodule: Based on the dynamic characteristic parameters of blood flow, the heartbeat integration method is used with an integration step of 0.01 seconds to calculate the heart's blood output per beat. Poiseuille's law is used, the blood viscosity is set to 3.5 mPa·s, the vascular radius is averaged, and the vascular resistance distribution is measured. K-means clustering is used with the cluster center set to 3 and the maximum number of iterations set to 100 to classify blood flow patterns. Based on the Reynolds number calculation, the Reynolds number threshold is set to 2300 to identify turbulent intervals. Wavelet packet decomposition is used with the mother wavelet set to Symlets 6 and the number of decomposition levels set to 3. Frequency band energy analysis is performed on the blood flow velocity signal to screen out abnormal flow velocity signals and generate the hemodynamic health status.

[0090] See also Figure 3 , the cardiac electrical activity monitoring module includes:

[0091] ECG signal acquisition submodule: It extracts ECG signals, RR intervals, and P-QRS-T durations through sensors, extracts cardiac waveform peaks, corrects low-frequency drift, removes interference signals, and performs cardiac cycle segmentation and abnormal band elimination to obtain ECG basic signal data.

[0092] Rhythm feature calculation submodule: Based on the basic ECG signal data, it measures the heartbeat interval, calculates the rhythm stability, classifies the waveform morphology, measures the atrioventricular conduction time and classifies the heartbeat rhythm to obtain the characteristic parameters of the cardiac electrical activity;

[0093] Abnormal signal evaluation submodule: Based on the characteristic parameters of cardiac electrical activity, it analyzes arrhythmias, screens for cardiac arrest, calculates electrical axis deviation, and performs abnormal heartbeat signal screening and conduction abnormality detection to obtain the health status of cardiac electrical activity;

[0094] ECG signal acquisition submodule: Based on the ECG signals, RR intervals, and P-QRS-T time courses collected by the sensor, wavelet transform is used with the decomposition layer number set to 3 to decompose the cardiac waveform and extract the peak of the cardiac waveform. The polynomial fitting method is used with the fitting order set to 3 to correct low-frequency drift. Bandpass filtering is used with the cutoff frequency set to 0.5Hz to 50Hz to remove interference signals. Short-time Fourier transform is used with the window length set to 256 sampling points to segment the cardiac cycle. The dynamic time warping algorithm DTW is used with the step size constraint set to 1 and the window width set to 15 sampling points to match the normal waveform template, eliminate abnormal bands, and generate basic ECG signal data.

[0095] Rhythm feature calculation submodule: Based on the basic ECG signal data, the peak detection algorithm is used, the detection threshold is set to 0.6 times the maximum peak, the heart beat interval is measured, the coefficient of variation is calculated, and the window length is set to 5 heart cycles. The rhythm stability is calculated, and the waveform morphology is classified using morphological cluster analysis using the K-means clustering algorithm with the cluster center set to 3. The PR interval calculation method is used based on the sliding window mean filter with the window size set to 5 sampling points. The atrioventricular conduction time is measured. The support vector machine is used with the kernel function set to the radial basis kernel and the training data set size set to 5000 samples. The heart beat rhythm is classified and the characteristic parameters of cardiac electrical activity are generated.

[0096] Abnormal signal assessment submodule: Based on the characteristic parameters of cardiac electrical activity, support vector machine is used with the kernel function set to polynomial kernel and the classification threshold set to 0.8 to analyze arrhythmias. Principal component analysis (PCA) is used to select the first three principal components to screen for cardiac arrest. The axis calculation formula is used to calculate the axis deviation based on the amplitude of the QRS waves in leads I and aVF. The autoregressive model (AR) with the order set to 4 is used to screen abnormal heartbeat signals. The outlier detection algorithm is used with the outlier factor threshold set to 0.7 to screen for conduction abnormalities and generate a healthy state of cardiac electrical activity.

[0097] See also Figure 3 , the vascular function assessment module includes:

[0098] Vascular compliance measurement submodule: Based on the hemodynamic health status and cardiac electrical activity health status, it extracts pulse wave peak value, calculates vascular diameter change, measures pressure response, compares elastic recoil rate, and calculates diastolic and systolic compliance to obtain vascular compliance parameters;

[0099] Blood flow distribution analysis submodule: Based on vascular compliance parameters, it measures pulse wave propagation time, calculates blood flow velocity gradient, evaluates local resistance, and performs flow attenuation ratio measurement and blood flow redistribution comparison to obtain blood perfusion distribution characteristics;

[0100] Vascular abnormality identification submodule: Based on blood perfusion distribution characteristics, it classifies vasoconstriction abnormalities, evaluates blood flow adaptability, calculates pulse wave phase differences, identifies turbulence and recirculation areas, detects abnormal blood flow morphology, and obtains vascular health status;

