Health monitoring and management and control method, device and equipment for smoking hypertension patient, medium and program product

Abstract

The invention provides a smoking hypertension patient health monitoring and control method, device and equipment, a medium and a program product, and relates to the technical field of health data processing, the method is applied to an intelligent mobile terminal, and the method comprises the following steps: converting user complaint information into structured data including life time, symptoms and mood feeling; based on the structured data and physiological data including blood pressure data collected by the wearable device, integrated health data displayed by a time axis is generated; determining smoking habit data based on the time and the place corresponding to the structured data; inputting the smoking habit data and the integrated health data into a natural language generation model to generate a chief complaint record; and performing classification and grading health management strategies through an AI algorithm based on the physiological data. Smoking behavior analysis and hypertension health monitoring are fused, personalized and dynamic health management of smoking hypertension patients is achieved, the accuracy of health state monitoring results of the patients is improved, and effective medical assistance is achieved for chronic diseases.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of health data processing, and particularly relates to a smoking hypertension patient health monitoring and management method, device, equipment, medium and program product. BACKGROUND

[0002] Although there are blood pressure monitoring applications at present, which have basic monitoring and recording functions, when serving smoking hypertension patients, they lack targeted design. Specifically, the absence of patient complaint recording function causes complaint information to be difficult to record and integrate in a timely, accurate and comprehensive manner, and the health status of the patient cannot be monitored according to the actual life events of the patient, resulting in inaccurate health status monitoring results of smoking hypertension patients. SUMMARY

[0003] The present application provides a smoking hypertension patient health monitoring and management method, device, equipment, medium and program product, which solves the defect that the health status of the patient cannot be monitored according to the actual life events of the patient in the prior art, and realizes monitoring the health status of the patient according to the actual life events, thereby improving the accuracy of the health status monitoring results of smoking hypertension patients. The present application provides a smoking hypertension patient health monitoring and management method, which is applied to an intelligent mobile terminal and includes the following steps: receiving user complaint information, the user complaint information being voice information or text information, converting the user complaint information into structured data, the structured data including at least one preset type of information extracted from the user complaint information, the preset type including life events, symptoms and mood; generating integrated health data displayed in a time axis based on the structured data and physiological data, the physiological data being collected by a wearable device, the physiological data at least including blood pressure data; determining smoking habit data based on the time and location corresponding to the structured data, the smoking habit data including the time and location of the occurrence of a smoking event; inputting the smoking habit data and the integrated health data into a language generation model to obtain a complaint record output by the language generation model; generating a hierarchical health management strategy based on the smoking habit data and the integrated health data.

[0004] According to the smoking hypertension patient health monitoring and management method provided by the present application, after the smoking habit data is determined, the method further includes: performing feature extraction on the preprocessed complaint record, physiological data, integrated health data and weather information to obtain multi-dimensional features associated with blood pressure; input the multi-dimensional features into a trained regularity reasoning model, and output an influence law of multiple factors on blood pressure based on the regularity reasoning model; generate a targeted blood pressure regularity AI summary report based on the influence law; The regularity reasoning model is trained based on multiple groups of historical data, and each group of historical data includes the chief complaint record, the physiological data, the integrated health data, weather information, and blood pressure data after a historical period.

[0005] According to the smoking habit data and the integrated health data, a hierarchical health management strategy is generated, which includes: statistically analyze the physiological data to determine the fluctuation mode corresponding to the physiological data; When the fluctuation mode corresponds to a first fluctuation mode, a first prompt information is generated based on a preset medication plan, and the first prompt information is used to remind regular medication; When the fluctuation mode corresponds to a second fluctuation mode or a third fluctuation mode, an individualized intervention strategy is generated based on the smoking habit data, the integrated health data, and real-time physiological data; Based on the smoking habit data, the integrated health data, and real-time physiological data, an individualized intervention strategy is generated, which includes: Based on the smoking habit data, the integrated health data, and real-time physiological data, input data is extracted, including physiological indicator channels, behavior data channels, and environmental data channels; input the input data into a cause analysis model to obtain the fluctuation reason corresponding to the physiological data; When the fluctuation mode corresponds to the second fluctuation mode, a lifestyle suggestion is generated based on the fluctuation reason, and a second prompt information is generated, which is a reminder to take medicine and the lifestyle suggestion; When the fluctuation mode corresponds to the third fluctuation mode, medical guidance information is generated based on the fluctuation reason and the current geographic location, and a third prompt information is generated and sent to a first preset terminal, which corresponds to an emergency contact person. The third prompt information includes real-time physiological data, fluctuation reasons, smoking habit data, integrated health data, and medical guidance information.

[0006] According to the smoking habit data and the integrated health data, a hierarchical health management strategy is generated, which includes: training a prediction model based on the structured data and the blood pressure data of a plurality of historical time points; inputting the structured data of a plurality of time points before the current time point into the trained prediction model to obtain a blood pressure data prediction result output by the prediction model; wherein the prediction model is an LSTM time series model; After the user complaint information is converted into structured data, the method further comprises: generating an intervention scheme based on the structured data, the physiological data and / or the blood pressure data prediction result; determining the execution of the intervention scheme based on the structured data and the physiological data after the intervention scheme is generated; adjusting the intervention scheme based on the execution.

[0007] The method for monitoring and controlling the health of a smoking hypertensive patient provided by the application further comprises: encrypting the collected health data and uploading it to the cloud storage, wherein the collected health data at least includes the physiological data, the integrated health data and the smoking habit data; wherein the process of encrypting the collected health data comprises: blocking the collected health data in the to-be-uploaded time period to obtain N data blocks, and sorting the data blocks based on the data content in the data blocks; obtaining N encrypted reference data, which is generated based on the collected health data in the uploaded time period; encrypting the data blocks based on the encrypted reference data in sequence according to the order of the data blocks to obtain encrypted data blocks.

