Intelligent hypertension health management system

Abstract

The invention, which relates to the medical health management field, discloses an intelligent hypertension health management system comprising a data acquisition module, a data processing module, an early warning module and an interaction module. By constructing the cerebral circulation-blood pressure association model, the potential association relationship between the cerebral hemodynamic parameters and the blood pressure data can be accurately mined, so that the possibility of blood pressure abnormality caused by cerebral vasculopathy is effectively judged, and when abnormality is detected, the system can send out an early warning signal in time, so that the early warning effect is improved. And according to the abnormal degree and the historical health data of the user, customized health suggestions are provided for the user through a personalized health suggestion generation algorithm, and the combination of the precise early warning and the personalized health suggestions helps the user to timely discover and process potential health problems and improve the health management effect.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of medical health management, in particular to an intelligent high blood pressure health management system. BACKGROUND

[0002] Traditional high blood pressure health management methods often have many limitations, on the one hand, traditional blood pressure monitoring equipment has single function, can only provide blood pressure data, and cannot fully reflect the physiological condition of the patient, especially cannot directly monitor the health status of the brain blood vessels, on the other hand, the traditional health management system has limited ability in data analysis, and it is difficult to mine the potential correlation between blood pressure data and other physiological parameters, resulting in insufficient accuracy and timeliness of early warning.

[0003] In summary, the traditional high blood pressure health management technology has the shortcomings of incomplete monitoring and limited data analysis capability, and it is difficult to meet the needs of modern high blood pressure patients for accurate and efficient health management, therefore, it is particularly important to develop an intelligent high blood pressure health management system. SUMMARY

[0004] The purpose of the present application is to make up for the shortcomings of the prior art, and to provide an intelligent high blood pressure health management system, which can realize real-time and continuous monitoring and management of the health status of high blood pressure patients by integrating data acquisition module, data processing module, early warning module and interactive module, especially, the system can accurately mine the potential correlation between cerebral hemodynamics parameters and blood pressure data by constructing a brain circulation-blood pressure correlation model, so as to effectively judge the possibility of blood pressure abnormality caused by brain blood vessel lesions, and timely issue an early warning signal when an abnormality is found, and provide personalized health advice according to the user's situation.

[0005] To solve the above technical problems, the present application provides the following technical scheme: an intelligent high blood pressure health management system, which comprises the following components: data acquisition module, data processing module, early warning module and interactive module; Data acquisition module: according to the preset time interval, the blood pressure data of the user is collected, including systolic pressure, diastolic pressure and pulse information, and the data is transmitted to the data processing module, based on near-infrared spectroscopy technology, including a plurality of near-infrared light sources and light detectors, the near-infrared light sources emit near-infrared light of different wavelengths to the specific area of the user's brain, the light detectors receive the reflected light scattered by the brain tissue, by analyzing the intensity and spectral characteristics of the reflected light, the blood oxygen saturation, cerebral blood flow, cerebral vascular resistance and cerebral hemodynamics parameters of the brain tissue are calculated, and the data is transmitted to the data processing module in real time; Data processing module: Preprocesses received blood pressure data and cerebral hemodynamic parameters to construct a cerebral circulation-blood pressure correlation model. This model, based on a machine learning algorithm, is trained using a large amount of historical blood pressure data, cerebral hemodynamic parameters, and corresponding clinical diagnosis results. The model analyzes real-time collected data, explores the potential correlation between cerebral hemodynamic parameters and blood pressure data, and determines the possibility of abnormal blood pressure caused by cerebral vascular disease. Based on the preprocessed data, a time series analysis algorithm is used to predict the changing trends of blood pressure and cerebral hemodynamic parameters, providing a more accurate basis for subsequent early warnings. Early Warning Module: Based on medical standards and extensive clinical data, it sets normal range thresholds, mild abnormality thresholds, and severe abnormality thresholds for cerebral hemodynamic parameters and blood pressure data. When the correlation analysis submodule determines the possibility of abnormal blood pressure caused by cerebral vascular disease, or the trend prediction submodule predicts that the relevant parameters are about to exceed the set thresholds, the early warning trigger unit issues different levels of early warning signals based on the degree of abnormality. These signals are displayed in eye-catching colors and icons on the interface of the interactive module. At the same time, the early warning information will include preliminary health advice. Interactive module: provides an intuitive user operation interface. The interface design supports chart display, which allows users to intuitively understand their own health status. Users can input instructions to the system through the interactive module, and the system will perform corresponding operations and feedback based on the user instructions.

