Cardiovascular medicine disease early warning analysis system based on big data

Abstract

The invention discloses a cardiovascular medicine disease early warning analysis system based on big data, and relates to the technical field of disease early warning, and the system comprises a data processing and modeling module which is used for collecting physiological signal data and home behavior data of a patient, and constructing a physiological baseline model representing the personalized health state of the patient; the signal decoupling and correcting module is used for calculating to obtain corrected physiological indexes without the influence of behavior disturbance based on the home behavior data collected in real time; the risk quantitative evaluation module is used for generating a composite risk score representing the current cardiac function deterioration degree based on the deviation degree of the corrected physiological indexes and the physiological baseline model; and the early warning and adaptive optimization module is used for carrying out risk level judgment and triggering corresponding early warning. According to the method, behaviors and physiological signals are decoupled, personalized baselines are constructed, multi-dimensional risks are quantified, closed-loop optimization is formed, the problem of high false alarm and missing alarm in traditional early warning is solved, and accurate early warning and long-term effective monitoring are achieved.

Description

Technical Field

[0001] The present invention relates to the field of disease early warning technology, and in particular to a cardiovascular disease early warning analysis system based on big data. Background Art

[0002] Heart failure, the terminal stage of many cardiovascular diseases, is crucial for post-hospital home management, especially for elderly patients living alone. Existing home monitoring systems typically use wearable devices to collect physiological indicators such as heart rate and blood pressure, setting fixed thresholds for early warning. This provides basic data collection and preliminary early warning methods for patients' home health monitoring.

[0003] Existing home monitoring systems face a core technical challenge. Fluctuations in physiological indicators caused by daily activities, such as housework and emotional agitation, can mimic physiological changes caused by actual deterioration in heart function, creating a fuzzy coupling effect between behavioral and physiological signals. Traditional early warning models are unable to effectively distinguish between these two signal fluctuations, resulting in high false alarm and missed alarm rates, reducing system reliability and timeliness, and making it difficult to meet the needs of real-world home care. Summary of the Invention

[0004] The purpose of the present invention is to provide a cardiovascular disease early warning analysis system and method based on big data, which solves the problems existing in the background technology.

[0005] To solve the above technical problems, the present invention provides a cardiovascular disease early warning analysis system based on big data, comprising: a data processing and modeling module for collecting physiological signal data and home behavior data of a patient, and constructing a physiological baseline model representing the patient's personalized health status based on the physiological signal data under a preset resting steady state; a signal decoupling and correction module for identifying the patient's current activity state based on the real-time collected home behavior data, and combining the current activity state with the real-time collected physiological signal data to calculate corrected physiological indicators that are stripped of the influence of behavioral disturbances through a preset behavior-physiological signal decoupling model; a risk quantification assessment module, configured to calculate and generate a standardized physiological deviation index based on the degree of deviation between the corrected physiological index and the physiological baseline model, and to weight a plurality of the standardized physiological deviation indices with preset clinical weights to generate a composite risk score representing the current degree of deterioration of cardiac function; The early warning and adaptive optimization module is configured to determine the risk level and trigger a corresponding early warning based on the composite risk score, and in response to manual feedback on the early warning, use the real physiological indicators confirmed by the feedback to adaptively optimize the physiological baseline model.

[0006] Preferably, the data processing and modeling module includes the following specific steps: S1, calculating the statistical mean of the time series of all physiological signal data collected in the preset resting steady state to generate a personalized mean; S2, calculating the statistical standard deviation of the time series of the physiological signal data to generate a personalized standard deviation; S3: Using the personalized mean and the personalized standard deviation together as parameters for defining the physiological baseline model.

[0007] Preferably, the signal decoupling and correction module is specifically configured to: S21, calculating, based on the data collected during the calibration period, a difference between the physiological signal data and the personalized mean in each of the activity states, and averaging the differences to generate an average physiological offset corresponding to each activity state; S22 : Subtracting the average physiological offset corresponding to the current activity state from the physiological signal data collected in real time to generate the corrected physiological index.

[0008] Preferably, the risk quantification assessment module is specifically configured to generate the standardized physiological deviation index by: The absolute difference between the corrected physiological index and the personalized mean is calculated, and the absolute difference is divided by the personalized standard deviation to complete normalization and generate the standardized physiological deviation index.