[0008] The method for monitoring and controlling the health of a smoking hypertensive patient provided by the application further comprises: determining recommended content in a preset knowledge base based on the structured data and the physiological data; pushing the recommended content to the user terminal.

[0009] The application further provides a device for monitoring and controlling the health of a smoking hypertensive patient, comprising: a complaint recording module configured to receive user complaint information, convert the user complaint information into structured data, and extract at least one type of information from the user complaint information, wherein the user complaint information is voice information or text information, the structured data includes at least one type of information, and the types include life events, symptoms and emotional experiences; an integrated display module configured to generate integrated health data displayed in a time axis based on the structured data and physiological data, the physiological data being collected by a wearable device, the physiological data at least including blood pressure data; a smoking data extraction module configured to determine smoking habit data based on time and location corresponding to the structured data, the smoking habit data including time and location of a smoking event; a health monitoring module configured to input the smoking habit data and the integrated health data into a language generation model to obtain a chief complaint record output by the language generation model; a health management module configured to generate a hierarchical health management strategy based on the smoking habit data and the integrated health data.

[0010] The application also provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the smoking hypertension patient health monitoring and management method according to any one of the above.

[0011] The application also provides a non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program is executable on a processor to implement the smoking hypertension patient health monitoring and management method according to any one of the above.

[0012] The application also provides a computer program product including a computer program, wherein the computer program is executable on a processor to implement the smoking hypertension patient health monitoring and management method according to any one of the above.

[0013] The smoking hypertension patient health monitoring and management method, device, equipment, medium, and program product provided by the application extract information such as life events, symptoms, and mood from user chief complaint information, convert the information into structured data, realize timely recording of the chief complaint information of the patient, generate integrated health data displayed in a time axis based on the structured data and physiological data collected by a wearable device, determine smoking habit data based on the structured data, generate a chief complaint record based on the smoking habit data and the integrated health data, and further generate a hierarchical health management strategy based on the smoking habit data and the integrated health data, realize analysis and integration of the chief complaint information of the patient, and monitor the health of the patient in combination with life events including smoking and physiological data collected by the patient, thereby improving the accuracy of the smoking hypertension patient health state monitoring result. BRIEF DESCRIPTION OF DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings described below are only some embodiments of the present application, and other drawings can also be obtained by those skilled in the art without creative effort.

[0015] Figure 1 is a flowchart of the smoking hypertension patient health monitoring and management method provided by the present application.

[0016] Figure 2 is a terminal interaction schematic diagram in the smoking hypertension patient health monitoring and management method provided by the present application.

[0017] Figure 3 is a health data integration schematic diagram in the smoking hypertension patient health monitoring and management method provided by the present application.

[0018] Figure 4 is a process schematic diagram of the generation of chief complaint record in the smoking hypertension patient health monitoring and management method provided by the present application.

[0019] Figure 5 is a structural schematic diagram of the smoking hypertension patient health monitoring and management device provided by the present application.

[0020] Figure 6 is a structural schematic diagram of the electronic device provided by the present application. DETAILED DESCRIPTION

[0021] In order to make the objectives, technical solutions and advantages of the present application clearer, the following will combine the drawings in the present application to clearly and completely describe the technical solutions in the present application. Obviously, the described embodiments are some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present application.

[0022] It should be understood that when used in the specification and the appended claims, the term "comprising" indicates the presence of the described features, integers, steps, operations, elements, and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0023] It should also be understood that the terms used in the present application specification are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the present application specification and the appended claims, the singular forms "a", "an" and "the" are intended to include the plural forms unless the context clearly indicates otherwise.

[0024] It should also be further understood that the term "and / or" as used in this specification and the appended claims refers to any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0025] As used in this specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrases "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0026] The following is combined Figures 1-4 This invention describes a method for health monitoring and management in hypertensive patients who smoke, as provided by the present invention. For example... Figure 1 As shown, the method for health monitoring and management of hypertensive patients who smoke includes the following steps: S110. Receive user complaint information, which is either voice or text information. Convert the user complaint information into structured data. The structured data includes at least one preset type of information extracted from the user complaint information. The preset types include life events, symptoms, and mood. S120. Based on structured data and physiological data, generate integrated health data displayed on a timeline. The physiological data is collected by wearable devices and includes at least blood pressure data. S130. Based on the time and location corresponding to the structured data, determine the smoking habit data, which includes the time and location of the smoking event. S140. Input smoking habit data and integrated health data into the language generation model to obtain the chief complaint record output by the language generation model; S150: Based on smoking habit data and integrated health data, generate a tiered health management strategy.

[0027] The method provided by this invention can be executed by a smart mobile terminal device (such as a smartphone) that has an application (APP) installed on it, also known as an APP client. When the application is run, it executes the method provided by this invention.

[0028] The method provided by this invention extracts information such as life events, symptoms, and mood from user complaints, transforms it into structured data, and enables timely recording of patients' complaints. Furthermore, it generates integrated health data displayed on a timeline based on the structured data and physiological data collected by wearable devices, determines smoking habit data based on the structured data, and generates complaint records based on the smoking habit data and integrated health data. This achieves the analysis and integration of patients' complaints, and, combined with the patient's life events, including smoking, and the collected physiological data, monitors the patient's health, improving the accuracy of health status monitoring results for hypertensive patients who smoke.

[0029] Specifically, such as Figure 2 As shown, in the method provided by the present invention, receiving user complaint information can be achieved by a complaint recording module set in a smart terminal or a wearable device. Since the smart wearable device can be worn by the patient at all times, the wearable device is the main device and the smart terminal is the auxiliary device. The patient can choose freely according to their usage habits and scenarios, making it more convenient for the patient to use.

