AI intelligent health monitoring management system based on big data

Abstract

The invention relates to the technical field of health management, and discloses an AI intelligent health monitoring management system based on big data, the system enhances the comprehensiveness and precision of health monitoring by integrating physiological feature data and medical feature data, and the traditional system generally depends on a single data source or detection mode, so that the system is not easy to use. The system can establish a relation between real-time data and historical data so as to provide more comprehensive health risk assessment, specifically, the physiological feature data acquisition module and the medical feature data acquisition module work cooperatively, the comprehensiveness and accuracy of the data are ensured, and the health risk assessment efficiency is improved. The abnormal reference coefficient is calculated through an intelligent algorithm and compared with the set threshold value, the sensitivity and accuracy of health problem early warning are greatly improved, and the system can recognize potential health risks more timely through dynamic data analysis.

Description

Technical Field

[0001] The present invention relates to the field of health management technology, and specifically to an AI intelligent health monitoring and management system based on big data. Background Art

[0002] With the continuous advancement of science and technology, big data and artificial intelligence (AI) are increasingly being used in all walks of life, especially in the field of health management. The development of health management technology involves multiple directions, including the collection, processing, analysis and establishment of early warning systems for medical data. The health monitoring system relies on big data technology, combined with wearable devices, physiological data collection and analysis algorithms, to provide users with accurate health warnings and interventions. Specifically, the AI ​​intelligent health monitoring and management system based on big data can provide users with personalized health guidance and emergency warnings through AI intelligent analysis based on real-time monitoring of key physiological indicators such as cardiovascular, respiratory, and sleep health, ensuring that health problems can be discovered and dealt with in the first place.

[0003] At present, most traditional health monitoring systems rely mainly on a single data source, such as using only real-time physiological data or a separate medical history record for analysis. This single data processing method makes the health assessment of the existing system relatively limited and unable to fully capture the health status of individuals. In addition, traditional health monitoring systems often lack timely and intelligent early warning mechanisms. Most of them only provide basic health data reports without in-depth health risk analysis and abnormal early warning. Furthermore, there is still a lot of room for improvement in the accuracy of health monitoring equipment and the real-time monitoring. Many systems fail to combine the comparison of real-time data with historical data, and lack sufficient sensitivity and dynamic risk identification capabilities.

[0004] Therefore, we proposed an AI intelligent health monitoring and management system based on big data to solve the above problems. Summary of the invention

[0005] The purpose of the present invention is to provide an AI intelligent health monitoring and management system based on big data to solve the problem that most traditional health monitoring systems proposed in the above background technology mainly rely on a single data source, such as only using real-time physiological data or separate medical history records for analysis. This single data processing method makes the health assessment of the existing system relatively limited and unable to fully capture the health status of individuals. In addition, traditional health monitoring systems often lack timeliness and intelligent early warning mechanisms. Most of them only provide basic health data reports without in-depth health risk analysis and abnormal early warning. Furthermore, there is still a lot of room for improvement in the accuracy of health monitoring equipment and the real-time monitoring. Many systems fail to combine the comparison of real-time data with historical data, and lack sufficient sensitivity and dynamic risk identification capabilities.

[0006] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: an AI intelligent health monitoring and management system based on big data, comprising a physiological characteristic data acquisition module, a medical characteristic data acquisition module, a cardiovascular health analysis module, a respiratory system analysis module, a sleep health analysis module and an intelligent alarm module;

[0007] The physiological characteristic data acquisition module is used to acquire the user's real-time cardiovascular health data, real-time respiratory system data, and real-time sleep health data, and pre-process the acquired data;

[0008] The medical characteristic data acquisition module is used to acquire the user's online medical data, thereby extracting the patient's medical cardiovascular health data, medical respiratory system data and medical sleep health data, and preprocessing the extracted data;

[0009] The cardiovascular health analysis module is used to integrate and calculate the user's real-time cardiovascular health data and the obtained cardiovascular health data from the medical consultation, so as to obtain the user's cardiovascular health data abnormal reference coefficient, and compare the calculated cardiovascular health data abnormal reference coefficient with a preset cardiovascular health data abnormal threshold, so as to generate a first comparison result, and determine whether the user currently has a cardiovascular abnormality according to the first comparison result;

[0010] The respiratory system analysis module is used to integrate and calculate the user's real-time respiratory system data and the acquired consultation respiratory system data, so as to obtain the user's respiratory system data abnormality reference coefficient, and compare the respiratory system data abnormality reference coefficient with a preset respiratory system data abnormality threshold, so as to generate a second comparison result, and determine whether the user has a respiratory system abnormality according to the second comparison result;

[0011] The sleep health analysis module is used to integrate and calculate the user's real-time sleep health data and the doctor's sleep health data, so as to obtain the user's sleep health data abnormal reference coefficient, and compare the sleep health data abnormal reference coefficient with a preset sleep health data abnormal threshold, so as to generate a third comparison result, and judge whether the user has a sleep health abnormality problem according to the third comparison result;

[0012] The intelligent alarm module is used to perform intelligent alarm according to the first comparison result, the second comparison result and the third comparison result.

[0013] Preferably, the physiological characteristic data acquisition module includes a cardiovascular data acquisition unit, a respiratory system data acquisition unit and a sleep health data acquisition unit;

[0014] The cardiovascular data acquisition unit is used to obtain the user's real-time cardiovascular data including real-time heart rate, real-time blood oxygen saturation and real-time heart rate variability;

[0015] The real-time heart rates are recorded as A1, A2, A3, ..., An according to the timestamps;

[0016] The real-time blood oxygen saturation is recorded as B1, B2, B3, ..., Bn according to the timestamp;

[0017] The real-time heart rate variability is recorded as C1, C2, C3, ..., Cn according to the timestamp;

[0018] The respiratory system data acquisition unit is used to obtain the user's real-time respiratory system data including real-time respiratory frequency, real-time respiratory depth and real-time respiratory efficiency;

[0019] The real-time respiratory rates are recorded as D1, D2, D3, ..., Dn according to the timestamps;

[0020] The real-time breathing depth is recorded as E1, E2, E3, ..., En according to the timestamp;

[0021] The real-time breathing efficiency is recorded as F1, F2, F3, ..., Fn according to the timestamp;

[0022] The sleep health data acquisition unit is used to obtain the user's real-time sleep health data including real-time deep sleep ratio, real-time sleep heart rate and real-time sleep stability;

[0023] The real-time deep sleep ratio is recorded as G1, G2, G3, ..., Gn according to the timestamp;

[0024] The real-time sleeping heart rate is recorded as H1, H2, H3, ..., Hn according to the timestamp;

[0025] The real-time sleep stabilities are recorded as I1, I2, I3, ..., In according to the timestamps.