[0009] Preferably, the risk quantification assessment module is further configured to generate the composite risk score: multiplying each of the standardized physiological deviation indices by a preset corresponding clinical weight coefficient to obtain a set of products; and then adding all the products in the set to generate the composite risk score; The clinical weight coefficient is preset based on an expert knowledge base in the field of cardiovascular disease management, or is calibrated through machine learning training on a historical case database.

[0010] Preferably, the early warning and adaptive optimization module is specifically configured to: comparing the composite risk score with a preset first risk threshold and a second risk threshold, wherein the second risk threshold is higher than the first risk threshold; When the composite risk score is higher than the second risk threshold, a level 2 warning is determined, and a detailed alert including the composite risk score and one or more standardized physiological deviation indices that contribute most to the composite risk score is sent to a preset emergency contact; When the composite risk score is higher than the first risk threshold and not higher than the second risk threshold, it is determined to be a level one warning, and a voice prompt is issued to the patient through the smart terminal and the frequency of collecting the physiological signal data is increased; When the composite risk score is not higher than the first risk threshold, it is determined to be a safe state and the current monitoring state is maintained; The first risk threshold and the second risk threshold are determined by performing receiver operating characteristic curve analysis on historical early warning data to achieve a balance between the sensitivity and specificity of the early warning.

[0011] Preferably, the early warning and adaptive optimization module is further configured to adaptively optimize the physiological baseline model and to: When the real physiological index confirmed by the manual feedback is received, the real physiological index and the personalized mean are weighted averaged to calculate and update the personalized mean.

[0012] Beneficial effects Compared with the prior art, the present invention has the following beneficial effects: 1. Through the signal decoupling and correction module, the patient's activity status can be identified in real time, the disturbance of behavior on physiological signals can be removed, and corrected physiological indicators can be generated, which fundamentally solves the problem of fuzzy signal coupling, significantly improves the accuracy of early warning, and reduces the false alarm and missed alarm rate.

[0013] 2. With the help of data processing and modeling modules, a personalized physiological baseline model is constructed based on the patient's resting steady-state data, abandoning the traditional one-size-fits-all fixed threshold model. Risk is assessed using the individual as a reference system, and key physiological changes in specific patients are keenly captured to achieve personalized and precise monitoring.

[0014] 3. The risk quantification assessment module clinically weights the multi-dimensional standardized physiological deviation index to generate a composite risk score. By integrating multi-dimensional information with clinical experience, it intuitively represents the comprehensive risk level of cardiac function deterioration and provides a reliable basis for clinical decision-making.

[0015] 4. The early warning and adaptive optimization module builds a closed-loop mechanism of "monitoring-early warning-intervention-feedback-optimization", dynamically optimizing the physiological baseline model based on manual feedback, so that it can continuously adapt to the patient's long-term physiological changes, ensuring the effectiveness and accuracy of the system's long-term monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention, and those skilled in the art can derive other drawings based on these drawings without inventive effort. Figure 1 It is a logic block diagram of the system of the present invention. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0018] Example 1: See also Figure 1 The present invention provides a cardiovascular disease early warning analysis system based on big data, comprising: a data processing and modeling module for collecting physiological signal data and home behavior data of a patient, and constructing a physiological baseline model representing the patient's personalized health status based on the physiological signal data under a preset resting steady state; a signal decoupling and correction module for identifying the patient's current activity state based on the real-time collected home behavior data, and combining the current activity state with the real-time collected physiological signal data to calculate corrected physiological indicators that are stripped of the influence of behavioral disturbances through a preset behavior-physiological signal decoupling model; a risk quantification assessment module, configured to calculate and generate a standardized physiological deviation index based on the degree of deviation between the corrected physiological index and the physiological baseline model, and to weight a plurality of the standardized physiological deviation indices with preset clinical weights to generate a composite risk score representing the current degree of deterioration of cardiac function; The early warning and adaptive optimization module is configured to determine the risk level and trigger a corresponding early warning based on the composite risk score, and in response to manual feedback on the early warning, use the real physiological indicators confirmed by the feedback to adaptively optimize the physiological baseline model.