[0030] In traditional health management scenarios, patients need to recall their symptoms and the course of their illness during consultations, while also providing doctors with as much relevant information about their life events as possible for reference. However, traditional patient complaint records suffer from several drawbacks, including significant patient recall errors, unclear descriptions of symptoms and the course of the illness, lack of correlation with current accurate heart rate, blood pressure, and activity levels, and low efficiency in doctor-patient communication. The method provided by this invention receives real-time user complaint information, transforms it into structured data, and integrates it with physiological data to obtain integrated health data that describes the patient's overall condition. This enables efficient, accurate, and timely recording of complaints, forming standardized and comprehensive complaint records. This allows daily health monitoring to go beyond relying solely on physiological data, incorporating actual life events for more accurate health monitoring results. Furthermore, it provides doctors with more accurate and comprehensive reference information as a basis for diagnosis and treatment during consultations. Through real-time, continuous, and correlated data collection, it breaks the spatial and temporal limitations of traditional medicine, enabling refined and dynamic insights into the patient's condition, greatly improving the accuracy of diagnosis and the individualization of treatment plans.

[0031] Specifically, the method provided by this invention offers patients a convenient function for recording their chief complaints, allowing them to record their physical discomfort symptoms, feelings, and life events through various methods such as text and voice. For example, patients can record the time and severity of headaches and dizziness, as well as their mood changes, diet, smoking habits, and medication usage on that day. Furthermore, the user's chief complaint information can also include image data, such as photos or videos, to more intuitively describe the condition. After receiving the user's chief complaint information, it is transformed into structured information. Specifically, the user's chief complaint information is classified and tagged to obtain information of various preset types, which are then associated and stored with corresponding time and health data to form a complete chief complaint record.

[0032] For example, patients can describe any discomfort they experience during or around the time of blood pressure measurement, including the following information: Symptom types: headache (location, nature), dizziness (is it a spinning sensation? Does it worsen when standing?), chest tightness, palpitations (heartbeat is fast / slow / irregular / leaky), shortness of breath, fatigue, blurred vision, tinnitus, nausea, stiff neck, etc.

[0033] Severity: Mild, Moderate, Severe; Does it affect daily activities?

[0034] Duration: How long the symptoms lasted.

[0035] Mood / feelings: Normal, Average, Bad.

[0036] Triggering or alleviating factors: What you were doing before the symptoms appeared (e.g., just finished exercising, emotional excitement, drinking coffee / alcohol, specific posture), and what you did to relieve the symptoms (e.g., rest, taking medication).

[0037] Other relevant information: whether the medication was taken as prescribed, the time and dosage of medication, whether any doses were missed, and whether any special events occurred (such as staying up late or experiencing high stress).

[0038] The process of structuring received user complaint information can be illustrated as follows: For example, a patient might open the complaint recording function and verbally state, "My headache worsened after smoking this morning." Structuring this information yields the following structured information: { "symptom": "headache" "intensity": 8, "trigger": "after smoking", "duration": "2 hours" "time": "2025-06-05 07:30:00" }

[0039] After obtaining structured data, it is integrated with physiological data. For example... Figure 2 As shown, integrated health data can be displayed through the chief complaint panoramic view module. Figure 3 As shown, the integrated health data view uses a timeline as a guide to display various patient information. This information can be shared with doctors and accessed by support staff, providing a comprehensive and systematic understanding of the patient's disease progression and health status changes, thus aiding in diagnostic decision-making. For example... Figure 3 As shown, through the panoramic view of the chief complaint displayed on the timeline, doctors can clearly see the patient's chief symptoms, smoking behavior, lifestyle habits and other relevant information before and after a certain blood pressure rise, so as to more accurately judge the condition and formulate a treatment plan.

[0040] The chief complaint record and chief complaint panoramic view functions provide more comprehensive and accurate information for communication between patients and doctors, helping doctors to better understand the patient's condition and develop more reasonable treatment plans. At the same time, it also makes it convenient for patients to give timely feedback to doctors about their condition.

[0041] When smoking is present in life events within structured data, smoking habit data can be determined based on the corresponding time and location of the smoking event. This smoking habit data includes the time and location of the smoking event. Smoking habit data is extracted from the structured data, and a chief complaint record is generated based on this data and integrated with health data. The chief complaint record is generated using a natural language generation model, taking the structured chief complaint data as input to produce a doctor-readable report. The natural language generation model can be pre-trained using multiple sets of training data, each set including sample smoking habit data, sample integrated health data, and sample reports.

[0042] like Figure 4 As shown, based on physiological data and integrated health data, in addition to generating chief complaint records, it can also generate targeted AI summary reports on blood pressure patterns for individual patients. Specifically, after determining smoking habit data, it includes: After preprocessing the chief complaint records, physiological data, integrated health data, and weather information, feature extraction was performed to obtain multi-dimensional features related to blood pressure. Multidimensional features are input into a trained pattern inference model, and the influence of multiple factors on blood pressure is output based on the pattern inference model. Generate targeted AI summary reports on blood pressure patterns based on the influencing factors; The regularity reasoning model is trained based on multiple sets of historical data. Each set of historical data includes the chief complaint record, physiological data, integrated health data, weather information, and blood pressure data after the historical period.

[0043] The app utilizes artificial intelligence algorithms to analyze and summarize multi-dimensional data, including received physiological data, patient complaints, weather change data (obtainable via a weather API), historical medication records, and exercise data, to generate a targeted AI summary report. The specific algorithm steps are as follows: Data preprocessing: The collected data of various types are cleaned to remove outliers and missing values. For missing values, interpolation methods (such as linear interpolation, polynomial interpolation, etc.) are used to fill in the missing values. At the same time, the data is normalized to map data of different dimensions to the same numerical range, which facilitates subsequent analysis.