[0101] Vascular compliance measurement submodule: Based on the hemodynamic health status and cardiac electrical activity health status, the pulse wave signal is decomposed using wavelet transform with the decomposition layer set to 3, the pulse wave peak is extracted, and a boundary tracking algorithm is used with the initial boundary threshold set to 0.2 times the maximum waveform amplitude. Edge detection is performed on the vascular image sequence, and changes in vascular diameter are calculated. Pressure-volume regression analysis is used based on the least squares regression model to calculate the vascular pressure response. Exponential decay fitting is used with the fitting function set to an exponential function to calculate the elastic recoil rate. The diastolic-systolic ratio is calculated based on the second-order derivative of the blood flow pressure waveform to calculate the compliance ratio of the diastolic and systolic periods, and generate vascular compliance parameters.

[0102] Blood flow distribution analysis submodule: Based on vascular compliance parameters, the pulse wave propagation time calculation formula is used to calculate the pulse wave propagation time based on the time difference between the R wave peak of the ECG signal and the rising edge of the pulse wave. The local velocity gradient calculation is used to numerically differentiate the blood flow velocity signal with a step size of 0.01 seconds to calculate the blood flow velocity gradient. Finite element fluid dynamics analysis is used based on the Navier-Stokes equation, setting the blood viscosity to 3.5 mPa·s to evaluate local resistance. The energy attenuation model is used to calculate the kinetic energy loss per unit volume based on the blood kinetic energy equation, and the flow attenuation ratio is measured. The blood flow redistribution matrix is used to calculate the changes in blood flow ratio in different vascular regions based on the fluid continuity equation to generate blood perfusion distribution characteristics.

[0103] Vascular abnormality identification submodule: Based on the blood perfusion distribution characteristics, support vector machine (SVM) is used, the kernel function is set to radial basis kernel, and the number of training samples is set to 5000 to classify vascular contraction abnormalities. The blood flow adaptability evaluation model is used to calculate the shear stress of the vascular wall, and the shear force threshold is set to 1.5 Pa to evaluate blood flow adaptability. The Hilbert transform is used to extract the instantaneous phase of the pulse wave signal and calculate the pulse wave phase difference. The turbulence identification Reynolds number is calculated, and the Reynolds number threshold is set to 2300 to calibrate the turbulent area. The backflow detection algorithm is used to calculate the backflow area based on the velocity vector direction reversal ratio to screen abnormal blood flow morphology and generate vascular health status.

[0104] See also Figure 3 , the arteriosclerosis progression detection module includes:

[0105] Vascular wall thickness measurement submodule: Based on the vascular health status, it separates the vascular wall layers, calculates the thickness change rate, compares the long-term trend, and measures the lumen cross-sectional area and calculates the degree of vascular contraction to obtain vascular structural parameters;

[0106] Plaque deposition analysis submodule: Based on vascular structural parameters, it measures vascular wall proliferation, calculates plaque formation rate, and assesses calcification degree. It also calculates the impact of vascular elasticity and classifies the degree of hardening to obtain vascular plaque deposition characteristics.

[0107] Blood flow stasis assessment submodule: Based on the characteristics of vascular plaque deposition, it calculates the flow velocity drop rate in the stenosis area, identifies the blood flow stagnation area, screens for high coagulation risk points, and compares the termination area to obtain the health status of arteriosclerosis;

[0108] Vascular wall thickness measurement submodule: Based on the vascular health status, an adaptive threshold segmentation algorithm is used, with the initial threshold set at 0.6 times the grayscale mean of the vascular image to separate the vascular wall layers. Morphological gradient operations are used with a 3×3 structuring element to enhance the vascular wall edges and detect thickness changes. A multi-scale edge detection algorithm, based on a Gaussian filter with a standard deviation of 1.2, is used to finely extract the vascular wall boundaries and calculate the rate of change of vascular wall thickness. Time series regression analysis is performed based on a second-order polynomial fitting model with a window length of 10 time points to compare long-term trends. Discrete Fourier transform is used to perform frequency domain analysis on the cross-sectional area series, and the lumen cross-sectional area is calculated after removing high-frequency noise. A radial symmetric transformation is used with a transformation radius of 5 pixels to calculate the degree of vascular contraction and generate vascular structural parameters.

[0109] Plaque deposition analysis submodule: Based on vascular structural parameters, local maximum gradient detection is used with a window size of 5×5 pixels to calculate the area of vascular wall proliferation. A dual-threshold plaque segmentation algorithm is used with a low threshold set to 0.5 times the grayscale mean and a high threshold set to 0.9 times to segment the plaque area. An exponential growth model is used to calculate the plaque formation rate based on time series data. The degree of calcification is assessed using a calcification density measurement method based on Hounsfield unit conversion with a density threshold set to 130 HU. A vascular elasticity regression model is used to calculate changes in vascular elasticity based on vascular wall shear stress with a shear stress threshold set to 1.5 Pa. A K-means clustering algorithm is used with a cluster center set to 3 to classify the degree of sclerosis and generate vascular plaque deposition characteristics.