[0019] In this embodiment, the system first establishes a unique health benchmark for each heart failure patient through the data processing and modeling module. Afterwards, the signal decoupling and correction module, as one of the cores of the present invention, can accurately identify and quantify the temporary impact of daily activities on physiological signals, and separate it from the original data, effectively solving the problem of false alarms and missed warnings caused by the "behavior-physiological signal fuzzy coupling effect". The risk quantification assessment module converts multi-dimensional, pure physiological indicators into a single risk score with clear clinical significance, providing a quantitative basis for subsequent decision-making. Ultimately, the early warning and adaptive optimization module constitutes a complete closed loop of "monitoring-early warning-intervention-feedback-optimization", which can not only respond to health risks in a timely manner, but also enable the system model to continue to evolve through learning and confirmed data, thereby achieving long-term, accurate and highly personalized home health monitoring.

[0020] Example 2: The data processing and modeling module includes the following specific steps: S1, calculating the statistical mean of the time series of all physiological signal data collected in the preset resting steady state to generate a personalized mean; S2, calculating the statistical standard deviation of the time series of the physiological signal data to generate a personalized standard deviation; S3: Using the personalized mean and the personalized standard deviation together as parameters for defining the physiological baseline model.

[0021] And the signal decoupling and correction module is specifically configured to: S21, calculating, based on the data collected during the calibration period, a difference between the physiological signal data and the personalized mean in each of the activity states, and averaging the differences to generate an average physiological offset corresponding to each activity state; S22 : Subtracting the average physiological offset corresponding to the current activity state from the physiological signal data collected in real time to generate the corrected physiological index.

[0022] In this embodiment, the data processing and modeling module uses Gaussian distribution as the physiological baseline model. The design logic of this model is that the physiological indicators of healthy individuals will fluctuate within a certain range around a stable central value under specific conditions. The mathematical expression of the model establishment is:

[0023] Indicates the The true physiological baseline of physiological indicators; is the number of the physiological indicator; represents Gaussian distribution, that is, normal distribution; For the The personalized mean of each indicator in the resting steady state is calculated by step S1; For the The square root of the personalized variance of the indicator in the resting steady state is Calculated in step S2.

[0024] Existing home monitoring technologies often use a "one-size-fits-all" fixed threshold based on population statistics. For example, a heart rate exceeding 100 beats per minute is uniformly set as abnormal. However, the technical pain point of this approach is that it completely ignores individual physiological differences. For an elderly heart failure patient with a long-term heart rate of 85 beats per minute, a heart rate of 95 beats per minute may be a danger signal; while for an athlete with a baseline heart rate of only 50 beats per minute, a heart rate of 80 beats per minute is completely normal. This non-personalized approach leads to underreporting of the former and false positives of the latter.

[0025] To solve the above-mentioned problems of false positives and false negatives, a unique reference system must be established for each user to represent his or her own normal health status.

[0026] This invention is inspired by basic medical and statistical theories. The physiological indicators of healthy individuals in an undisturbed resting state do not exhibit fixed values, but rather exhibit small, nearly random fluctuations around a stable central value. This distribution characteristic statistically closely matches the Gaussian (normal) distribution model. Therefore, this invention employs a Gaussian distribution to establish a mathematical model for each physiological indicator of each user.

[0027] The physical significance of the physiological baseline model is that it no longer defines "normal" as a fixed value, but as a value that includes personalized mean values ( ) and personalized fluctuation range ( In business terms, this means that the early warning system's judgment basis has changed from "whether it exceeds the general red line" to "whether it has significantly deviated from its normal state," thus achieving truly personalized and precise monitoring.

[0028] Where, Representative The true physiological baseline of physiological indicators (such as heart rate and respiratory rate); represents a Gaussian distribution.

[0029] (Personalized Mean): This parameter is derived from the system's initial calibration phase, which involves collecting multiple consecutive days of user physiological signal data in a preset resting steady state and calculating the arithmetic mean of this time series data (step S1). This parameter is a floating-point value.

[0030] (Personalized variance): The variable source of this parameter is Similarly, it is obtained by calculating the statistical variance (step S2) of the same period of resting steady-state data, which represents the natural fluctuation range of the indicator of the user in a healthy state.

[0031] The physiological baseline model constructed here is the cornerstone of the entire early warning analysis system. It serves as the core input for the subsequent risk quantification assessment module and is the origin and benchmark for all risk calculations.