[0006] Furthermore, the cerebral hemodynamics acquisition unit in the data acquisition module is a head-mounted device, which is equipped with multiple near-infrared light sources and light detectors. The near-infrared light source emits near-infrared light with wavelengths of 760nm and 850nm to specific areas of the user's brain. The light detector receives the reflected light after scattering by the brain tissue, and calculates the blood oxygen saturation, cerebral blood flow, cerebral vascular resistance and cerebral hemodynamic parameters of the brain tissue by analyzing the intensity and spectral characteristics of the reflected light. The light source and detector layout of the unit have been optimized to cover the main blood supply areas of the brain to improve detection accuracy and comfort. Specifically, by conducting 300 simulated detection experiments on 100 users of different body shapes and head shapes, the layout was optimized based on the detection error and user feedback, so that the detection error was reduced to less than 5%.

[0007] Furthermore, the data processing module adopts an improved outlier detection algorithm based on local anomaly factors, which introduces a dynamic weight factor , adaptively adjust the abnormality judgment standard according to the data distribution density, the formula is: ; in for point The local density of is the mean of the local density of all points, To adjust the parameters, 2000 sets of clinical samples containing normal data and known abnormal data were trained to determine The optimal value is 0.5. Compared with the traditional local anomaly factor algorithm, the accuracy of identifying abnormal data is increased by 20%, which effectively improves the data quality and ensures the accuracy of subsequent data analysis.

[0008] Furthermore, the cerebral circulation-blood pressure association model constructed in the data processing module is based on the improved deep attention network algorithm. The model input layer receives the preprocessed blood pressure data and cerebral hemodynamic parameters. In the attention mechanism layer, the formula Calculate the attention weight, where No. The input and The attention weights between the inputs, is the natural exponential function, is the index variable for summation, is the characteristic dimension of blood pressure data, is the characteristic dimension of cerebral hemodynamic parameters, , is the activation function, and is a trainable parameter, For the general and Perform splicing operations. For the The hidden state of time steps, For the The hidden state of each input was determined through 50 rounds of iterative training on historical data of 1,000 hypertensive patients, enabling the model to more accurately explore the potential correlation between cerebral hemodynamic parameters and blood pressure data. Compared with traditional neural network models, the accuracy of judging abnormal blood pressure caused by cerebral vascular lesions has increased by 15%.

[0009] Furthermore, the data processing module predicts the changing trends of blood pressure and cerebral hemodynamic parameters using a prediction algorithm based on the fusion of chaos theory and grey model. The algorithm first adopts an improved false neighbor method and uses the formula Determine the embedding dimension of the data phase space reconstruction ,in No. dimensional embedding space The false neighbor ratio of a point, is the index of the data point, is the total number of data points, is the index variable for summation, For dimensional embedding space reconstruction vectors, For dimensional embedding space reconstruction vector, when Less than the threshold for the first time Time That is the optimal embedding dimension, combined with the autocorrelation function method to determine the delay time , then reconstructed the gray prediction model and used chaotic sequence optimization. Compared with a single prediction algorithm, the prediction error of the changing trend of blood pressure and cerebral hemodynamic parameters was reduced by 18%, providing a more accurate basis for early warning.

[0010] Furthermore, the threshold setting unit of the early warning module sets normal range thresholds, mild abnormality thresholds and severe abnormality thresholds for cerebral hemodynamic parameters and blood pressure data based on medical standards and a large amount of clinical data. For cerebral vascular resistance parameters, the normal range is determined based on statistical analysis of test data from 5,000 healthy adults. The mild abnormality threshold is set to 110% of the upper limit of the normal range, and the severe abnormality threshold is set to 120% of the upper limit of the normal range. The thresholds will be dynamically adjusted through a weighted average algorithm based on individual differences in user age, gender, and underlying diseases. The weights are determined through comparative analysis of clinical data of people with different characteristics, so that the early warning thresholds are more in line with individual actual conditions, thereby improving the effectiveness of the early warning.

[0011] Furthermore, when the warning trigger unit of the warning module issues a warning signal, it provides customized health advice to the user based on the degree of abnormality and the user's historical health data through a personalized health advice generation algorithm. The algorithm comprehensively considers the user's dietary preferences, exercise habits, and medication history factors, and uses a decision tree algorithm to build a recommendation model. The branch nodes and thresholds of the decision tree are determined by training the treatment effects and feedback data of 1,000 patients in different situations. When it is detected that the user's blood pressure rises due to a high-salt diet, the system will recommend low-salt alternative recipes based on the user's taste preferences to enhance the user's compliance with health management.