[0044] Feature engineering: Extracting features related to blood pressure changes from preprocessed data. For example, calculating the correlation between daily average blood pressure, blood pressure fluctuation range, heart rate variability, exercise intensity and blood pressure, smoking amount and blood pressure, and emotional state and blood pressure.

[0045] Model building: Machine learning algorithms, such as time series analysis algorithms (e.g., ARIMA, LSTM) or deep learning algorithms (e.g., a model combining convolutional neural networks (CNN) and recurrent neural networks (RNN), are used to construct a blood pressure change prediction model in the form of time series data. The extracted features are used as input to the model, and blood pressure values ​​are used as output. The model is trained using a large amount of historical data, enabling it to learn the complex relationships between different factors and blood pressure changes.

[0046] Pattern Summary: Using a trained model, patient blood pressure data is analyzed to summarize individual patterns of blood pressure changes. For example, the analysis reveals trends in blood pressure under specific weather conditions (e.g., high or low temperatures), specific emotional states (e.g., anxiety or tension), specific medication usage times, or after smoking, as well as the interaction effects of different factors on blood pressure. Ultimately, a personalized pattern report is generated and presented visually to patients and healthcare professionals in the form of charts and text.

[0047] The pattern inference model can be trained using the user's historical data from the previous three months as training data. During the training process, the pattern inference model infers a blood pressure change pattern based on the input historical data at each training session. Based on this blood pressure change pattern, it can generate predicted blood pressure data for subsequent periods. The difference between the predicted blood pressure data and the actual blood pressure data is used to guide the update of the pattern inference model, enabling it to evolve towards outputting more accurate blood pressure change patterns.

[0048] Based on the generated regularity reports and current physiological data and patient complaints, the app develops personalized intervention strategies. For example, if it detects significant blood pressure fluctuations during a specific period and the patient has not taken medication on time, it reminds the patient to take their medication as prescribed and provides medication guidance. If it detects that the patient's anxiety is causing elevated blood pressure, it pushes relevant content such as relaxation training and psychological counseling. The intervention strategies cover multiple aspects, including lifestyle adjustments, medication, and psychological intervention.

[0049] The app sends the generated intervention strategy to the wearable device and displays it on its own interface. The wearable device reminds the patient to implement the intervention measures through vibration, sound, pop-up windows, etc., such as reminding the patient to take medication on time, engage in appropriate exercise, and rest. At the same time, the app further urges the patient to implement the intervention strategy through push notifications and setting reminder alarms, and records the patient's implementation of the intervention measures for subsequent analysis and adjustment of the intervention plan.

[0050] like Figure 2 As shown, the application can include a health indicator monitoring and visualization module. This module receives physiological data collected by the data acquisition module in the smart wearable application. This data can include blood pressure, heart rate, blood oxygen saturation, steps taken, sleep quality, etc. Dynamic views of blood pressure, heart rate, and blood oxygen saturation are displayed in intuitive charts (such as line graphs, bar charts, and pie charts) and a dynamic interface, allowing patients to clearly understand the trends of their various indicators and achieve real-time dynamic monitoring of their health. Smoking habit data, such as smoking time, location, and number of cigarettes, is also presented in a visual manner, helping patients intuitively understand their smoking behavior and its changes. A dynamic dashboard can be integrated to display blood pressure trend heatmaps, smoking behavior calendars, symptom distribution radar charts, etc. Furthermore, a health conversion dashboard can be set up to display positive incentive indicators such as the number of days since quitting smoking and the number of weeks with target blood pressure. By integrating dimensions such as blood pressure stability, smoking frequency, and symptom frequency, a comprehensive health score can also be generated for patient reference.

[0051] Through the visualization of monitoring indicators and data analysis functions, patients and doctors can quickly and accurately understand the patient's health status, promptly identify problems and take measures, thereby improving the efficiency and accuracy of health management.

[0052] Furthermore, based on the collected physiological data, personalized interventions can be implemented for patients, specifically including: Statistical analysis was performed on the physiological data to determine the fluctuation patterns corresponding to the physiological data; When the fluctuation pattern corresponds to the first fluctuation pattern, a first prompt message is generated based on the preset medication plan. The first prompt message is used to remind people to maintain regular medication. When the fluctuation pattern corresponds to the second or third fluctuation pattern, a personalized intervention strategy is generated based on the smoking habit data, the integrated health data, and the real-time physiological data. Based on the smoking habit data, the integrated health data, and the real-time physiological data, a personalized intervention strategy is generated, including: Input data is extracted based on the smoking habit data, the integrated health data, and the real-time physiological data. The input data includes data from the physiological indicator channel, the behavioral data channel, and the environmental data channel. The input data is fed into the cause analysis model to obtain the cause of fluctuations in the physiological data; When the fluctuation pattern corresponds to the second fluctuation pattern, a lifestyle suggestion is generated based on the cause of the fluctuation, and a second reminder message is generated, which is a reminder to take medication and the lifestyle suggestion. When the fluctuation pattern corresponds to the third fluctuation pattern, medical guidance information is generated based on the cause of the fluctuation and the current geographical location, and a third prompt message is generated and sent to the first preset terminal. The first preset terminal corresponds to the emergency contact. The third prompt message includes the real-time physiological data, the cause of the fluctuation, the smoking habit data, the integrated health data, and the medical guidance information.