[0032] By establishing a physiological baseline model, the system creates a customized health profile for each user, eliminating the need for broad, uniform standards. This technology enables the system to discern subtle but critical physiological changes that are clinically significant for a specific patient, significantly improving the specificity and sensitivity of early warnings through quantitative comparison.

[0033] The signal decoupling and correction module uses a correction model based on state average offset. Its design philosophy is that the impact of a specific activity on physiological indicators has a certain statistical regularity in the short term, and the true physiological state can be approximately restored by subtracting the average impact of the activity. The calculation formula for the corrected physiological indicators is:

[0034] Indicates time; Indicates Moment The corrected physiological value of each physiological index is the corrected physiological index; Indicates Moment The original observation values of physiological indicators; Indicates The patient's activity state at all times, such as resting steady state, daily activity state or high-load transient state.

[0035] Indicates the Physiological indicators in active state The average physiological offset under the above conditions is calculated in step S21.

[0036] Existing technologies that directly analyze raw physiological signals suffer from a technical difficulty in distinguishing between pathological changes and physiological fluctuations. For example, the accelerated heart rate of a heart failure patient ascending stairs can mimic the tachycardia caused by deteriorating cardiac function. This is known as the "behavioral-physiological signal ambiguity coupling effect." This confusion is the root cause of the high false alarm rate in existing systems.

[0037] In order to accurately assess the true health status, it is necessary to first accurately "strip off" the temporary, benign effects caused by daily activities from the original physiological signals.

[0038] The design concept of this invention is based on the principle of "noise cancellation" in signal processing. If the noise pattern—the average effect of a specific activity on a physiological signal—is known in advance, it can be subtracted from the noisy signal to restore a cleaner original signal. This invention treats the effects of activity as a form of physiological noise.

[0039] The physical meaning of is that it estimates At all times, even when the user is active After eliminating the influence of this activity, its physiological indicators are equivalent to the values in the resting state. It is a "pure" physiological indicator that better reflects the user's internal cardiac function compensation state.

[0040] Parameter definition: : Moment Corrected physiological values of physiological indicators; : The original observation value collected by the sensor at the moment; : Core parameter, representing the activity status For the first The average physiological deviation caused by the physiological indicators.

[0041] Variable source: The variables are derived from step S21. During the calibration period, the system requires the user to wear the device to perform different types of activities while collecting physiological data and activity status. The system then calculates the physiological indicators relative to the personalized resting mean in each activity state. The average value of the difference between (such as rest, daily activities, high-load transient) value.

[0042] Calculated corrected physiological indicators This will serve as a direct input for calculating the standardized physiological deviation index in the next risk quantification assessment module.

[0043] The technical effect is a significant improvement in the accuracy of early warnings. By effectively removing activity interference, the system can focus on identifying pathological changes truly caused by deteriorating cardiac function, fundamentally resolving the "fuzzy coupling" problem, making the triggering of early warning events more precise and greatly improving the system's reliability and practical value.

[0044] Example 3: In order to generate the standardized physiological deviation index, the risk quantification assessment module is specifically configured to: The absolute difference between the corrected physiological index and the personalized mean is calculated, and the absolute difference is divided by the personalized standard deviation to complete normalization and generate the standardized physiological deviation index.

[0045] Furthermore, the risk quantification assessment module is further configured to generate the composite risk score: multiplying each of the standardized physiological deviation indices by a preset corresponding clinical weight coefficient to obtain a set of products; and then adding all the products in the set to generate the composite risk score; The clinical weight coefficient is preset based on an expert knowledge base in the field of cardiovascular disease management, or is calibrated through machine learning training on a historical case database.

[0046] In this embodiment, the risk quantification assessment module first calculates the standardized physiological deviation index. This index is designed using statistical standardization methods to transform physiological indicators of different units and different degrees of variation into the same dimensionless evaluation scale, allowing for direct comparison and integration. Its calculation formula is:

[0047] : Moment The standardized physiological deviation index of each indicator is a dimensionless floating-point value; : The corrected physiological value calculated in the previous step; : Personalized mean from physiological baseline model; : Personalized standard deviation from the physiological baseline model (note that square root of ).

[0048] After obtaining calibrated physiological indicators, directly comparing their absolute deviations from the mean presents a new technical pain point. For example, if the heart rate deviates by 10 beats / minute and the respiratory rate deviates by 5 breaths / minute, which of these two deviations is more dangerous? Because their units (dimensions) and inherent volatility (degree of variability) are completely different, they cannot be directly compared or added together.