[0026] Preferably, the medical characteristic data collection module includes a cardiovascular report analysis unit, a respiratory system report analysis unit and a sleep report analysis unit;

[0027] The cardiovascular report analysis unit is used to obtain the user's cardiovascular health data including standard cardiovascular data including standard heart rate ZA, standard blood oxygen saturation ZB and standard heart rate variability ZC;

[0028] The respiratory system report analysis unit is used to obtain the user's respiratory system data including standard respiratory frequency ZD, standard respiratory depth ZE and standard respiratory efficiency ZF;

[0029] The sleep report analysis unit is used to obtain the user's medical sleep health data, including the standard deep sleep ratio ZG, the standard sleep heart rate ZH, and the standard sleep stability ZI.

[0030] Preferably, the cardiovascular health analysis module includes a cardiovascular data calculation unit and a cardiovascular data comparison unit;

[0031] The cardiovascular data calculation unit is used to integrate and calculate the user's real-time cardiovascular health data and the obtained medical cardiovascular health data, so as to obtain the abnormal reference coefficient of the user's cardiovascular health data, including the first abnormal reference coefficient XYZ1 of cardiovascular health data and the second abnormal reference coefficient XYZ2 of cardiovascular health data. When any parameter in the user's real-time cardiovascular health data is greater than 40% of any parameter in the medical cardiovascular health data or any parameter in the user's real-time cardiovascular health data is less than 40% of any parameter in the medical cardiovascular health data, the AI automatically enables the second abnormal reference coefficient XYZ2 of cardiovascular health data. If the condition is not met, the AI defaults to using the first abnormal reference coefficient XYZ1 of cardiovascular health data;

[0032] Among them, the first abnormal reference coefficient XYZ1 of cardiovascular health data is obtained by first integrating and calculating the window mean value of the user's real-time cardiovascular health data, and then integrating and calculating the window mean value with the medical cardiovascular health data;

[0033] The second abnormal reference coefficient XYZ2 of cardiovascular health data is obtained by integrating and calculating the user's real-time cardiovascular health data and the obtained medical cardiovascular health data;

[0034] The cardiovascular data comparison unit is used to compare the first abnormal reference coefficient XYZ1 of cardiovascular health data and the second abnormal reference coefficient XYZ2 of cardiovascular health data with the preset abnormal threshold of cardiovascular health data respectively, so as to generate the first comparison result. The specific method is as follows:

[0035] When XYZ1≥Y1, it means that the current user does not have abnormal cardiovascular health data;

[0036] When XYZ1<Y1, it means that the current user has abnormal cardiovascular health data;

[0037] When XYZ2>Y1×120%, it means that the current user does not have abnormal cardiovascular health data;

[0038] When XYZ2≤Y2×120%, it means that the current user has abnormal cardiovascular health data.

[0039] Preferably, the first cardiovascular health data abnormal reference coefficient XYZ1 and the second cardiovascular health data abnormal reference coefficient XYZ2 are respectively calculated and obtained by the following formulas:

[0040]

[0041] Where: ZA is the standard heart rate, ZB is the standard blood oxygen saturation, ZC is the standard heart rate variability, PA is the user's heart rate window mean, PB is the user's blood oxygen saturation window mean, PC is the user's heart rate variability window mean, An is the user's real-time heart rate at the nth timestamp, Bn is the user's real-time blood oxygen saturation at the nth timestamp, and Cn is the user's real-time heart rate variability at the nth timestamp;

[0042] n is the total number of timestamps, m is the window size, k is the kth data point in the window, a1, a2 and a3 are weight values, and the values ​​of a1, a2 and a3 are adjusted and set by the user.

[0043] Preferably, the respiratory system analysis module includes a respiratory system data calculation unit and a respiratory system data comparison unit;

[0044] The respiratory system data calculation unit is used to integrate and calculate the user's real-time respiratory system data and the acquired visit respiratory system data, so as to obtain the user's respiratory system data abnormal reference coefficient, including a first respiratory system data abnormal reference coefficient HYZ1 and a second respiratory system data abnormal reference coefficient HYZ2. When any data under the user's real-time respiratory system data is greater than 40% of any data under the visit respiratory system data or any data under the user's real-time respiratory system data is less than 40% of any data under the visit respiratory system data, the AI ​​automatically enables the second respiratory system data abnormal reference coefficient HYZ2. If the condition is not met, the AI ​​defaults to using the first respiratory system data abnormal reference coefficient HYZ1.

[0045] The first respiratory system data abnormal reference coefficient HYZ1 is obtained by first integrating and calculating the user's real-time respiratory system data to obtain the window mean, and then integrating and calculating the window mean with the patient's respiratory system data;

[0046] The second respiratory system data abnormal reference coefficient HYZ2 is obtained by integrating and calculating the user's real-time respiratory system data and the obtained patient respiratory system data;

[0047] The respiratory system data comparison unit is used to compare the first respiratory system data abnormality reference coefficient HYZ1 and the second respiratory system data abnormality reference coefficient HYZ2 with the preset respiratory system data abnormality threshold value, so as to generate a second comparison result, specifically in the following manner:

[0048] When HYZ1 ≥ Y2, it indicates that there is no abnormal respiratory system data for the current user;

[0049] When HYZ1 < Y2, it indicates that there is abnormal respiratory system data for the current user;

[0050] When HYZ2 > Y2 × 120%, it indicates that there is no abnormal respiratory system data for the current user;

[0051] When HYZ2 ≤ Y2 × 120%, it indicates that there is abnormal respiratory system data for the current user.

[0052] Preferably, the first abnormal respiratory system data reference coefficient HYZ1 and the second abnormal respiratory system data reference coefficient HYZ2 are respectively obtained by calculating through the following formulas:

[0053]

[0054] In the formula: ZD is the standard breathing frequency, ZE is the standard breathing depth, ZF is the standard breathing efficiency, PD is the average value of the breathing frequency window of the user, PE is the average value of the breathing depth window of the user, PF is the average value of the breathing efficiency window of the user, Dn is the real-time breathing frequency of the user at the nth timestamp, En is the real-time breathing depth of the user at the nth timestamp, Fn is the real-time breathing efficiency of the user at the nth timestamp;

[0055] n is the total number of timestamps, m is the window size, k is the kth data point within the window, b1, b2, and b3 are weight values, and the values of b1, b2, and b3 are adjusted and set by the user.