[0110] Blood flow stasis assessment submodule: Based on the characteristics of vascular plaque deposition, the finite difference method is used with a step size of 0.01 seconds to perform numerical differentiation of the blood flow velocity and calculate the velocity drop rate in the stenotic area. The hemodynamic simulation model is used based on the Navier-Stokes equation, and the blood viscosity is set to 3.5 mPa·s to identify blood flow stagnation areas. The coagulation risk prediction algorithm is used to screen high coagulation risk points based on shear rate calculation. The termination area comparison analysis is used to compare the blood flow pattern in the termination area based on the normalized distribution of blood flow terminal velocity to generate the health status of arteriosclerosis.

[0111] See also Figure 3 , the cardiovascular event prediction module includes:

[0112] Hemodynamic change assessment submodule: Based on the hemodynamic health status, cardiac electrical activity health status, vascular health status, and arteriosclerosis health status, it extracts blood flow acceleration, calculates pulsation intensity changes, and performs local blood flow dynamic deviation analysis. It also monitors cardiac output, measures changes in stroke volume over a short period of time, calculates pulse pressure fluctuation amplitude, and evaluates circulating blood perfusion levels to obtain hemodynamic fluctuation characteristics.

[0113] Abnormal heart rhythm detection submodule: Based on the characteristics of hemodynamic fluctuations, it calculates the heartbeat interval, extracts RR interval data, detects the difference in beat intervals, and measures instantaneous heart rate changes to classify abnormal heart rhythms. It screens abnormal heartbeat events by comparing ECG waveform patterns, measures the cumulative frequency of abnormal beats, and matches arrhythmia types to obtain abnormal heart rhythm identification results.

[0114] Vascular Occlusion Risk Analysis Submodule: Based on the results of abnormal heart rhythm identification, it assesses the impact of vascular stenosis. It uses the isolation forest to calculate changes in blood flow resistance, determine areas of insufficient blood supply, and detect local blood flow rate decreases to calculate occlusion risk. It also assesses vascular wall stability, determines plaque rupture trends, and screens for thrombosis areas to obtain cardiovascular event risk assessment results.

[0115] Hemodynamic change assessment submodule: Based on the hemodynamic health status, cardiac electrical activity health status, vascular health status and arteriosclerosis health status, the second-order derivative calculation is used to perform numerical differentiation of the blood flow velocity signal with a step size of 0.01 seconds to extract the blood flow acceleration. The pulse signal envelope analysis is used to calculate the pulsation signal envelope and measure the change in pulsation intensity. Local weighted regression is used with a smoothing parameter of 0.2 to perform local blood flow dynamic deviation analysis. The heartbeat integration method is used with an integration step size of 0.01 seconds to calculate the cardiac output per unit time. Wavelet decomposition is used to analyze the changes in stroke volume within the cardiac cycle. Fourier transform is used with a frequency domain cutoff of 50Hz to calculate the pulse pressure fluctuation amplitude. The blood perfusion ratio is used to evaluate the circulating blood perfusion level based on the normalized distribution of blood flow per unit volume, and to generate hemodynamic fluctuation characteristics.

[0116] Abnormal heart rhythm detection submodule: Based on the characteristics of hemodynamic fluctuations, the peak detection algorithm is used, and the detection threshold is set to 0.6 times the maximum peak value. The heart beat interval is measured, and the time series difference calculation is used with a step size of 0.01 seconds. The RR interval data is extracted, and the standard deviation calculation is used with a window size of 5 heart cycles to detect the difference in the beat interval. The instantaneous heart rate variation analysis is used based on weighted sliding mean filtering with a window size of 5 sampling points to measure the instantaneous heart rate variation. The support vector machine is used with the kernel function set to the radial basis kernel and the training data set size set to 5000 samples to classify abnormal heart rhythms. The dynamic time warping algorithm is used with a step size constraint of 1 and a window width of 10 sampling points. The ECG waveform patterns are compared to screen abnormal heart beat events. The cumulative frequency calculation is used based on the time series counting method to measure the cumulative frequency of abnormal beats. K-means clustering is used with the cluster center set to 3 to match the arrhythmia types and generate abnormal heart rhythm recognition results.