[0053] The received physiological data can be analyzed in real time using a pre-set algorithm. Fluctuations in physiological data are categorized into normal and abnormal fluctuations. When the fluctuation is normal (threshold: systolic blood pressure ≤ 140 mmHg or diastolic blood pressure ≤ 90 mmHg), i.e., the first fluctuation mode, a reminder to maintain regular medication is generated, such as affirming the patient's treatment effect and encouraging continued regular medication. When the fluctuation is abnormal, it is divided into two modes: the second fluctuation mode (corresponding to mild abnormalities, such as systolic blood pressure 140-159 mmHg or diastolic blood pressure 90-100 mmHg) and the third fluctuation mode (corresponding to moderate to severe abnormalities, such as systolic blood pressure ≥ 160 mmHg or diastolic blood pressure ≥ 100 mmHg).

[0054] When abnormal fluctuations occur, i.e., the second or third fluctuation pattern, smoking habit data, integrated health data, and real-time physiological data are input into the cause analysis model to obtain the corresponding cause of the physiological data fluctuations. Specifically, based on smoking habit data, integrated health data, and real-time physiological data, three channels of input data can be extracted: Physiological indicator channels: time series of systolic blood pressure, diastolic blood pressure, heart rate, and blood oxygen saturation (sampling frequency: 1 time / minute); Behavioral data channels: exercise intensity, exercise duration, medication records (binary: whether medication was taken), smoking frequency; Environmental data channels: emotional state (obtained through voice / facial recognition) and daily routine (obtained through activity monitoring).

[0055] For three-channel input data, features are extracted separately through the feature extraction layer of the cause analysis model: Physiological indicator channel: Blood pressure change trend features and heart rate variability features were extracted using a bidirectional LSTM network; Behavioral data channel: CNN network is used to extract motion pattern features and medication adherence features; Environmental data channel: Use attention mechanisms to extract features of drastic emotional changes and abnormal sleep patterns.

[0056] Next, the features of the three channels are input into the fusion layer of the cause analysis model. The fusion layer adopts a gated fusion mechanism, which dynamically allocates weights according to the importance of data from different channels. The data processing of the fusion layer can be expressed by the formula: F=σ(W1*P+b1)⊙P+σ(W2*B+b2)⊙B+σ(W3*E+b3)⊙E, where F is the fusion feature, P is the physiological indicator feature, B is the behavioral data feature, E is the environmental data feature, σ is the activation function, W is the weight matrix, b is the bias vector, and ⊙ is the element-wise multiplication.

[0057] The fused features are then input into the output layer of the causal analysis model. This output layer includes a causal inference layer, where a Bayesian causal network is constructed to define the causal relationships of various factors on blood pressure. Key causal nodes include smoking behavior, medication use, exercise intensity, emotional state, and sleep patterns. A causal priority ranking mechanism is designed, and the contribution of each factor is calculated based on the strength of evidence. The contribution calculation formula is: C = w1*S + w2*M + w3*E + w4*R + w5*T, where C is the contribution, S is the smoking factor, M is the medication factor, E is the emotional factor, R is the sleep pattern factor, T is the time factor, and w1-w5 are weight coefficients. The output layer also includes an interpretability layer, which uses an attention mechanism to visualize the degree of influence of each factor on blood pressure abnormalities and generates a natural language mechanism to explain the main causes and influencing mechanisms of blood pressure abnormalities. For example, medication factors: due to multiple missed doses; sleep pattern factors: prolonged gaming; emotional factors: intense arguments, etc.

[0058] The causal analysis model can be trained based on multiple sets of training data. Each set of training data includes sample input data and the corresponding causal labels for the sample input data.

[0059] Based on the analysis of physiological data, personalized intervention strategies are generated. For normal fluctuations (the first fluctuation mode), a reminder to maintain regular medication use can be generated. For mild abnormal fluctuations (the second fluctuation mode), an intervention strategy including medication reminders and lifestyle recommendations (such as smoking cessation guidance and exercise adjustments) can be generated. For moderate to severe fluctuations (the third fluctuation mode), an emergency contact intervention strategy can be generated: notifying emergency contacts through multiple channels such as telephone, voice, and SMS. In the third fluctuation mode, the analysis results of the fluctuation causes and the patient's historical data can be sent to emergency contacts for reference. Emergency medical guidance can also be provided to emergency contacts, including first aid measures and information on the nearest hospital to the current location.

[0060] Personalized intervention strategies can be generated through a strategy generation model, which can be trained using reinforcement learning. In the reinforcement learning process, the state space is blood pressure status plus patient characteristics, the action space is a set of intervention strategies, and the reward function is the blood pressure recovery speed and patient satisfaction.

[0061] For example, when blood pressure exceeds the preset normal range (e.g., systolic blood pressure >140 mmHg or diastolic blood pressure >90 mmHg), a comprehensive judgment is made based on the patient's historical blood pressure data, age, gender, and physical condition. If abnormal blood pressure is determined, a notification will be immediately sent to the patient via vibration, sound, and pop-up window. Simultaneously, the app interface will display specific information and suggestions regarding the abnormal blood pressure, such as "Your blood pressure is outside the normal range. Please rest immediately and measure again later." Furthermore, the abnormal blood pressure information can be synchronized to the patient's pre-set emergency contact's mobile phone for timely assistance.

[0062] Furthermore, when a patient's blood pressure is detected to be persistently abnormal (e.g., multiple consecutive measurements exceeding the normal range), heart rate exhibits abnormal fluctuations (e.g., heart rate is too fast or too slow for an extended period), or blood oxygen levels are consistently low (e.g., a significant increase in the time spent with nighttime blood oxygen saturation below 90%), and combined with other physiological data indicating a potential serious threat to health, a medical reminder is automatically generated. A strong medical reminder is sent to the patient, informing them of the need to seek medical attention at a hospital as soon as possible, and providing information about nearby hospitals, including their name, address, distance, and department descriptions. Simultaneously, the system generates a report with a single click based on the patient's recent routine data (e.g., blood pressure, heart rate, blood oxygen, exercise data, chief complaints, etc.), making it convenient for the patient to provide to the doctor during their medical visit.