[0056] Preferably, the sleep health analysis module includes a sleep health data calculation unit and a sleep health data comparison unit;

[0057] The sleep health data calculation unit is used to integrate and calculate the real-time sleep health data of the user and the obtained medical sleep health data, so as to obtain the sleep health data abnormal reference coefficient of the user, including the first sleep health data abnormal reference coefficient SYZ1 and the second sleep health data abnormal reference coefficient SYZ2. When any parameter in the real-time sleep health data of the user is greater than 40% of any parameter in the medical sleep health data of the user or any parameter in the real-time sleep health data of the user is less than 40% of any parameter in the medical sleep health data of the user, the AI automatically enables the second sleep health data abnormal reference coefficient SYZ2. If the condition is not met, the AI defaults to using the first sleep health data abnormal reference coefficient SYZ1;

[0058] Among them, the first sleep health data anomaly reference coefficient SYZ1 is obtained by first integrating and calculating the window mean of the user's real-time sleep health data, and then integrating and calculating the window mean with the diagnosed sleep health data;

[0059] The second sleep health data anomaly reference coefficient SYZ2 is obtained by integrating and calculating the user's real-time sleep health data and the diagnosed sleep health data obtained;

[0060] The sleep health data comparison unit is used to compare the first sleep health data anomaly reference coefficient SYZ1 and the second sleep health data anomaly reference coefficient SYZ2 with the preset sleep health data anomaly threshold respectively, so as to generate a third comparison result. The specific method is as follows:

[0061] When SYZ1≥Y3, it means that the current user does not have abnormal sleep health data;

[0062] When SYZ1<Y3, it means that the current user has abnormal sleep health data;

[0063] When SYZ2>Y3×120%, it means that the current user does not have abnormal sleep health data;

[0064] When SYZ2≤Y3×120%, it means that the current user has abnormal sleep health data.

[0065] Preferably, the first sleep health data anomaly reference coefficient SYZ1 and the second sleep health data anomaly reference coefficient SYZ2 are respectively obtained by the following formulas:

[0066]

[0067] In the formula: ZG is the standard deep sleep ratio, ZH is the standard sleep heart rate, ZI is the standard sleep stability, PG is the window mean of the user's deep sleep ratio, PH is the window mean of the user's sleep heart rate, PI is the window mean of the user's sleep stability, Gn is the user's real-time deep sleep ratio at the nth timestamp, Hn is the user's real-time sleep heart rate at the nth timestamp, In is the user's real-time sleep stability at the nth timestamp;

[0068] n is the total number of timestamps, m is the window size, k is the kth data point within the window, c1, c2, and c3 are weight values, and the values of c1, c2, and c3 are adjusted and set by the user.

[0069] Preferably, the intelligent alarm module intelligently identifies the current time period through AI and conducts intelligent alarm according to the time period. The specific method is as follows:

[0070] When it is not nighttime, AI identifies the first comparison result and the second comparison result. When the first cardiovascular health data abnormal reference coefficient XYZ1 and the first respiratory system data abnormal reference coefficient HYZ1 are enabled, if the first comparison result and the second comparison result are both in data abnormality, AI adjusts the bracelet LED screen color to yellow and gives a sound prompt at a frequency of once every two seconds. If either of the first comparison result and the second comparison result is in data abnormality, AI does not give an alarm. When the second cardiovascular health data abnormal reference coefficient XYZ2 and the second respiratory system data abnormal reference coefficient HYZ2 are enabled, if either is in, AI adjusts the bracelet LED screen color to red, gives a sound prompt twice every second, and sends a message to the emergency contact and makes a call to a smart phone.

[0071] When it is at night, AI identifies the second comparison result and the third comparison result. When the first respiratory system data abnormal reference coefficient HYZ1 and the first sleep health data abnormal reference coefficient SYZ1 are enabled, if the second comparison result and the third comparison result are both in, AI adjusts the bracelet LED screen color to yellow, and makes a sound prompt at a frequency of once every two seconds, and the prompt tone is increased by 30% compared to non-night time. If either the second comparison result or the third comparison result is in data abnormality, AI does not alarm. When the second respiratory system data abnormal reference coefficient HYZ2 and the second sleep health data abnormal reference coefficient SYZ2 are enabled, if either the second comparison result or the third comparison result is in, AI adjusts the bracelet LED screen color to red, and makes a sound prompt twice every second, and the prompt tone is increased by 150% compared to non-night time, and calls the emergency contact on a smart phone.

[0072] Compared with the prior art, the present invention has the following beneficial effects:

[0073] 1. The system enhances the comprehensiveness and accuracy of health monitoring by integrating physiological characteristic data with medical characteristic data. Traditional systems usually rely on a single data source or detection method, while the system can establish a connection between real-time data and historical data to provide a more comprehensive health risk assessment. Specifically, the physiological characteristic data acquisition module and the medical characteristic data acquisition module work together to ensure the comprehensiveness and accuracy of the data. The abnormal reference coefficient is calculated by an intelligent algorithm and compared with the set threshold, which greatly improves the sensitivity and accuracy of health problem warning. Traditional methods often rely on fixed thresholds or simple judgment criteria, which are prone to miss some subtle health abnormalities. The system can identify potential health risks more promptly through dynamic, data-based analysis, especially the comprehensive analysis of the cardiovascular health analysis module, the respiratory system analysis module and the sleep health analysis module, which not only improves the sensitivity to cardiovascular, respiratory and sleep problems, but also provides users with more personalized health assessments. The introduction of the intelligent alarm module realizes the comprehensive monitoring and intelligent alarm of multi-dimensional health problems, which not only enhances the real-time nature of health monitoring, but also ensures that when health abnormalities occur, users can receive warnings and take actions in the first time, thereby reducing the severity of health risks.

[0074] 2. By distinguishing between nighttime and non-nighttime states, the intelligent alarm module can intelligently adapt to the alarm needs of different time periods. The alarm volume is automatically increased at night to ensure that users or emergency contacts can detect it in time in a quiet environment, while avoiding disturbing normal daytime activities due to unnecessary prompt sounds. The intelligent alarm module integrates the first comparison results, the second comparison results, and the third comparison results. Through comprehensive judgment of multi-dimensional data, it reduces the false alarms that may be caused by single data anomalies and improves the accuracy and reliability of anomaly identification. According to the category of abnormal reference coefficients, the intelligent alarm module adopts different alarm responses. When the first reference coefficient is enabled, the alarm is only triggered when all data are abnormal to avoid overreaction to slight data fluctuations. When the second reference coefficient is enabled, any data abnormality triggers a high-intensity alarm. This hierarchical alarm design ensures a rapid response to serious health risks while avoiding unnecessary interference to users. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 It is a system flow chart of the present invention.