[0012] Furthermore, the user interface unit of the interactive module supports simultaneous display of multiple devices, including mobile phone apps, smart watches and dedicated health management terminals, and each device interface adopts an adaptive layout algorithm to automatically adjust the size and position of interface elements according to the device screen size, resolution and display ratio, ensuring that users can clearly and conveniently view real-time blood pressure data, cerebral hemodynamic parameters, warning information and historical health data reports on different devices. The algorithm determines relevant parameters by conducting compatibility tests and parameter optimization on 50 common smart devices on the market, thereby improving the user experience.

[0013] Compared with the prior art, the intelligent high blood pressure health management system has the following beneficial effects: I. The system can accurately mine the potential correlation between cerebral hemodynamic parameters and blood pressure data by constructing a brain circulation-blood pressure correlation model, thereby effectively determining the possibility of blood pressure abnormalities caused by brain vascular lesions, and when detecting abnormalities, the system will timely issue a warning signal, and according to the abnormal degree and user historical health data, provide customized health suggestions for the user through a personalized health suggestion generation algorithm. The combination of accurate early warning and personalized health suggestions helps users to discover and handle potential health problems in time, and improves the effect of health management.

[0014] II. The interaction module of the system supports multi-device synchronous display, including mobile phone APP, smart watch and special health management terminal, etc., ensuring that users can clearly and conveniently view real-time blood pressure data, cerebral hemodynamic parameters, warning information and historical health data reports on different devices. At the same time, the adaptive layout algorithm is used in the interface of each device to automatically adjust the size and position of the interface elements according to the screen size, resolution and display ratio of the device. This design greatly improves the user experience, allowing users to conveniently manage their health status anytime and anywhere.

[0015] Other advantages, objects and features of the present application will be set forth in part in the following specification, and in part will become apparent to those skilled in the art upon examination of the following specification, or can be learned by practice of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creating any inventive labor.

[0017] Figure 1 It is a functional implementation flowchart of an intelligent high blood pressure health management system.

[0018] Figure 2 It is a structural schematic diagram of an intelligent high blood pressure health management system. DETAILED DESCRIPTION

[0019] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following will combine the drawings and preferred embodiments to specifically describe the specific embodiments, structures, features and effects of the present application.

[0020] Embodiment one

[0021] Community residents wear head-mounted cerebral hemodynamic acquisition devices, which integrate multiple near-infrared light sources and light detectors to emit near-infrared light of different wavelengths to specific areas of the brain. At the same time, a cuff-type blood pressure monitor collects systolic blood pressure, diastolic blood pressure and pulse data at preset time intervals. The collected blood pressure data and cerebral hemodynamic parameters (including blood oxygen saturation, cerebral blood flow, cerebral vascular resistance, etc.) are transmitted to the system data processing module in real time.

[0022] The data processing module first pre-processes the received data to remove noise and outliers, and then uses the improved outlier detection algorithm based on the local anomaly factor. The formula is: This algorithm introduces a dynamic weight factor , adaptively adjust the abnormality judgment standard according to the data distribution density, the formula is: ; in for point The local density of is the mean of the local density of all points, In order to adjust the parameters, the abnormal judgment criteria are adaptively adjusted according to the data distribution density, and abnormal data points are identified. Then, the cerebral circulation-blood pressure association model based on the improved deep attention network algorithm is used. The formula is: The model input layer receives the preprocessed blood pressure data and cerebral hemodynamic parameters. In the attention mechanism layer, the formula Calculate the attention weight, where , is the activation function, and As a trainable parameter, the potential correlation between cerebral hemodynamic parameters and blood pressure data is analyzed to determine whether there is a possibility of abnormal blood pressure caused by cerebral vascular lesions. Finally, a prediction algorithm based on the fusion of chaos theory and grey model is adopted. The formula is: The algorithm first adopts the improved false neighbor method, and then uses the formula Determine the embedding dimension of the data phase space reconstruction ,when Less than the threshold for the first time Time That is the optimal embedding dimension, combined with the autocorrelation function method to determine the delay time , then construct a grey prediction model and use chaotic sequence optimization to predict the changing trends of blood pressure and cerebral hemodynamic parameters.

[0023] The early warning module dynamically adjusts the threshold based on medical standards and clinical data, combined with individual differences such as residents' age and gender, and sets the ranges of normal, mild abnormality and severe abnormality. When the associated model determines that there is a possibility of blood pressure abnormality or the predicted parameters are about to exceed the threshold, the early warning trigger unit will issue different levels of early warning signals according to the degree of abnormality, which are displayed in eye-catching colors and icons on the interactive module interface, and at the same time generate preliminary health guidance including diet and exercise recommendations.