[0117] Vascular occlusion risk analysis submodule: Based on the results of abnormal heart rhythm identification, hemodynamic modeling is used, and the blood viscosity is set to 3.5 mPa·s based on the Poiseuille equation to evaluate the impact of vascular stenosis. Isolation forest is used, and the outlier factor threshold is set to 0.7 to calculate the change in blood flow resistance. The local blood flow supply ratio is used to determine the area of insufficient blood supply based on the blood flow supply ratio per unit volume per unit time. The finite difference method is used with a step size of 0.01 seconds to detect the decrease in local blood flow rate. The occlusion probability is calculated based on the cumulative resistance model to calculate the occlusion risk. The vascular wall stability index is used to determine the plaque rupture trend based on the vascular wall stress distribution equation. The thrombus recognition algorithm is used to screen the thrombus formation area based on the shear rate change to generate cardiovascular event risk assessment results.

[0118] Isolation forest, according to the formula:

[0119]

[0120] Where: R f is the change value of blood flow resistance, η is blood viscosity, L is blood vessel length, r is blood vessel radius, ρ is blood density, v is local blood flow velocity, is the dynamic pressure, is the rate of change of vascular wall shear stress, S t To detect the cross-sectional area of blood vessels, S ref is the reference cross-sectional area, α is the dynamic pressure correction value weight, β is the shear stress weight, and γ is the vascular stenosis influence weight;

[0121] Execution process: First, the basic blood flow resistance of the vascular segment is calculated by obtaining the patient's vascular imaging data and hemodynamic parameters, and the initial resistance value is calculated by blood viscosity η, vascular length L and radius r, and then the dynamic pressure correction term is introduced. To reflect the effect of local blood flow velocity v on resistance, the blood density ρ is obtained by measuring the individual blood components of the patient, the local blood flow velocity v is calculated by Doppler ultrasound and blood flow simulation data, and the dynamic pressure weight α is obtained by regression analysis of a large-scale cardiovascular event database to optimize the correction ratio of dynamic pressure to blood flow resistance. At the same time, the change rate of vascular wall shear stress is introduced. Reflecting the stress state of the blood vessels, the shear stress σ is calculated by the velocity gradient, and the shear stress correction weight β is determined by the vascular wall stress modeling experiment to accurately evaluate the stability of the blood vessels. Taking into account the effect of the degree of vascular stenosis on the resistance, the local vascular cross-sectional area ratio is calculated, where the detection cross-sectional area S t Determined by angiography or CT, reference cross-sectional area S ref The mean of adjacent non-lesioned vascular segments is taken, and the stenosis influence weight γ is obtained by statistical analysis of large-scale patient data to optimize the influence ratio of local stenosis on resistance and obtain the corrected blood flow resistance change value R f , used to assess the degree of vascular stenosis, the risk of thrombosis and stratified warning of cardiovascular events.

[0122] See also Figure 3 , the layered warning response module includes:

[0123] High-risk case screening submodule: Based on the cardiovascular event risk assessment results, a gradient boosting tree is used to calculate the risk level. By analyzing the degree of blood flow restriction and the abnormal heart rhythm index, individual risk levels are divided and screened. At the same time, emergency intervention assessment is performed. By measuring the short-term risk fluctuation rate, the proportion of high-risk individuals is calculated to obtain the high-risk individual classification results;

[0124] Medical intervention matching submodule: Based on the high-risk individual classification results, intervention needs are determined. By comparing medical history records, appropriate medical measures are screened, and emergency treatment priorities are determined. Vascular lesion characteristics are analyzed to obtain medical intervention implementation plans.

[0125] Individual health management and adjustment submodule: Based on the medical intervention implementation plan, individualized health adjustments are carried out. By analyzing changes in health status, lifestyle optimization plans are formulated. At the same time, dynamic health management adjustments are made. By tracking changes in intervention effects, intervention details are optimized, and long-term follow-up plans are carried out to obtain cardiovascular stratified early warning strategies.

[0126] High-risk case screening submodule: Based on the results of cardiovascular event risk assessment, a gradient boosting tree is used with a learning rate of 0.1, a maximum depth of 5, and 100 iterations. The input risk features are hierarchically calculated to generate a risk level score. Blood flow restriction analysis is used to calculate the local blood flow velocity decrease rate, with a threshold of 20% reduction to determine the degree of blood flow restriction. The heart rhythm abnormality index is calculated based on the RR interval variation coefficient analysis, with an abnormal threshold of 0.15 to classify individual heart rhythm stability. The K-means clustering algorithm is used with the cluster center set to 3 to divide individual risk levels. The short-term risk fluctuation rate is calculated based on a sliding window mean filter with a window size of 10 time points to calculate the short-term risk change rate. The high-risk individual proportion statistics are used to calculate the proportion of high-risk individuals based on the cumulative distribution function to generate high-risk individual classification results.