[0049] In order to measure and integrate the risk levels of different physiological indicators on the same scale, the deviation degree of each indicator needs to be standardized.

[0050] This paper explicitly incorporates the crucial statistical concept of the "Z-score" as its scientific basis. The core idea of the Z-score is to measure how many standard deviations a data point is from its mean. It is a natural, dimensionless metric that perfectly addresses the issue of scaling different indicators.

[0051] The physical meaning of is that it measures not the “absolute value” of the deviation, but the “significance” or “relative fluctuation” of the deviation. A value of 2.0 has the same business meaning, regardless of whether it corresponds to heart rate or breathing: "The current indicator has deviated from the individual's normal level by two standard deviations." This is a universally comparable risk measure.

[0052] : Moment The standardized physiological deviation index of each indicator is a dimensionless floating-point value; : The corrected physiological value calculated in the previous step; : Personalized mean from physiological baseline model; : Personalized standard deviation from the physiological baseline model (note that square root of ).

[0053] By dividing the numerator (unit: X) by the denominator (unit: X), dimensionlessness is finally achieved.

[0054] Value range: The value range of is a real number greater than or equal to 0. The larger the value, the further away from the personal norm, and the higher the risk.

[0055] A calculated set of (One for each indicator) will serve as input to the next step "weighted fusion module" to calculate the composite risk score.

[0056] The technical effect lies in enabling the "co-evaluation" of multi-dimensional physiological information. It transforms data of different units and varying volatility into a unified evaluation scale, laying the mathematical foundation for subsequent meaningful weighted fusion and comprehensive evaluation, making comprehensive risk assessment possible.

[0057] The module then calculates a composite risk score using a weighted fusion model. This model is designed to integrate information from multiple physiological dimensions and assign risk contributions to different indicators based on clinical medical knowledge, thereby generating a comprehensive, single risk assessment result. The calculation formula is:

[0058] For the Physiological indicators in Standardized physiological deviation index at the moment; 、 、 Same as above; For Composite risk score at the moment; is the total number of core physiological indicators used for risk assessment, which is a system constant preset according to clinical guidelines.

[0059] For the The clinical weight coefficient of each indicator is preset by the medical expert knowledge base or obtained through machine learning training of historical case data.

[0060] Complex diseases like heart failure often develop as a result of the interaction of multiple physiological systems. Monitoring only a single indicator can easily lead to overgeneralization and fail to fully reflect the overall risk.

[0061] A model is needed that can integrate the risk information contained in multiple "standardized physiological deviation indices" into a single total score that can comprehensively characterize the current comprehensive risk level of deterioration of cardiac function.

[0062] Weighted summation is a classic method in multi-attribute decision-making theory and index construction. Its technical significance lies in its ability to intuitively reflect the importance of different factors in the final decision. This paper draws on this concept to assign corresponding weights to the risk contributions of different physiological indicators.

[0063] The business significance of is that it no longer presents a bunch of scattered data to doctors or users, but provides an intuitive and quantifiable single risk snapshot. For example, A score of 7.5 is more dangerous than a score of 3.2, and this score is directly related to clinical decision-making.

[0064] Weight coefficient This is the core of the formula. Its value is not set in vain. The present invention provides a clear determination method: When the system is initialized or there is a lack of sufficient historical data, the weights can be set by a group of cardiovascular experts based on clinical guidelines. For example, for patients with heart failure, the weights of dyspnea and nocturnal paroxysmal dyspnea are will be weighted higher than heart rate variability.

[0065] Machine learning training: A more advanced and accurate method is to use massive amounts of annotated historical case data and use machine learning algorithms such as logistic regression, support vector machine (SVM), or gradient boosting tree for supervised learning training. The algorithm will automatically optimize and find a set of weights that can maximize the accuracy of the prediction model. .

[0066] The weight coefficients are designed as background configurable items, and the system has adaptive capabilities. It can regularly retrain the model with newly accumulated case data and automatically update the weights to adapt to the development of disease cognition or changes in specific patient groups.

[0067] Calculated composite risk score , which will be directly used for the threshold judgment in the next step.

[0068] The system presets at least two risk thresholds (first risk threshold and second risk threshold). By comparing with these thresholds, automatic classification of "safe status", "first level warning" and "second level warning" is achieved.