[0024] Residents can view chart analyses of their own blood pressure and cerebral blood flow parameters through the visual interface of the community health management terminal. If the system detects that a resident's blood pressure is increasing due to a long-term high-salt diet, the interactive module will recommend low-salt alternative recipes based on their taste preferences using a recommendation model constructed using a decision tree algorithm, and remind them to have regular check-ups. Community doctors can also obtain resident data through the terminal background for further diagnosis and intervention.

[0025] Example 2

[0026] Elderly patients use smart health monitors with integrated head-mounted collection units at home. The device automatically collects blood pressure data and cerebral hemodynamic parameters at a set frequency (such as every 30 minutes). The near-infrared light source emits light of a specific wavelength to the main blood supply areas of the brain. The light detector receives the reflected light and calculates the relevant parameters. All data is transmitted to the home gateway in real time via the wireless network and then synchronized to the cloud data processing module.

[0027] The cloud-based data processing module pre-processes the received data from elderly patients, uses an improved local anomaly factor algorithm to detect outliers to ensure data accuracy, and then uses the cerebral circulation-blood pressure association model constructed by the improved deep attention network algorithm to analyze the correlation characteristics between cerebral blood flow and blood pressure in elderly patients, and determine whether there is a risk of blood pressure fluctuations due to cerebral vascular disease. At the same time, a prediction algorithm that integrates chaos theory and gray model is used to predict the short-term change trends of patients' blood pressure and cerebral blood flow parameters.

[0028] The early warning module dynamically adjusts the threshold based on the characteristics of elderly patients who have more underlying diseases, taking into account their age, gender and medical history. When the correlation analysis finds that the patient's cerebral blood flow parameters are abnormal and may cause a sudden increase in blood pressure, or the trend forecast shows that the systolic blood pressure is about to exceed the severe abnormal threshold, the early warning trigger unit immediately issues a high-level early warning signal, which is also pushed to the patient's family's mobile phone APP, the patient's smart watch and home smart terminal. The interface reminds in the form of a red icon and voice broadcast. The health advice includes emergency blood pressure reduction measures and medical guidance.

[0029] Patients can view charts and trend analysis of their own health data through smart watches or mobile phone apps. The interface uses an adaptive layout algorithm to adapt to different device screens. If the system detects that the patient's blood pressure fluctuates due to reduced exercise recently, the interactive module will combine their exercise habits and medication history to generate a personalized exercise plan through a decision tree algorithm, such as recommending suitable indoor Tai Chi classes and simultaneously reminding them to take medication on time. Family members can also remotely view patient data through the app and communicate with community doctors online to achieve real-time monitoring and intervention at home.

[0030] The above description is merely a preferred embodiment of the present invention and does not constitute any form of limitation to the present invention. Although the present invention has been disclosed as above in terms of a preferred embodiment, it is not intended to limit the present invention. Any person skilled in the art can, without departing from the scope of the technical solution of the present invention, make some changes or modifications to equivalent embodiments using the technical contents disclosed above. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. An intelligent hypertension health management system, characterized in that: The system includes the following components: data acquisition module, data processing module, early warning module and interaction module; Data acquisition module: collects the user's blood pressure data, including systolic pressure, diastolic pressure, and pulse information, at preset time intervals and transmits the data to the data processing module. Based on near-infrared spectral imaging technology, it includes multiple near-infrared light sources and light detectors. The near-infrared light sources emit near-infrared light of different wavelengths to specific areas of the user's brain. The light detectors receive the reflected light after scattering by the brain tissue. By analyzing the intensity and spectral characteristics of the reflected light, the blood oxygen saturation of the brain tissue, cerebral blood flow, cerebral vascular resistance, and cerebral hemodynamic parameters are calculated, and the data is transmitted to the data processing module in real time; Data processing module: Preprocesses received blood pressure data and cerebral hemodynamic parameters to construct a cerebral circulation-blood pressure correlation model. This model, based on a machine learning algorithm, is trained using a large amount of historical blood pressure data, cerebral hemodynamic parameters, and corresponding clinical diagnosis results. The model analyzes real-time collected data, explores the potential correlation between cerebral hemodynamic parameters and blood pressure data, and determines the possibility of abnormal blood pressure caused by cerebral vascular disease. Based on the preprocessed data, a time series analysis algorithm is used to predict the changing trends of blood pressure and cerebral hemodynamic parameters. Early Warning Module: Based on medical standards and extensive clinical data, it sets normal range thresholds, mild abnormality thresholds, and severe abnormality thresholds for cerebral hemodynamic parameters and blood pressure data. When the correlation analysis submodule determines the possibility of abnormal blood pressure caused by cerebral vascular disease, or the trend prediction submodule predicts that the relevant parameters are about to exceed the set thresholds, the early warning trigger unit issues different levels of early warning signals based on the degree of abnormality. These signals are displayed in eye-catching colors and icons on the interface of the interactive module. At the same time, the early warning information will include preliminary health advice. Interactive module: provides an intuitive user operation interface. The interface design supports chart display. Users can input instructions to the system through the interactive module, and the system performs corresponding operations and feedback based on the user instructions.