[0127] Medical intervention matching submodule: Based on the results of high-risk individual classification, the medical history data matching algorithm is used. Based on cosine similarity calculation, individual medical history records are compared to screen similar cases. Medical knowledge graph reasoning is used. Based on the graph embedding model, the embedding dimension is set to 128 to screen suitable medical measures. The priority determination model is used based on logistic regression analysis. The input features include lesion severity, treatment response score, and medical history complexity. The priority of emergency treatment is calculated. Vascular lesion feature analysis is used. Based on CT vascular image segmentation, a U-Net neural network is used with the input channel number set to 1 to extract the lesion area, classify different lesion patterns, and generate a medical intervention execution plan.

[0128] Individual health management adjustment submodule: Based on the medical intervention implementation plan, using health status trend analysis, based on the time series prediction model, setting the number of hidden layer neurons to 64, analyzing the trend of changes in health status, using lifestyle optimization suggestion generation model, based on the rule engine method, input variables include diet structure, exercise frequency and sleep duration, output personalized health adjustment suggestions, using the health management dynamic correction model, based on Bayesian optimization, adjust health management parameters, using the intervention effect tracking algorithm, based on time-weighted scoring, the window size is set to 30 days, analyze the impact of intervention measures on health status, using the long-term follow-up planning model, based on the Markov decision process, set the state transition probability matrix, conduct long-term simulation of individual health status, and generate a cardiovascular stratified early warning strategy.

[0129] Gradient boosting tree, according to the formula:

[0130]

[0131] Where: R c is the individual cardiovascular risk score, w j is the feature weight, g(y j ) is the gradient boosting tree prediction function, λ is the regularization coefficient, is the heart rate change rate, V var is the blood flow velocity variability, V base is the baseline blood flow velocity, is the blood flow variability ratio, Q lesion is the blood flow of the diseased vessel segment, Q healthy is the blood flow in the healthy blood vessel segment, is the blood flow difference ratio, θ1 is the heart rate change rate weight, θ2 is the blood flow variability ratio weight, and θ3 is the blood flow difference ratio weight;

[0132] Execution process: First, collect the patient's vascular imaging data and hemodynamic parameters, and calculate the blood flow resistance change value R f Assess the health of blood vessels, calculate the initial blood flow resistance based on blood viscosity η, blood vessel length L and radius r, and introduce a dynamic pressure correction term Reflects the effect of blood flow velocity v on resistance, further combined with the shear stress change rate Assess the stress state of the vascular wall while considering the effect of local stenosis by using the vascular cross-sectional area ratio Correct the resistance and use the gradient boosting tree method to perform weighted summation of features such as blood flow restriction degree, arrhythmia index, short-term risk fluctuation rate, and add a heart rate variability correction term To measure the stability of autonomic nervous function, combined with the blood flow resistance change rate dR f / dt evaluates short-term vascular status fluctuations and abnormal pulse ratio Reflect the severity of arrhythmia, calculate the individual's comprehensive risk score, and calculate the cardiovascular risk level score Process feature data based on the gradient boosting tree method and additionally introduce the heart rate change rate Measures heart rate fluctuations in a short period of time, combined with blood flow velocity variability ratio Assess local blood flow stability and blood flow difference ratio Quantify the blood supply capacity of diseased and healthy blood vessels, and finally based on R s and R c Determine an individual's cardiovascular disease risk level, achieve accurate screening and early warning for high-risk groups, and provide a scientific basis for personalized cardiovascular health management.

[0133] See also Figure 3 , long-term cardiovascular status maintenance module includes:

[0134] Hemodynamic trend analysis submodule: Based on the cardiovascular stratification warning strategy, it extracts blood pressure fluctuations, detects cardiac cycle stability, and analyzes pulse wave propagation. It also performs arterial compliance comparison, blood flow pattern classification, and local blood flow obstruction calibration to obtain hemodynamic trend characteristics.

[0135] Chronic lesion monitoring submodule: Based on hemodynamic trend characteristics, it uses a long-short-term memory network to calculate the rate of vascular wall thickening, assess the progression of vascular stenosis, and screen for blood supply restriction. It also demarcates abnormal perfusion areas, analyzes the arteriosclerosis index, and identifies early vascular occlusion to obtain chronic lesion monitoring results.