[0076] In the figure: 1. Physiological characteristic data acquisition module; 11. Cardiovascular data acquisition unit; 12. Respiratory system data acquisition unit; 13. Sleep health data acquisition unit; 2. Medical characteristic data acquisition module; 21. Cardiovascular report analysis unit; 22. Respiratory system report analysis unit; 23. Sleep report analysis unit; 3. Cardiovascular health analysis module; 31. Cardiovascular data calculation unit; 32. Cardiovascular data comparison unit; 4. Respiratory system analysis module; 41. Respiratory system data calculation unit; 42. Respiratory system data comparison unit; 5. Sleep health analysis module; 51. Sleep health data calculation unit; 52. Sleep health data comparison unit; 6. Intelligent alarm module. DETAILED DESCRIPTION

[0077] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0078] Example 1: Please refer to Figure 1 , an AI intelligent health monitoring and management system based on big data, comprising a physiological characteristic data collection module 1, a medical characteristic data collection module 2, a cardiovascular health analysis module 3, a respiratory system analysis module 4, a sleep health analysis module 5 and an intelligent alarm module 6;

[0079] The physiological characteristic data acquisition module 1 is used to acquire the user's real-time cardiovascular health data, real-time respiratory system data, and real-time sleep health data, and pre-process the acquired data;

[0080] The medical characteristic data acquisition module 2 is used to acquire the user's online medical data, thereby extracting the patient's medical cardiovascular health data, medical respiratory system data and medical sleep health data, and pre-processing the extracted data;

[0081] The cardiovascular health analysis module 3 is used to integrate and calculate the user's real-time cardiovascular health data and the obtained cardiovascular health data from the medical consultation, so as to obtain the abnormal reference coefficient of the user's cardiovascular health data, and compare the obtained abnormal reference coefficient of the cardiovascular health data with the preset abnormal threshold of the cardiovascular health data, so as to generate a first comparison result, and determine whether the user currently has cardiovascular abnormality according to the first comparison result;

[0082] The respiratory system analysis module 4 is used to integrate and calculate the user's real-time respiratory system data and the acquired consultation respiratory system data, so as to obtain the user's respiratory system data abnormal reference coefficient, and compare the respiratory system data abnormal reference coefficient with a preset respiratory system data abnormal threshold, so as to generate a second comparison result, and determine whether the user has a respiratory system abnormality according to the second comparison result;

[0083] The sleep health analysis module 5 is used to integrate and calculate the user's real-time sleep health data and the doctor's sleep health data, so as to obtain the user's sleep health data abnormal reference coefficient, and compare the sleep health data abnormal reference coefficient with the preset sleep health data abnormal threshold, so as to generate a third comparison result, and judge whether the user has a sleep health abnormality problem according to the third comparison result;

[0084] The intelligent alarm module 6 is used for performing intelligent alarm according to the first comparison result, the second comparison result and the third comparison result.

[0085] In this embodiment: the physiological characteristic data acquisition module 1 obtains the physiological data related to the user's cardiovascular, respiratory system and sleep health in real time through devices such as health bracelets, and performs preprocessing on it, such as denoising and data formatting. This module completes the efficient collection and preliminary cleaning of data, providing accurate and high-quality input data for subsequent analysis. Its advantages lie in real-time and high efficiency, providing a solid data foundation for the health monitoring system.

[0086] Medical feature data collection module 2 extracts medical data related to cardiovascular, respiratory and sleep health from online medical records, and pre-processes the extracted data to ensure the structuring and standardization of the data. This module realizes the effective use of historical medical data, makes up for the short-term limitations that real-time data may have, provides more comprehensive background information support for the system, and makes health monitoring more accurate and personalized.

[0087] The cardiovascular health analysis module 3 integrates and calculates the real-time cardiovascular data with the cardiovascular data of the visit, obtains the abnormal reference coefficient of the cardiovascular health data, and compares it with the preset abnormal threshold to generate the first comparison result. Through this mechanism, the module accurately judges the user's cardiovascular health status and provides timely warnings when risks occur. This integrated analysis of multi-source data improves the sensitivity and accuracy of cardiovascular health monitoring and significantly reduces risk omissions.

[0088] The respiratory system analysis module 4 integrates and calculates the real-time respiratory data and the visiting respiratory data, obtains the abnormal reference coefficient of the respiratory system data, and compares it with the preset abnormal threshold to generate a second comparison result. This module judges the user's respiratory health status through comprehensive analysis, especially in the early identification of respiratory system problems, helping users to timely discover potential risks and take intervention measures.

[0089] The sleep health analysis module 5 integrates and calculates the real-time sleep data and the medical sleep data, obtains the abnormal reference coefficient of the sleep health data, and compares it with the preset threshold to generate a third comparison result. This module provides users with a comprehensive assessment of sleep quality through a scientific analysis model, identifies potential sleep disorders, and thus improves overall health.

[0090] Based on the analysis results of cardiovascular, respiratory and sleep systems, the intelligent alarm module 6 makes comprehensive judgments on multi-source data and intelligently generates alarm signals. It not only realizes the comprehensive integration of cross-domain health data, but also ensures the timeliness and accuracy of health risk warnings, allowing users to receive reminders of health abnormalities in the first place, thereby effectively avoiding potential health threats.

[0091] The system enhances the comprehensiveness and accuracy of health monitoring by integrating physiological characteristic data with medical characteristic data. Traditional systems usually rely on only a single data source or detection method, while the system can establish a connection between real-time data and historical data, thereby providing a more comprehensive health risk assessment. Specifically, the physiological characteristic data acquisition module 1 and the medical characteristic data acquisition module 2 work together to ensure the comprehensiveness and accuracy of the data. The abnormal reference coefficient is calculated by an intelligent algorithm and compared with the set threshold, which greatly improves the sensitivity and accuracy of health problem warning. Traditional methods often rely on fixed thresholds or simple judgment criteria, which are prone to miss some subtle health abnormalities. The system can identify potential health risks more promptly through dynamic, data-based analysis. In particular, the comprehensive analysis of the cardiovascular health analysis module 3, the respiratory system analysis module 4 and the sleep health analysis module 5 not only improves the sensitivity to cardiovascular, respiratory and sleep problems, but also provides users with a more personalized health assessment. The introduction of the intelligent alarm module 6 realizes the comprehensive monitoring and intelligent alarm of multi-dimensional health problems. This not only enhances the real-time nature of health monitoring, but also ensures that when health abnormalities occur, users can receive warnings and take actions in the first time, thereby reducing the severity of health risks.