[0063] In one possible implementation, reminders can also be sent based on a medication schedule. A medication schedule can be generated by accepting patient settings, including drug name, dosage, and other information. Based on the medication schedule and the patient's actual situation, such as whether they have taken their medication on time and their current physical condition, a medication reminder notification can be sent to the patient via a mobile app a certain time in advance (e.g., 15 minutes). If the patient fails to take their medication on time, a reminder will be sent again after a certain period, and the missed medication will be recorded.

[0064] The method provided by this invention can improve the clinical value of hypertension monitoring; it can output different levels of results based on the monitoring results, including providing the attending physician with a visualized medical record of blood pressure and smoking status with one click; issuing warning values ​​to patients and suggesting them seek medical attention for progressive increases or drastic fluctuations in blood pressure over a period of time; issuing critical value reminders to patients and their primary caregivers and suggesting them seek emergency medical attention for high-risk changes in blood pressure and heart rate; and realizing functions such as medication reminders, missed dose recording, and prevention of incorrect doses, thus achieving personalized management of the two chronic diseases of smoking and hypertension.

[0065] Based on the collected structured data and blood pressure data, machine learning algorithms can be used to train and learn the patterns and correlations between life events and blood pressure changes, constructing personalized health prediction models to provide targeted health monitoring for individual patients. Specifically, after converting user complaint information into structured data, this includes: The prediction model is trained based on structured data and blood pressure data from multiple historical moments. The structured data from multiple time points prior to the current moment are input into the trained prediction model to obtain the blood pressure prediction results output by the prediction model.

[0066] The prediction model is an LSTM time series model, which can extract information from time series data to achieve more accurate predictions.

[0067] In other words, the method provided by this invention continuously learns from the patient's historical health data, behavioral patterns, and treatment feedback information. As data accumulates, the model's predictive accuracy continuously improves. Through AI learning, it can predict the patient's blood pressure trends over a future period and the potential health risks caused by smoking, providing a more scientific basis for personalized intervention. For example, it can predict the extent to which a patient's blood pressure may rise after several days of continuous smoking, allowing for early intervention. It can also predict personalized blood pressure trends based on individual patient data using machine learning, identify high-risk periods, and discover possible causes of abnormal blood pressure. For smoking and blood pressure data within the individual patient data, a "smoking-blood pressure" correlation analysis algorithm can be trained. This algorithm, by inputting smoking data, outputs predicted blood pressure changes, quantifying the long-term and short-term effects of smoking behavior on blood pressure fluctuations and discovering the correlation between blood pressure and smoking.

[0068] For individual patient data, AI algorithms can be used to further refine blood pressure fluctuation prevention strategies. This involves training a blood pressure fluctuation prevention strategy generation model, which takes integrated health data as input and outputs a blood pressure fluctuation prevention strategy. The model can be trained using reinforcement learning. During training, based on the patient's current integrated health data, the model generates a sample blood pressure fluctuation prevention strategy. A reward value is generated based on the effectiveness of this sample strategy, and the model is updated based on this reward value. This allows the model to output appropriate sample blood pressure fluctuation prevention strategies based on the input integrated health data.

[0069] Furthermore, after converting user complaint information into structured data, it also includes: Intervention plans are generated based on prediction results from structured data, physiological data, and / or blood pressure data. Based on the structured data and physiological data generated after the intervention plan was generated, the implementation status of the intervention plan was determined; The intervention plan will be adjusted based on the implementation results.

[0070] Based on the blood pressure prediction results output by the predictive model and / or the patient's daily health data, a personalized intervention plan can be generated for the patient. The intervention plan includes smoking cessation guidance, blood pressure fluctuation prevention recommendations, and lifestyle modification plans.

[0071] For smoking cessation interventions, personalized smoking cessation plans can be developed based on the patient's smoking habits and willingness to quit, providing methods for quitting and psychological support. Regarding the prevention of blood pressure fluctuations, reasonable medication recommendations, exercise plans, and dietary schemes can be given based on the patient's blood pressure data and physical condition. For example, knowledge of nicotine replacement therapy (NRT) can be dynamically adjusted based on patient adherence, personalized exercise / dietary plans (such as high-fiber diets) can be recommended for lifestyle regulation, and mindfulness meditation courses can be provided when stress-induced smoking is detected.

[0072] After the intervention plan is generated, patients can be urged to implement the intervention plan through push notifications, message reminders, etc., and the implementation status of patients can be evaluated and adjusted regularly.

[0073] AI learning and personalized intervention capabilities can provide accurate health predictions and personalized intervention plans based on the individual patient's situation, meeting the patient's individual needs and improving treatment outcomes.

[0074] The AI ​​learning and training process can be executed in the cloud, which reduces the memory and computing resource requirements of the terminal executing the method provided in this invention, ensuring the smooth operation of the terminal and the effectiveness of AI algorithm training.

[0075] For data collected locally on a smart mobile terminal, long-term storage on the local device would result in significant consumption of storage resources. Furthermore, since the training process for AI algorithms takes place in the cloud, the method provided in this invention requires uploading the health data collected by the smart mobile terminal to the cloud. To ensure data security and prevent leakage of collected health data, one possible implementation of the method provided in this invention involves encrypting the collected health data before uploading it to the cloud. Specifically, this includes: The collected health data is encrypted and uploaded to the cloud for storage. The collected health data includes at least physiological data, integrated health data, and smoking habit data. The process of encrypting the collected health data includes: The collected health data within the upload period is randomly divided into N data blocks, and the data blocks are sorted based on the data content in the data blocks. Obtain N encrypted reference data, which are generated based on the health data collected within the uploaded time period; The data blocks are encrypted sequentially based on the encrypted reference data to obtain encrypted data blocks.