[0136] Health baseline adjustment submodule: Based on the results of chronic disease monitoring, individual health baseline correction, vascular function matching optimization and long-term monitoring frequency setting are carried out. Dynamic follow-up adjustment, health intervention correction and lifestyle optimization matching are also carried out to obtain a long-term cardiovascular status optimization plan;

[0137] Hemodynamic trend analysis submodule: Based on the cardiovascular stratified early warning strategy, wavelet transform is used to decompose the continuous blood pressure signal and extract the blood pressure fluctuation characteristics. The autoregressive moving average model with the order set to 2 is used to predict the blood pressure change trend. The peak detection algorithm is used to detect the peak of the cardiac cycle, calculate the distance between adjacent heartbeats, and perform cardiac cycle stability detection. The pulse wave propagation time is calculated based on the time difference between the ECG R wave peak and the pulse wave rising edge. The arterial compliance comparison algorithm is used, based on the vascular compliance calculation, the comparison threshold is set to 5% of the vascular elasticity change rate, and the blood flow pattern is classified. The local resistance calculation is used. Based on the hemodynamic equation, the shear stress threshold is set to 1.5 Pa, the local blood flow obstruction area is calibrated, and the hemodynamic trend characteristics are generated.

[0138] Chronic lesion monitoring submodule: Based on hemodynamic trend characteristics, long-short-term memory networks and time series difference calculations are used with a step size of 0.01 seconds to calculate the rate of vascular wall thickening. Finite element fluid dynamics analysis is performed based on the Navier-Stokes equation, with blood viscosity set to 3.5 mPa·s to simulate blood flow conditions and assess the progression of vascular stenosis. The blood supply index is calculated based on the blood supply ratio per unit volume to screen for areas with restricted blood supply. CT blood perfusion analysis is performed based on multiphase CT imaging data with a density threshold of 130 HU to demarcate abnormal perfusion areas. The arteriosclerosis index is calculated based on the elastic recoil rate of the vascular wall with an elastic change threshold of 20%. Blood flow occlusion trend analysis is performed based on the cumulative resistance calculation model to screen for early characteristics of vascular occlusion and generate chronic lesion monitoring results.

[0139] Health baseline adjustment submodule: Based on the results of chronic disease monitoring, an individual health baseline regression model is adopted. Based on multiple regression analysis, the input variables include age, vascular compliance and basal blood pressure level. The individual health baseline is adjusted. The vascular function matching optimization algorithm is used. Based on hemodynamic simulation, the initial boundary condition is set as mean arterial pressure to optimize vascular function matching. A long-term monitoring frequency setting model is adopted. Based on the Markov decision process, the state transition probability matrix is set to formulate a long-term monitoring frequency. The health dynamic tracking algorithm is used. Based on time-weighted scoring, the window size is set to 30 days. The health follow-up strategy is dynamically adjusted. An individualized intervention correction model is adopted. Based on Bayesian optimization, health intervention parameters are adjusted. The lifestyle optimization recommendation system is used. Based on the rule engine method, the input variables include diet structure, exercise frequency and sleep duration. The optimal lifestyle plan is matched to generate a long-term cardiovascular status optimization plan.

[0140] The above are merely preferred embodiments of the present invention and do not limit the present invention in any other form. Any technician familiar with the profession may use the technical content disclosed above to change or modify it into an equivalent embodiment with equivalent changes and apply it to other fields. However, any simple modification, equivalent change and modification made to the above embodiment based on the technical essence of the present invention without departing from the content of the technical solution of the present invention shall still fall within the scope of protection of the technical solution of the present invention.

Claims

1. A stratified early warning system for cardiovascular disease, characterized in that: The system comprises: Obtain hemodynamic health status through individual hemodynamic modeling module; Obtain the health status of cardiac electrical activity through the cardiac electrical activity monitoring module; Based on the individual hemodynamic modeling module and the cardiac electrical activity monitoring module, the vascular function assessment module obtains the vascular health status; Obtain arteriosclerosis health status through arteriosclerosis progression detection module; Obtain hemodynamic health status through individual hemodynamic modeling module; Based on the individual hemodynamic modeling module, the cardiac electrical activity monitoring module, the vascular function assessment module and the arteriosclerosis progression detection module, a cardiovascular event prediction module is used to obtain a cardiovascular event risk assessment result; Obtain cardiovascular stratified early warning strategies through the stratified early warning response module; Get access to long-term cardiovascular state optimization plans through the Long-term Cardiovascular State Maintenance module.