[0092] Example 2: Please refer to Figure 1 , the physiological characteristic data acquisition module 1 includes a cardiovascular data acquisition unit 11, a respiratory system data acquisition unit 12 and a sleep health data acquisition unit 13;

[0093] The cardiovascular data acquisition unit 11 is used to obtain the user's real-time cardiovascular data including real-time heart rate, real-time blood oxygen saturation and real-time heart rate variability;

[0094] The real-time heart rates are recorded as A1, A2, A3, ..., An according to the timestamps;

[0095] The real-time blood oxygen saturation is recorded as B1, B2, B3, ..., Bn according to the timestamp;

[0096] The real-time heart rate variability is recorded as C1, C2, C3, ..., Cn according to the timestamp;

[0097] The respiratory system data acquisition unit 12 is used to obtain the user's real-time respiratory system data including real-time respiratory frequency, real-time respiratory depth and real-time respiratory efficiency;

[0098] The real-time respiratory rates are recorded as D1, D2, D3, ..., Dn according to the timestamps;

[0099] The real-time breathing depth is recorded as E1, E2, E3, ..., En according to the timestamp;

[0100] The real-time breathing efficiency is recorded as F1, F2, F3, ..., Fn according to the timestamp;

[0101] The sleep health data acquisition unit 13 is used to obtain the user's real-time sleep health data including real-time deep sleep ratio, real-time sleep heart rate and real-time sleep stability;

[0102] The real-time deep sleep ratio is recorded as G1, G2, G3, ..., Gn according to the timestamp;

[0103] The real-time sleeping heart rate is recorded as H1, H2, H3, ..., Hn according to the timestamp;

[0104] The real-time sleep stabilities are recorded as I1, I2, I3, ..., In according to the timestamps.

[0105] The medical characteristic data collection module 2 includes a cardiovascular report analysis unit 21, a respiratory system report analysis unit 22 and a sleep report analysis unit 23;

[0106] The cardiovascular report analysis unit 21 is used to obtain the user's cardiovascular health data including standard cardiovascular data including standard heart rate ZA, standard blood oxygen saturation ZB and standard heart rate variability ZC;

[0107] The respiratory system report analysis unit 22 is used to obtain the user's respiratory system data including the standard respiratory frequency ZD, the standard respiratory depth ZE and the standard respiratory efficiency ZF;

[0108] The sleep report analysis unit 23 is used to obtain the user's sleep health data including the standard deep sleep ratio ZG, the standard sleep heart rate ZH and the standard sleep stability ZI.

[0109] In this embodiment: Traditional health monitoring systems often rely on a single data source, making it difficult to fully assess an individual's health status. However, this system uses two-way data integration between the physiological feature data acquisition module 1 and the medical feature data acquisition module 2 to not only obtain multi-dimensional physiological data such as the user's cardiovascular, respiratory system, and sleep health in real time, but also extract key health indicators from historical medical reports. This design of combining multiple data sources significantly improves the comprehensiveness and accuracy of system health monitoring.

[0110] The system records the real-time collected data and medical report data in a refined manner according to timestamps and specific indicators. This structured data storage method not only facilitates subsequent in-depth analysis, but also enables users to quickly locate health anomalies by comparing real-time data with standard data, helping them understand potential health risks.

[0111] Through modular design, the system can integrate and calculate real-time physiological data and standard medical data and dynamically compare them. This design makes up for the deficiency of traditional monitoring systems that cannot dynamically evaluate health status in real time. In particular, when health abnormalities are found, it can provide rapid feedback, greatly improving the timeliness of health management.

[0112] Each data collection module focuses on a health field, which refines the functional division and facilitates expansion. This design can support users to flexibly adjust the monitoring scope according to their own needs and provide more personalized health management services.

[0113] The system links historical medical data with real-time health data and calculates abnormal reference coefficients of health data through AI algorithms. This linkage mechanism not only helps users understand their current health status, but also predicts potential risks through trend analysis, providing users with the possibility of early intervention to avoid the deterioration of health problems.

[0114] Example 3: Please refer to Figure 1 , the cardiovascular health analysis module 3 includes a cardiovascular data calculation unit 31 and a cardiovascular data comparison unit 32;

[0115] The cardiovascular data calculation unit 31 is used to integrate and calculate the user's real-time cardiovascular health data and the obtained medical cardiovascular health data, so as to obtain the abnormal reference coefficient of the user's cardiovascular health data, including the first cardiovascular health data abnormal reference coefficient XYZ1 and the second cardiovascular health data abnormal reference coefficient XYZ2. When any parameter in the user's real-time cardiovascular health data is greater than 40% of any parameter in the medical cardiovascular health data or any parameter in the user's real-time cardiovascular health data is less than 40% of any parameter in the medical cardiovascular health data, the AI automatically enables the second cardiovascular health data abnormal reference coefficient XYZ2. If the condition is not met, the AI defaults to using the first cardiovascular health data abnormal reference coefficient XYZ1;

[0116] Among them, the first cardiovascular health data abnormal reference coefficient XYZ1 is obtained by first integrating and calculating the window mean of the user's real-time cardiovascular health data, and then integrating and calculating the window mean with the medical cardiovascular health data;

[0117] The second cardiovascular health data abnormal reference coefficient XYZ2 is obtained by integrating and calculating the user's real-time cardiovascular health data and the obtained medical cardiovascular health data;

[0118] The cardiovascular data comparison unit 32 is used to compare the first cardiovascular health data abnormal reference coefficient XYZ1 and the second cardiovascular health data abnormal reference coefficient XYZ2 with the preset cardiovascular health data abnormal threshold respectively, so as to generate the first comparison result. The specific method is as follows:

[0119] When XYZ1≥Y1, it means that the current user does not have abnormal cardiovascular health data;

[0120] When XYZ1<Y1, it means that the current user has abnormal cardiovascular health data;

[0121] When XYZ2>Y1×120%, it means that the current user does not have abnormal cardiovascular health data;

[0122] When XYZ2≤Y2×120%, it means that the current user has abnormal cardiovascular health data.

[0123] The first cardiovascular health data abnormal reference coefficient XYZ1 and the second cardiovascular health data abnormal reference coefficient XYZ2 are obtained by the following formulas respectively:

[0124]

[0125]

[0126] Where: ZA is the standard heart rate, ZB is the standard blood oxygen saturation, ZC is the standard heart rate variability, PA is the user's heart rate window mean, PB is the user's blood oxygen saturation window mean, PC is the user's heart rate variability window mean, An is the user's real-time heart rate at the nth timestamp, Bn is the user's real-time blood oxygen saturation at the nth timestamp, and Cn is the user's real-time heart rate variability at the nth timestamp;

[0127] n is the total number of timestamps, m is the window size, k is the kth data point in the window, a1, a2 and a3 are weight values, and the values ​​of a1, a2 and a3 are adjusted and set by the user.