2. The intelligent hypertension health management system according to claim 1, characterized in that: The cerebral hemodynamics acquisition unit in the data acquisition module is a head-mounted device, which is equipped with multiple near-infrared light sources and light detectors. The near-infrared light source emits near-infrared light with wavelengths of 760nm and 850nm to specific areas of the user's brain. The light detector receives the reflected light after scattering by the brain tissue and calculates the blood oxygen saturation, cerebral blood flow, cerebral vascular resistance and cerebral hemodynamic parameters of the brain tissue by analyzing the intensity and spectral characteristics of the reflected light. The layout of the light source and detector of this unit has been optimized to cover the main blood supply areas of the brain.

3. The intelligent hypertension health management system according to claim 1, characterized in that: The data processing module adopts an outlier detection algorithm based on the local anomaly factor improvement, which introduces a dynamic weight factor , adaptively adjust the abnormality judgment standard according to the data distribution density, the formula is: ； in for point The local density of is the mean of the local density of all points, is the adjustment parameter.

4. The intelligent hypertension health management system according to claim 1, characterized in that: The cerebral circulation-blood pressure association model constructed in the data processing module is based on the improved deep attention network algorithm. The model input layer receives the preprocessed blood pressure data and cerebral hemodynamic parameters. In the attention mechanism layer, the formula Calculate the attention weight, where No. The input and The attention weights between the inputs, is the natural exponential function, is the index variable for summation, is the characteristic dimension of blood pressure data, is the characteristic dimension of cerebral hemodynamic parameters, , is the activation function, and is a trainable parameter, For the general and Perform splicing operations. For the The hidden state of time steps, For the The hidden state of the input.

5. The intelligent hypertension health management system according to claim 1, characterized in that: The data processing module predicts the changing trend of blood pressure and cerebral hemodynamic parameters using a prediction algorithm based on the fusion of chaos theory and grey model. The algorithm first adopts an improved false neighbor method and uses the formula Determine the embedding dimension of the data phase space reconstruction ,in No. dimensional embedding space The false neighbor ratio of a point, is the index of the data point, is the total number of data points, is the index variable for summation, For dimensional embedding space reconstruction vectors, For dimensional embedding space reconstruction vector, when Less than the threshold for the first time Time That is the optimal embedding dimension, combined with the autocorrelation function method to determine the delay time , then construct the grey prediction model and use chaotic sequence optimization.

6. The intelligent hypertension health management system according to claim 1, characterized in that: The threshold setting unit of the early warning module sets normal range thresholds, mild abnormality thresholds and severe abnormality thresholds for cerebral hemodynamic parameters and blood pressure data based on medical standards and a large amount of clinical data. The thresholds will be dynamically adjusted through a weighted average algorithm based on individual differences in user age, gender and underlying diseases.

7. The intelligent hypertension health management system according to claim 1, characterized in that: When the warning trigger unit of the warning module issues a warning signal, it provides customized health advice to the user based on the degree of abnormality and the user's historical health data through a personalized health advice generation algorithm. The algorithm comprehensively considers the user's dietary preferences, exercise habits, and medication history factors, and uses a decision tree algorithm to build a recommendation model. When it is detected that the user's blood pressure rises due to a high-salt diet, the system will recommend low-salt alternative recipes based on the user's taste preferences, thereby enhancing the user's compliance with health management.

8. The intelligent hypertension health management system according to claim 1, characterized in that: The user interface unit of the interactive module supports simultaneous display on multiple devices, including mobile phone apps, smart watches, and dedicated health management terminals. Each device interface uses an adaptive layout algorithm to automatically adjust the size and position of interface elements according to the device screen size, resolution, and display ratio.

Images