2. The cardiovascular disease stratification early warning system according to claim 1, characterized in that: Individual hemodynamic modeling module: extracts blood pressure, heart rate, and blood flow velocity, analyzes blood pressure fluctuations, heart rate stability, and pulse waveforms, measures cardiac output, vascular resistance, and blood flow patterns, and obtains hemodynamic health status; Cardiac electrical activity monitoring module: extracts ECG signals, RR intervals, and P-QRS-T durations, assesses cardiac rhythm, screens for arrhythmias, calculates conduction delays, matches atrial fibrillation patterns, detects precursors to cardiac arrest, and obtains the health status of cardiac electrical activity; Vascular function assessment module: Based on the hemodynamics and cardiac electrical activity health status, it extracts vascular compliance and pulse wave velocity, measures arterial dilatation capacity, vascular elasticity, and blood perfusion, classifies abnormal vasoconstriction, analyzes blood flow patterns, and obtains vascular health status; Arteriosclerosis progression detection module: Based on the vascular health status, extracts vascular wall thickness and stenosis rate, analyzes plaque deposition, calculates arteriosclerosis score, detects thrombosis tendency, calculates local blood flow obstruction, and obtains arteriosclerosis health status; Cardiovascular event prediction module: Based on the hemodynamic health status, cardiac electrical activity health status, vascular health status, and arteriosclerosis health status, the module uses isolation forests to calculate hemodynamic changes, analyze arrhythmias, determine the probability of acute vascular occlusion, assess the risk of sudden arrhythmias, and obtain cardiovascular event risk assessment results; Hierarchical early warning response module: Based on the cardiovascular event risk assessment, a gradient boosting tree is used to screen high-risk cases, match medical interventions, set risk buffers, adjust health management strategies, and obtain a cardiovascular stratified early warning strategy; Long-term cardiovascular status maintenance module: Based on the cardiovascular stratified early warning strategy, a long-term short-term memory network is used to analyze hemodynamic trends, monitor chronic lesions, adjust health baselines, optimize monitoring plans, and obtain long-term cardiovascular status optimization plans.

3. The cardiovascular disease stratification early warning system according to claim 2, characterized in that: The individual hemodynamic modeling module includes: Physiological signal acquisition submodule: This module extracts blood pressure, heart rate, and blood flow velocity through sensors, segments the blood pressure curve into segments and calculates peak values, removes high-frequency noise signals, completes missing data, and performs short-term mean analysis to evaluate signal stability and obtain basic blood flow signal data. Blood flow characteristic calculation submodule: Based on the blood flow basic signal data, it measures the blood pressure change rate, calculates the heart beat interval, analyzes blood flow velocity abnormalities, and performs vascular compliance measurement and blood flow velocity fluctuation matching to obtain blood flow dynamic characteristic parameters; Hemodynamic assessment submodule: Based on the dynamic characteristic parameters of blood flow, calculate the heart's blood output per beat, measure the vascular resistance distribution, classify the blood flow pattern, and perform turbulent interval identification and abnormal flow velocity screening to obtain the hemodynamic health status.

4. The cardiovascular disease stratification early warning system according to claim 2, characterized in that: The cardiac electrical activity monitoring module includes: ECG signal acquisition submodule: It extracts ECG signals, RR intervals, and P-QRS-T durations through sensors, extracts cardiac waveform peaks, corrects low-frequency drift, removes interference signals, and performs cardiac cycle segmentation and abnormal band elimination to obtain ECG basic signal data. Rhythm feature calculation submodule: based on the basic ECG signal data, measures the heartbeat interval, calculates the rhythm stability, classifies the waveform morphology, and performs atrioventricular conduction time measurement and heartbeat rhythm classification to obtain characteristic parameters of cardiac electrical activity; Abnormal signal evaluation submodule: Based on the characteristic parameters of cardiac electrical activity, analyze arrhythmias, screen cardiac arrest, calculate electrical axis deviation, and perform abnormal heartbeat signal screening and conduction abnormality detection to obtain the health status of cardiac electrical activity.

5. The cardiovascular disease stratification early warning system according to claim 2, characterized in that: The vascular function assessment module includes: Vascular compliance measurement submodule: Based on the hemodynamic health status and cardiac electrical activity health status, it extracts pulse wave peak value, calculates vascular diameter change, measures pressure response, compares elastic recoil rate, and calculates diastolic and systolic compliance to obtain vascular compliance parameters; Blood flow distribution analysis submodule: Based on the vascular compliance parameters, it measures the pulse wave propagation time, calculates the blood flow velocity gradient, evaluates the local resistance, and performs flow attenuation ratio measurement and blood flow redistribution comparison to obtain blood perfusion distribution characteristics; Vascular abnormality identification submodule: Based on the blood perfusion distribution characteristics, it classifies vascular contraction abnormalities, evaluates blood flow adaptability, calculates pulse wave phase difference, identifies turbulence and recirculation areas, detects abnormal blood flow morphology, and obtains vascular health status.

6. The cardiovascular disease stratification early warning system according to claim 2, characterized in that: The arteriosclerosis progression detection module includes: Vascular wall thickness measurement submodule: Based on the vascular health status, separates the vascular wall layers, calculates the thickness change rate, compares the long-term trend, and measures the lumen cross-sectional area and calculates the degree of vascular contraction to obtain vascular structural parameters; Plaque deposition analysis submodule: Based on the vascular structural parameters, it measures vascular wall proliferation, calculates plaque formation rate, and assesses calcification degree, and also calculates vascular elasticity impact and classifies sclerosis degree to obtain vascular plaque deposition characteristics; Blood flow retardation assessment submodule: Based on the characteristics of vascular plaque deposition, the flow velocity reduction rate in the stenosis area is calculated, the blood flow stagnation area is identified, and high coagulation risk point screening and termination area comparison are performed to obtain the health status of arteriosclerosis.