[0128] In this embodiment: the system designs two mechanisms: the first cardiovascular health data abnormal reference coefficient XYZ1 and the second cardiovascular health data abnormal reference coefficient XYZ2, and dynamically switches through conditional judgment. This design can flexibly adjust the calculation model according to the deviation between the user's real-time cardiovascular data and historical data, so as to more accurately capture abnormal changes in health data, making up for the inability of traditional systems to adapt to complex health conditions.

[0129] Through the cardiovascular data calculation unit 31, the system integrates and calculates the real-time heart rate An, real-time blood oxygen saturation Bn, real-time heart rate variability Cn with the standard heart rate ZA, standard blood oxygen saturation ZB and standard heart rate variability ZC to obtain the window mean and dynamic reference coefficient respectively. This design based on multi-level integration helps to comprehensively evaluate the health status and improve the accuracy of abnormality detection.

[0130] The cardiovascular data comparison unit 32 compares the reference coefficient with the preset abnormal threshold to realize intelligent judgment of the health status. In particular, when the real-time data exceeds the upper and lower limits of 40% of the historical data, the system automatically switches to the second abnormal reference coefficient to further improve the sensitivity and reaction speed to serious abnormal conditions, ensuring that abnormal conditions are identified in time.

[0131] Through the adjustable design of the weight parameters in the formula, users can adjust the weights of different indicators according to their individual health conditions, thereby customizing health data analysis. This function meets the personalized needs of different users, and is especially suitable for monitoring needs of special health conditions or specific populations.

[0132] The system introduces a time window processing method that can smooth out the interference caused by short-term fluctuations while capturing the dynamic trend of cardiovascular health data. This design provides a more accurate assessment of the user's short-term health changes and is particularly suitable for monitoring short-term abnormalities in cardiovascular data.

[0133] The formula calculation design is based on the combination of real-time and historical data weights, making full use of the analysis capabilities in the context of big data. Through the integrated calculation of standard data and real-time data, the system not only achieves efficient processing of large amounts of data but also improves the analysis depth and result credibility of health data.

[0134] Example 4: Please refer to Figure 1 , the respiratory system analysis module 4 includes a respiratory system data calculation unit 41 and a respiratory system data comparison unit 42;

[0135] The respiratory system data calculation unit 41 is used to integrally calculate the real-time respiratory system data of the user and the obtained medical treatment respiratory system data, so as to obtain the abnormal reference coefficient of the user's respiratory system data, including the first abnormal reference coefficient HYZ1 of the respiratory system data and the second abnormal reference coefficient HYZ2 of the respiratory system data. When any data under the user's real-time respiratory system data is greater than 40% of any data under the medical treatment respiratory system data or any data under the user's real-time respiratory system data is less than 40% of any data under the medical treatment respiratory system data, the AI automatically enables the second abnormal reference coefficient HYZ2 of the respiratory system data. If the condition is not met, the AI defaults to using the first abnormal reference coefficient HYZ1 of the respiratory system data;

[0136] Among them, the first abnormal reference coefficient HYZ1 of the respiratory system data is obtained by first integrally calculating the window mean value of the user's real-time respiratory system data and then integrally calculating the window mean value with the medical treatment respiratory system data;

[0137] The second abnormal reference coefficient HYZ2 of the respiratory system data is obtained by integrally calculating the user's real-time respiratory system data and the obtained medical treatment respiratory system data;

[0138] The respiratory system data comparison unit 42 is used to compare the first abnormal reference coefficient HYZ1 of the respiratory system data and the second abnormal reference coefficient HYZ2 of the respiratory system data with the preset abnormal threshold of the respiratory system data respectively, so as to generate a second comparison result. The specific method is as follows:

[0139] When HYZ1 ≥ Y2, it means that the current user does not have abnormal respiratory system data;

[0140] When HYZ1 < Y2, it means that the current user has abnormal respiratory system data;

[0141] When HYZ2 > Y2 × 120%, it means that the current user does not have abnormal respiratory system data;

[0142] When HYZ2 ≤ Y2 × 120%, it means that the current user has abnormal respiratory system data.

[0143] The first respiratory system data abnormal reference coefficient HYZ1 and the second respiratory system data abnormal reference coefficient HYZ2 are calculated and obtained by the following formulas:

[0144]

[0145] Where: ZD is the standard respiratory rate, ZE is the standard respiratory depth, ZF is the standard respiratory efficiency, PD is the user's respiratory rate window mean, PE is the user's respiratory depth window mean, PF is the user's respiratory efficiency window mean, Dn is the user's real-time respiratory rate at the nth timestamp, En is the user's real-time respiratory depth at the nth timestamp, and Fn is the user's real-time respiratory efficiency at the nth timestamp;

[0146] n is the total number of timestamps, m is the window size, k is the kth data point in the window, b1, b2 and b3 are weight values, and the values ​​of b1, b2 and b3 are adjusted and set by the user.

[0147] In this embodiment: the system designs the first respiratory system data abnormality reference coefficient HYZ1 and the second respiratory system data abnormality reference coefficient HYZ2, and automatically switches through the dynamic condition judgment mechanism. This double-layer mechanism enhances the sensitivity to abnormal conditions, allowing the system to quickly capture severe respiratory abnormalities and provide users with more timely health feedback.

[0148] The respiratory system data calculation unit 41 dynamically obtains the window mean by integrating and calculating the real-time data and the historical standard data. This method balances the short-term fluctuation and long-term trend of the data and improves the stability and accuracy of abnormal analysis.

[0149] The respiratory system data comparison unit 42 realizes intelligent judgment of the health status by comparing the abnormal reference coefficient with the preset threshold value. By analyzing the abnormal reference coefficient, it can quickly distinguish whether the user has a situation of too fast breathing rate, insufficient depth or abnormal efficiency, providing a scientific basis for early intervention.

[0150] The weight parameters in the formula allow users to adjust the importance of different indicators according to their personal health status or needs. For asthma patients, the weight of breathing efficiency can be increased, and for those with sleep apnea, the weight of breathing depth can be increased. This design realizes the personalized configuration of the monitoring algorithm to adapt to different health scenarios.