[0069] Compared with the existing technology that provides scattered indicators, the present invention provides It quantitatively compares multi-dimensional information and incorporates clinical experience knowledge. Its technical effect is to output a more comprehensive, intuitive, and information-dense comprehensive risk level, providing a more reliable and integrated basis for subsequent precise graded warnings and clinical decision-making.

[0070] Example 4: The early warning and adaptive optimization module is specifically configured to: comparing the composite risk score with a preset first risk threshold and a second risk threshold, wherein the second risk threshold is higher than the first risk threshold; When the composite risk score is higher than the second risk threshold, a level 2 warning is determined, and a detailed alert including the composite risk score and one or more standardized physiological deviation indices that contribute most to the composite risk score is sent to a preset emergency contact; When the composite risk score is higher than the first risk threshold and not higher than the second risk threshold, it is determined to be a level one warning, and a voice prompt is issued to the patient through the smart terminal and the frequency of collecting the physiological signal data is increased; When the composite risk score is not higher than the first risk threshold, it is determined to be a safe state and the current monitoring state is maintained; The first risk threshold and the second risk threshold are determined by performing receiver operating characteristic curve analysis on historical early warning data to achieve a balance between the sensitivity and specificity of the early warning.

[0071] The early warning and adaptive optimization module is further configured to adaptively optimize the physiological baseline model and is used to: When the real physiological index confirmed by the manual feedback is received, the real physiological index and the personalized mean are weighted averaged to calculate and update the personalized mean.

[0072] In this embodiment, the risk level determination function of the early warning and adaptive optimization module is based on the composite risk score. The threshold is usually determined by performing receiver operating characteristic curve analysis on historical data to balance the sensitivity and specificity of the warning. The model's adaptive optimization function uses an exponentially weighted moving average algorithm. The design logic of this algorithm is to use newly confirmed real data to fine-tune the old baseline mean, and use a smaller learning rate factor to ensure the smoothness of the model update, so that it can slowly adapt to the patient's long-term physiological changes rather than reacting violently to a single feedback. The calculation formula for updating the personalized mean is:

[0073] :Updated New personalized means for each indicator; : The old personalized mean before updating; : learning rate factor, a preset small constant; Key inputs are not arbitrary real-time data. Instead, they are derived from actual physiological status values confirmed by medical staff or family members after the system triggers an alert, through manual intervention and verification. This is validated, high-quality feedback data.

[0074] Existing technology flaws / pain points: Existing monitoring models are mostly static and immutable once deployed. However, human health evolves dynamically. Whether due to the natural progression of a condition, aging, or improvement through effective treatment, the physiological baseline undergoes long-term, slow changes. As a result, static models gradually become inaccurate, leading to a decline in accuracy.

[0075] A closed-loop feedback mechanism is needed to enable the physiological baseline model to "self-evolve" and "continuously learn" so as to dynamically adapt to the user's long-term health trend changes and maintain the long-term effectiveness of the model.

[0076] This paper draws on the classic "exponentially weighted moving average" algorithm from time series analysis as its technical inspiration. The core idea of EWMA is to assign higher weights to recent data and exponentially decaying weights to distant data, thereby effectively tracking long-term trends while smoothing short-term fluctuations.

[0077] Physical meaning: The physical meaning of this formula is to make the personalized mean Towards a "new true value" confirmed by manual feedback from professionals , making small, steady adjustments. This ensures that the model’s baseline can slowly and steadily follow the patient’s true long-term physiological changes, rather than reacting dramatically to a single feedback.

[0078] Determination of: learning rate Is the key adjustable parameter in this formula, its value range is between (0,1), and its selection is a trade-off: smaller This means that the update is slow, the model is very stable, can effectively filter out accidental and noisy feedback, and focuses more on long-term trends; larger This means that the model is more sensitive to new feedback and can adapt to changes faster.

[0079] In practice, The value of is typically set empirically based on clinical needs. For example, for chronic disease management where the condition changes slowly, a very small value may be chosen to ensure model stability. This value can also be configured as a backend configurable option.

[0080] This update process forms a complete closed-loop management system of "monitoring-early warning-intervention-feedback-optimization." This update process is triggered when the system receives human feedback.