7. The cardiovascular disease stratification early warning system according to claim 2, characterized in that: The cardiovascular event prediction module includes: Hemodynamic change assessment submodule: Based on the hemodynamic health status, cardiac electrical activity health status, vascular health status, and arteriosclerosis health status, the module extracts blood flow acceleration, calculates pulsation intensity changes, and performs local blood flow dynamic deviation analysis. It also monitors cardiac output, measures changes in stroke volume over a short period of time, calculates pulse pressure fluctuation amplitude, and assesses circulating blood perfusion levels to obtain hemodynamic fluctuation characteristics. Abnormal heart rhythm detection submodule: Based on the hemodynamic fluctuation characteristics, it calculates the heartbeat interval, extracts RR interval data, detects the difference in beat intervals, and measures instantaneous heart rate changes to classify abnormal heart rhythms. It screens abnormal heartbeat events by comparing ECG waveform patterns, measures the cumulative frequency of abnormal beats, and matches arrhythmia types to obtain abnormal heart rhythm identification results. Vascular occlusion risk analysis submodule: Based on the results of abnormal heart rhythm identification, the impact of vascular stenosis is assessed. Isolation forest is used to calculate changes in blood flow resistance, determine areas of insufficient blood supply, and perform local blood flow rate reduction detection to calculate occlusion risk. By evaluating the stability of the vascular wall, determining the trend of plaque rupture, and screening the area of thrombosis, the risk assessment results of cardiovascular events are obtained.

8. The cardiovascular disease stratification early warning system according to claim 2, characterized in that: The hierarchical early warning response module includes: High-risk case screening submodule: Based on the cardiovascular event risk assessment results, a gradient boosting tree is used to calculate the risk level. By analyzing the degree of blood flow restriction and the abnormal heart rhythm index, individual risk levels are divided and screened. At the same time, emergency intervention assessment is performed. By measuring the short-term risk fluctuation rate, the proportion of high-risk individuals is calculated to obtain high-risk individual classification results; Medical intervention matching submodule: Based on the high-risk individual classification results, intervention needs are determined, and by comparing medical history records, appropriate medical measures are screened, and emergency treatment priority is determined. Vascular lesion characteristics are analyzed to obtain a medical intervention implementation plan. Individual health management adjustment submodule: Based on the medical intervention implementation plan, individualized health adjustments are made. By analyzing changes in health status, a lifestyle optimization plan is formulated. At the same time, dynamic health management corrections are made. By tracking changes in intervention effects, intervention details are optimized, and long-term follow-up plans are carried out to obtain cardiovascular stratified early warning strategies.

9. The cardiovascular disease stratification early warning system according to claim 2, characterized in that: The gradient boosting tree is based on the formula: Where: R c is the individual cardiovascular risk score, w j is the feature weight, g(y j ) is the gradient boosting tree prediction function, λ is the regularization coefficient, is the heart rate change rate, V var is the blood flow velocity variability, V base is the baseline blood flow velocity, is the blood flow variability ratio, Q lesion is the blood flow of the diseased vessel segment, Q healthy is the blood flow in the healthy blood vessel segment, is the blood flow difference ratio, θ1 is the heart rate change rate weight, θ2 is the blood flow variability ratio weight, and θ3 is the blood flow difference ratio weight.

10. The cardiovascular disease stratification early warning system according to claim 2, characterized in that: The long-term cardiovascular status maintenance module includes: Hemodynamic trend analysis submodule: Based on the cardiovascular stratification warning strategy, it performs blood pressure fluctuation extraction, cardiac cycle stability detection, and pulse wave propagation analysis, and also performs arterial compliance comparison, blood flow pattern classification, and local blood flow obstruction calibration to obtain hemodynamic trend characteristics; Chronic lesion monitoring submodule: Based on the hemodynamic trend characteristics, a long-short-term memory network is used to calculate the rate of vascular wall thickening, assess the progression of vascular stenosis, and screen for blood supply restriction. It also performs perfusion abnormality area calibration, arteriosclerosis index analysis, and early identification of vascular occlusion to obtain chronic lesion monitoring results. Health baseline adjustment submodule: Based on the chronic disease monitoring results, individual health baseline correction, vascular function matching optimization and long-term monitoring frequency setting are carried out, and dynamic follow-up adjustment, health intervention correction and lifestyle optimization matching are carried out to obtain a long-term cardiovascular status optimization plan.

Images