[0151] The system uses a time window-based analysis method that can smooth out short-term fluctuations while accurately capturing abnormal dynamic changes in the respiratory system. This short-term dynamic analysis capability is particularly suitable for detecting acute changes in respiratory rate, depth, and efficiency.

[0152] By integrating real-time data with standard data, the system comprehensively analyzes the respiratory system using multiple metrics, significantly enhancing the detection ability for abnormal conditions. Compared with traditional single-parameter monitoring, the system can more comprehensively evaluate the user's respiratory health status, avoiding missed or misjudged cases caused by insufficient single-item data.

[0153] With the big data analysis ability of real-time data and historical data, the system supports the efficient processing and mining of diverse health data. In the analysis of respiratory system data, this function provides a reliable analysis model for complex health states and lays a foundation for future data expansion and model optimization.

[0154] Example Five: Please refer to Figure 1 , the sleep health analysis module 5 includes a sleep health data calculation unit 51 and a sleep health data comparison unit 52;

[0155] The sleep health data calculation unit 51 is used to integrate and calculate the user's real-time sleep health data and the obtained medical sleep health data, thereby obtaining the sleep health data anomaly reference coefficients of the user, including the first sleep health data anomaly reference coefficient SYZ1 and the second sleep health data anomaly reference coefficient SYZ2. When any parameter in the user's real-time sleep health data is greater than 40% of any parameter in the medical sleep health data or any parameter in the user's real-time sleep health data is less than 40% of any parameter in the medical sleep health data, the AI automatically enables the second sleep health data anomaly reference coefficient SYZ2. If the condition is not met, the AI defaults to using the first sleep health data anomaly reference coefficient SYZ1;

[0156] Among them, the first sleep health data anomaly reference coefficient SYZ1 is obtained by first integrating and calculating the window mean of the user's real-time sleep health data, and then integrating and calculating the window mean with the medical sleep health data;

[0157] The second sleep health data anomaly reference coefficient SYZ2 is obtained by integrating and calculating the user's real-time sleep health data and the obtained medical sleep health data;

[0158] The sleep health data comparison unit 52 is used to compare the first sleep health data anomaly reference coefficient SYZ1 and the second sleep health data anomaly reference coefficient SYZ2 with the preset sleep health data anomaly threshold respectively, thereby generating a third comparison result. The specific method is as follows:

[0159] When SYZ1 ≥ Y3, it means that the current user does not have any sleep health data anomalies;

[0160] When SYZ1 < Y3, it means that the current user has sleep health data anomalies;

[0161] When SYZ2>Y3×120%, it means that the current user has no abnormal sleep health data;

[0162] When SYZ2≤Y3×120%, it means that the current user has abnormal sleep health data.

[0163] The first sleep health data abnormal reference coefficient SYZ1 and the second sleep health data abnormal reference coefficient SYZ2 are respectively calculated and obtained by the following formulas:

[0164]

[0165] Wherein: ZG is the standard deep sleep ratio, ZH is the standard sleep heart rate, ZI is the standard sleep stability, PG is the user's deep sleep ratio window mean, PH is the user's sleep heart rate window mean, PI is the user's sleep stability window mean, Gn is the user's real-time deep sleep ratio at the nth timestamp, Hn is the user's real-time sleep heart rate at the nth timestamp, and In is the user's real-time sleep stability at the nth timestamp;

[0166] n is the total number of timestamps, m is the window size, k is the kth data point in the window, c1, c2 and c3 are weight values, and the values ​​of c1, c2 and c3 are adjusted and set by the user.

[0167] In this embodiment: by introducing the first sleep health data abnormal reference coefficient SYZ1 and the second sleep health data abnormal reference coefficient SYZ2, the system can dynamically switch the reference standard. When the real-time sleep data exceeds the consultation sleep data by 40%, the second reference coefficient is automatically enabled to adapt to significant fluctuations in sleep status. This mechanism improves the flexibility and accuracy of the system in dealing with abnormal conditions.

[0168] The sleep health data calculation unit 51 calculates the window mean value and combines the standard sleep health data to smooth the interference of short-term fluctuations on the analysis results while retaining long-term trend information. This method enhances the stability and reliability of data anomaly analysis.

[0169] The system provides an adjustable function for weight parameters, and users can customize the importance of indicators according to their personal needs. For example, the weight of deep sleep ratio can be increased for patients with long-term insomnia, and the weight of sleeping heart rate can be increased for patients with heart rate-related sleep disorders. This design meets the personalized monitoring needs of different users and improves the pertinence and practicality of monitoring results.

[0170] The sleep health data comparison unit 52 generates a third comparison result by comparing the abnormal reference coefficient with the preset threshold value to determine whether the user has problems such as insufficient deep sleep, abnormal heart rate fluctuation, or low sleep stability. This intelligent abnormality detection method based on big data provides users with timely and accurate sleep health assessment.

[0171] The system integrates real-time sleep health data with historical medical data to form a comprehensive abnormality analysis capability. Compared with systems that rely solely on real-time data or historical data, this design more comprehensively reflects the user's sleep health dynamics and long-term trends, significantly reducing the possibility of misjudgment and missed judgment.

[0172] The dynamic analysis design based on time windows ensures that the system can quickly identify drastic changes in sleep status within a short period of time, helping users to detect potential health risks in a timely manner.

[0173] With the help of big data analysis of real-time data and historical data, the system realizes multi-dimensional and efficient data processing and model optimization. This function enables the sleep health analysis module 5 to adapt to application scenarios with different data volumes and complexities, and provides a technical foundation for future expansion of functions.

[0174] Example 6: Please refer to Figure 1 The intelligent alarm module 6 uses AI to intelligently identify the current time period and makes intelligent alarms according to the time period. The specific methods are as follows:

[0175] When it is not nighttime, AI identifies the first comparison result and the second comparison result. When the first cardiovascular health data abnormal reference coefficient XYZ1 and the first respiratory system data abnormal reference coefficient HYZ1 are enabled, if the first comparison result and the second comparison result are both in data abnormality, AI adjusts the bracelet LED screen color to yellow and gives a sound prompt at a frequency of once every two seconds. If either of the first comparison result and the second comparison result is in data abnormality, AI does not give an alarm. When the second cardiovascular health data abnormal reference coefficient XYZ2 and the second respiratory system data abnormal reference coefficient HYZ2 are enabled, if either is in, AI adjusts the bracelet LED screen color to red, gives a sound prompt twice every second, and sends a message to the emergency contact and makes a call to a smart phone.