[0076] Specifically, in one possible implementation of the method provided by this invention, health data is uploaded in time slots. That is, at regular intervals, the health data collected within that time slot is uploaded to the cloud. For the health data within the time slot to be uploaded, it is first divided into N data blocks. These blocks are then sorted based on their data content, for example, according to the amount of data in each block and a pre-agreed rule. Next, encrypted reference data is constructed based on the health data from the previously uploaded time slot closest to the time slot to be uploaded. The encrypted reference data for the first previously uploaded time slot can be constructed based on the patient's basic information. The health data from the previously uploaded time slot is also divided into N blocks and sorted according to a preset rule. This establishes a correspondence between the data blocks of the health data from the previously uploaded time slot and the time slot to be uploaded. For the data blocks of the health data from the time slot to be uploaded, encrypted reference data extracted from the corresponding data blocks in the previously uploaded time slot is used.

[0077] Therefore, if an unauthorized attacker intercepts an encrypted data block, at least three conditions must be met to decrypt it: first, the sorting rules must be known; second, all other health data blocks collected within the same time period as the encrypted data block must be intercepted to determine their order; and third, all previously uploaded health data blocks from the time period preceding the encrypted data block must be obtained. This effectively ensures the security of collected health data.

[0078] The method provided by this invention, in order to improve the effectiveness of health monitoring, can also establish a knowledge base to provide patients with health knowledge. Specifically, it includes: Based on structured data and physiological data, recommended content is determined from a pre-set knowledge base; Recommended content is pushed to the user's device.

[0079] A comprehensive medical knowledge base has been built, covering medical knowledge, health management methods, and disease prevention measures related to hypertension and smoking. The content in the knowledge base is reviewed and updated by a professional medical team to ensure the accuracy and authority of the information. Patients can search for relevant knowledge through methods such as searching and browsing by category. The system will also proactively push relevant knowledge content based on the patient's health condition and needs, helping patients improve their health awareness and self-management abilities.

[0080] The knowledge base provides patients with a wealth of health knowledge, helping them understand disease-related information and health management methods, improving their health awareness and self-management abilities, and promoting the development of good living habits.

[0081] The following describes the health monitoring and management device for smoking-induced hypertension patients provided by the present invention. The health monitoring and management device for smoking-induced hypertension patients described below can be referred to in correspondence with the health monitoring and management method for smoking-induced hypertension patients described above. For example... Figure 5 As shown, the health monitoring and management device for smoking-related hypertension patients provided by this invention includes a chief complaint recording module 510, an integrated display module 520, a smoking data extraction module 530, a health monitoring module 540, and a health management module 550. Wherein: The chief complaint recording module 510 is used to receive user chief complaint information, which can be voice or text information, and to convert the user chief complaint information into structured data. The structured data includes at least one preset type of information extracted from the user chief complaint information. The preset types include life events, symptoms, and mood feelings. The integrated display module 520 is used to generate integrated health data displayed on a timeline based on structured data and physiological data. The physiological data is collected by wearable devices and includes at least blood pressure data. The smoking data extraction module 530 is used to determine smoking habit data based on the time and location corresponding to the structured data. The smoking habit data includes the time and location of the smoking event. The health monitoring module 540 is used to input smoking habit data and integrated health data into the language generation model to obtain the chief complaint record output by the language generation model; The health management module 550 is used to generate tiered health management strategies based on smoking habit data and integrated health data.

[0082] Figure 6 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 6 As shown, the electronic device may include a processor 610, a communications interface 620, a memory 630, and a communication bus 640. The processor 610, communications interface 620, and memory 630 communicate with each other via the communication bus 640. The processor 610 can call logical instructions stored in the memory 630 to execute a health monitoring and management method for patients with hypertension who smoke. The method for health monitoring and management of hypertensive patients who smoke includes: receiving user complaint information, which can be in the form of voice or text; converting the user complaint information into structured data, which includes at least one preset type of information extracted from the user complaint information, including life events, symptoms, and mood; generating integrated health data displayed on a timeline based on the structured data and physiological data, which is collected by wearable devices and includes at least blood pressure data; determining smoking habit data based on the time and location corresponding to the structured data, which includes the time and location of smoking events; inputting the smoking habit data and integrated health data into a language generation model to obtain the complaint record output by the language generation model; and generating a tiered health management strategy based on the smoking habit data and integrated health data.

[0083] Furthermore, the logical instructions in the aforementioned memory 630 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0084] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the health monitoring and management method for smoking-induced hypertension patients provided by the above methods. This health monitoring and management method for smoking-induced hypertension patients includes: receiving user complaint information, which may be voice or text information; converting the user complaint information into structured data, the structured data including at least one preset type of information extracted from the user complaint information, the preset types including life events, symptoms, and mood feelings; generating integrated health data displayed on a timeline based on the structured data and physiological data, the physiological data being collected by a wearable device and including at least blood pressure data; determining smoking habit data based on the time and location corresponding to the structured data, the smoking habit data including the time and location of smoking events; inputting the smoking habit data and integrated health data into a language generation model to obtain the complaint record output by the language generation model; and generating a tiered health management strategy based on the smoking habit data and integrated health data.

[0085] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the health monitoring and management method for smoker-induced hypertension patients provided by the above methods. This health monitoring and management method for smoker-induced hypertension patients includes: receiving user complaint information, which may be voice or text information; converting the user complaint information into structured data, the structured data including at least one preset type of information extracted from the user complaint information, the preset types including life events, symptoms, and mood feelings; generating integrated health data displayed on a timeline based on the structured data and physiological data, the physiological data being collected by a wearable device and including at least blood pressure data; determining smoking habit data based on the time and location corresponding to the structured data, the smoking habit data including the time and location of smoking events; inputting the smoking habit data and integrated health data into a language generation model to obtain the complaint record output by the language generation model; and generating a tiered health management strategy based on the smoking habit data and integrated health data.