[0081] Compared to static models, the adaptive optimization module designed in this invention is one of its core advances. Its technical effect is to endow the system with the ability to self-evolve and continuously optimize. The model can synchronize with the patient's long-term health trends, ensuring continued effectiveness and accuracy throughout the long-term monitoring cycle, achieving truly dynamic and intelligent health management.

[0082] 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 cardiovascular disease early warning analysis system based on big data, characterized by: include: A data processing and modeling module is used to collect the patient's physiological signal data and home behavior data, and construct a physiological baseline model representing the patient's personalized health status based on the physiological signal data in a preset resting steady state; a signal decoupling and correction module for identifying the patient's current activity state based on the real-time collected home behavior data, and combining the current activity state with the real-time collected physiological signal data to calculate corrected physiological indicators that are stripped of the influence of behavioral disturbances through a preset behavior-physiological signal decoupling model; a risk quantification assessment module, configured to calculate and generate a standardized physiological deviation index based on the degree of deviation between the corrected physiological index and the physiological baseline model, and to weight a plurality of the standardized physiological deviation indices with preset clinical weights to generate a composite risk score representing the current degree of deterioration of cardiac function; The early warning and adaptive optimization module is configured to determine the risk level and trigger a corresponding early warning based on the composite risk score, and in response to manual feedback on the early warning, use the real physiological indicators confirmed by the feedback to adaptively optimize the physiological baseline model.

2. A cardiovascular disease early warning analysis system based on big data according to claim 1, characterized in that: The data processing and modeling module includes the following specific steps: S1, calculating the statistical mean of the time series of all physiological signal data collected in the preset resting steady state to generate a personalized mean; S2, calculating the statistical standard deviation of the time series of the physiological signal data to generate a personalized standard deviation; S3: Using the personalized mean and the personalized standard deviation together as parameters for defining the physiological baseline model.

3. A cardiovascular disease early warning analysis system based on big data according to claim 2, characterized in that: The signal decoupling and correction module is specifically configured to: S21, calculating, based on the data collected during the calibration period, a difference between the physiological signal data and the personalized mean in each of the activity states, and averaging the differences to generate an average physiological offset corresponding to each activity state; S22 : Subtracting the average physiological offset corresponding to the current activity state from the physiological signal data collected in real time to generate the corrected physiological index.

4. A cardiovascular disease early warning analysis system based on big data according to claim 3, characterized in that: The risk quantification assessment module is specifically configured to generate the standardized physiological deviation index and is used to: The absolute difference between the corrected physiological index and the personalized mean is calculated, and the absolute difference is divided by the personalized standard deviation to complete normalization and generate the standardized physiological deviation index.

5. The cardiovascular disease early warning analysis system based on big data according to claim 4, characterized in that: To generate the composite risk score, the risk quantification assessment module is further configured to: multiplying each of the standardized physiological deviation indices by a preset corresponding clinical weight coefficient to obtain a set of products; and then adding all the products in the set to generate the composite risk score; The clinical weight coefficient is preset based on an expert knowledge base in the field of cardiovascular disease management, or is calibrated through machine learning training on a historical case database.

6. A cardiovascular disease early warning analysis system based on big data according to claim 5, characterized in that: The early warning and adaptive optimization module is specifically configured to: comparing the composite risk score with a preset first risk threshold and a second risk threshold, wherein the second risk threshold is higher than the first risk threshold; When the composite risk score is higher than the second risk threshold, a level 2 warning is determined, and a detailed alert including the composite risk score and one or more standardized physiological deviation indices that contribute most to the composite risk score is sent to a preset emergency contact; When the composite risk score is higher than the first risk threshold and not higher than the second risk threshold, it is determined to be a level one warning, and a voice prompt is issued to the patient through the smart terminal and the frequency of collecting the physiological signal data is increased; When the composite risk score is not higher than the first risk threshold, it is determined to be a safe state and the current monitoring state is maintained; The first risk threshold and the second risk threshold are determined by performing receiver operating characteristic curve analysis on historical early warning data to achieve a balance between the sensitivity and specificity of the early warning.

7. The cardiovascular disease early warning analysis system based on big data according to claim 6, characterized in that: The early warning and adaptive optimization module is further configured to adaptively optimize the physiological baseline model and to: When the real physiological index confirmed by the manual feedback is received, the real physiological index and the personalized mean are weighted averaged to calculate and update the personalized mean.

Images