[0086] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0087] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0088] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for health monitoring and management of hypertensive patients who smoke, characterized in that, Applied to smart mobile terminals, the method includes: Receive user complaint information, which may be voice or text information, and convert the user complaint information into structured data. The structured data includes at least one preset type of information extracted from the user complaint information. The preset types include life events, symptoms, and mood feelings. Based on the structured data and physiological data, integrated health data displayed on a timeline is generated. The physiological data is collected by wearable devices and includes at least blood pressure data. Based on the time and location corresponding to the structured data, smoking habit data is determined, including the time and location of the smoking event. The smoking habit data and the integrated health data are input into the language generation model to obtain the chief complaint record output by the language generation model; Based on the smoking habit data and the integrated health data, a tiered health management strategy is generated.

2. The method for health monitoring and management of hypertensive patients who smoke, as described in claim 1, is characterized in that... After determining the smoking habit data, the method further includes: After preprocessing the chief complaint record, the physiological data, the integrated health data, and the weather information, feature extraction is performed to obtain multi-dimensional features related to blood pressure. The multi-dimensional features are input into a trained pattern reasoning model, and the pattern of influence of multiple factors on blood pressure is output based on the pattern reasoning model. Based on the aforementioned patterns of influence, a targeted AI summary report on blood pressure patterns is generated through AI learning. The pattern reasoning model is trained based on multiple sets of historical data. Each set of historical data includes the chief complaint record, physiological data, integrated health data, weather information, and blood pressure data after the historical period.

3. The method for health monitoring and management of hypertensive patients who smoke, as described in claim 2, is characterized in that... The generation of a tiered health management strategy based on the smoking habit data and the integrated health data includes: Statistical analysis was performed on the physiological data to determine the fluctuation patterns corresponding to the physiological data; When the fluctuation pattern corresponds to the first fluctuation pattern, a first prompt message is generated based on the preset medication plan. The first prompt message is used to remind people to maintain regular medication. When the fluctuation pattern corresponds to the second or third fluctuation pattern, a personalized intervention strategy is generated based on the smoking habit data, the integrated health data, and the real-time physiological data. Based on the smoking habit data, the integrated health data, and the real-time physiological data, a personalized intervention strategy is generated, including: Input data is extracted based on the smoking habit data, the integrated health data, and the real-time physiological data. The input data includes data from the physiological indicator channel, the behavioral data channel, and the environmental data channel. The input data is fed into the cause analysis model to obtain the cause of fluctuations in the physiological data; When the fluctuation pattern corresponds to the second fluctuation pattern, a lifestyle suggestion is generated based on the cause of the fluctuation, and a second reminder message is generated, which is a reminder to take medication and the lifestyle suggestion. When the fluctuation pattern corresponds to the third fluctuation pattern, medical guidance information is generated based on the cause of the fluctuation and the current geographical location, and a third prompt message is generated and sent to the first preset terminal. The first preset terminal corresponds to the emergency contact. The third prompt message includes the real-time physiological data, the cause of the fluctuation, the smoking habit data, the integrated health data, and the medical guidance information.

4. The method for health monitoring and management of hypertensive patients who smoke, as described in claim 1, is characterized in that... The method further includes: The collected health data is encrypted and uploaded to the cloud for storage. The collected health data includes at least the physiological data, the integrated health data, and the smoking habit data. The process of encrypting the collected health data includes: The collected health data within the upload period is divided into blocks to obtain N data blocks, and the data blocks are sorted based on the data content in the data blocks; Obtain N encrypted reference data, which are generated based on the collected health data within the uploaded time period; The data blocks are encrypted sequentially based on the encrypted reference data in the order of the data blocks to obtain encrypted data blocks.

5. The method for health monitoring and management of hypertensive patients who smoke, as described in claim 1, is characterized in that... After converting the user's complaint information into structured data, the method further includes: The prediction model is trained based on the structured data and blood pressure data from multiple historical moments. The structured data from multiple time points prior to the current time are input into the trained prediction model to obtain the blood pressure data prediction result output by the prediction model. The prediction model is an LSTM time series model. After converting the user's complaint information into structured data, the method further includes: An intervention plan is generated based on the predicted results of the structured data, the physiological data, and / or the blood pressure data. Based on the structured data and physiological data generated after the intervention plan is generated, the execution status of the intervention plan is determined; The intervention plan is adjusted based on the implementation results.

6. The method for health monitoring and management of hypertensive patients who smoke, as described in claim 1, is characterized in that... The method further includes: Based on the structured data and the physiological data, recommended content is determined from a preset knowledge base; The recommended content is pushed to the user's terminal.

7. A health monitoring and management device for smokers with hypertension, characterized in that, include: The chief complaint recording module is used to receive user chief complaint information, which is voice information or text information, and convert the user chief complaint information into structured data. The structured data includes at least one preset type of information extracted from the user chief complaint information. The preset types include life events, symptoms, and mood feelings. An integrated display module is used to generate integrated health data displayed on a timeline based on the structured data and physiological data. The physiological data is collected by wearable devices and includes at least blood pressure data. The smoking data extraction module is used to determine smoking habit data based on the time and location corresponding to the structured data, wherein the smoking habit data includes the time and location of the smoking event. The health monitoring module is used to input the smoking habit data and the integrated health data into the language generation model and obtain the chief complaint record output by the language generation model; The health management module is used to generate a tiered health management strategy based on the smoking habit data and the integrated health data.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the health monitoring and management method for smoking hypertensive patients as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the health monitoring and management method for hypertensive patients who smoke, as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the health monitoring and management method for hypertensive patients who smoke, as described in any one of claims 1 to 6.

Images