[0176] When it is at night, AI identifies the second comparison result and the third comparison result. When the first respiratory system data abnormal reference coefficient HYZ1 and the first sleep health data abnormal reference coefficient SYZ1 are enabled, if the second comparison result and the third comparison result are both in, AI adjusts the bracelet LED screen color to yellow, and makes a sound prompt at a frequency of once every two seconds, and the prompt tone is increased by 30% compared to non-night time. If either the second comparison result or the third comparison result is in data abnormality, AI does not alarm. When the second respiratory system data abnormal reference coefficient HYZ2 and the second sleep health data abnormal reference coefficient SYZ2 are enabled, if either the second comparison result or the third comparison result is in, AI adjusts the bracelet LED screen color to red, and makes a sound prompt twice every second, and the prompt tone is increased by 150% compared to non-night time, and calls the emergency contact on a smart phone.

[0177] In this embodiment: by distinguishing between nighttime and non-nighttime states, the intelligent alarm module 6 can intelligently adapt to the alarm requirements of different time periods. For example, the alarm volume is automatically adjusted up at night to ensure that users or emergency contacts can detect it in a quiet environment in time, while avoiding disturbing normal daytime activities due to unnecessary reminder sounds. The intelligent alarm module 6 integrates the first comparison result, the second comparison result, and the third comparison result. Through comprehensive judgment of multi-dimensional data, it reduces the false alarms that may be caused by single data anomalies, and improves the accuracy and reliability of anomaly identification. According to the category of abnormal reference coefficients, the intelligent alarm module 6 adopts different alarm responses. When the first reference coefficient is enabled, the alarm is only triggered when all data are abnormal, avoiding overreaction to slight data fluctuations. When the second reference coefficient is enabled, any data abnormality triggers a high-intensity alarm. This hierarchical alarm design ensures a rapid response to serious health risks while avoiding unnecessary interference to users.

[0178] The alarm mode includes three ways: visual prompts, auditory prompts and emergency communications, ensuring that alarm information in different environments and states can be effectively transmitted. In particular, when the data is abnormally serious, the emergency contact can be directly dialed through the smart phone, which greatly improves the efficiency of responding to sudden health events. The intelligent alarm module 6 can automatically switch the alarm state according to the real-time data changes. For example, during non-nighttime, when only one of the first comparison result and the second comparison result is abnormal, the system does not trigger the alarm to avoid frequent invalid prompts. This design balances the sensitivity of health monitoring and user experience, improves the practicality and friendliness of the system, and the intelligent alarm module 6 supports dynamic adjustment of the alarm intensity according to the time period, such as significantly increasing the alarm volume and frequency at night. This personalized adjustment mechanism can not only ensure the timely delivery of important information, but also adapt to the needs of users in different scenarios. With the help of AI technology, the intelligent alarm module 6 can quickly integrate multi-dimensional data, dynamically judge the reference coefficient and comparison results, and accurately identify changes in the user's health status. This intelligent alarm trigger logic based on big data analysis effectively reduces false alarms and missed alarms, and significantly improves the reliability and response efficiency of the system.

[0179] The contents not described in detail in this specification belong to the prior art known to professional and technical personnel in this field.

[0180] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An AI intelligent health monitoring and management system based on big data, characterized by: It includes a physiological characteristic data collection module (1), a medical characteristic data collection module (2), a cardiovascular health analysis module (3), a respiratory system analysis module (4), a sleep health analysis module (5) and an intelligent alarm module (6); The physiological characteristic data acquisition module (1) is used to acquire the user's real-time cardiovascular health data, real-time respiratory system data and real-time sleep health data, and to pre-process the acquired data; The medical characteristic data acquisition module (2) is used to acquire the user's online medical data, thereby extracting the patient's cardiovascular health data, respiratory system data and sleep health data, and pre-processing the extracted data; The cardiovascular health analysis module (3) is used to integrate and calculate the user's real-time cardiovascular health data and the obtained cardiovascular health data from the medical consultation, thereby obtaining the user's cardiovascular health data abnormality reference coefficient, and compare the calculated cardiovascular health data abnormality reference coefficient with a preset cardiovascular health data abnormality threshold, thereby generating a first comparison result, and judging whether the user currently has a cardiovascular abnormality based on the first comparison result; The respiratory system analysis module (4) is used to integrate and calculate the user's real-time respiratory system data and the acquired consultation respiratory system data, thereby obtaining the user's respiratory system data abnormality reference coefficient, and compare the respiratory system data abnormality reference coefficient with a preset respiratory system data abnormality threshold, thereby generating a second comparison result, and judging whether the user has a respiratory system abnormality based on the second comparison result; The sleep health analysis module (5) is used to integrate and calculate the user's real-time sleep health data and the consultation sleep health data, so as to obtain the user's sleep health data abnormal reference coefficient, and compare the sleep health data abnormal reference coefficient with a preset sleep health data abnormal threshold, so as to generate a third comparison result, and determine whether the user has a sleep health abnormality problem according to the third comparison result; The intelligent alarm module (6) is used for performing intelligent alarm according to the first comparison result, the second comparison result and the third comparison result.

2. The AI ​​intelligent health monitoring and management system based on big data according to claim 1 is characterized by: The physiological characteristic data acquisition module (1) comprises a cardiovascular data acquisition unit (11), a respiratory system data acquisition unit (12) and a sleep health data acquisition unit (13); The cardiovascular data acquisition unit (11) is used to obtain the user's real-time cardiovascular data including real-time heart rate, real-time blood oxygen saturation and real-time heart rate variability; The real-time heart rates are recorded as A1, A2, A3, ..., An according to the timestamps; The real-time blood oxygen saturation is recorded as B1, B2, B3, ..., Bn according to the timestamp; The real-time heart rate variability is recorded as C1, C2, C3, ..., Cn according to the timestamp; The respiratory system data acquisition unit (12) is used to obtain real-time respiratory system data of the user including real-time respiratory frequency, real-time respiratory depth and real-time respiratory efficiency; The real-time respiratory rates are recorded as D1, D2, D3, ..., Dn according to the timestamps; The real-time breathing depth is recorded as E1, E2, E3, ..., En according to the timestamp; The real-time breathing efficiency is recorded as F1, F2, F3, ..., Fn according to the timestamp; The sleep health data acquisition unit (13) is used to obtain the user's real-time sleep health data including real-time deep sleep ratio, real-time sleep heart rate and real-time sleep stability; The real-time deep sleep ratio is recorded as G1, G2, G3, ..., Gn according to the timestamp; The real-time sleeping heart rate is recorded as H1, H2, H3, ..., Hn according to the timestamp; The real-time sleep stabilities are recorded as I1, I2, I3, ..., In according to the timestamps.

3. The AI ​​intelligent health monitoring and management system based on big data according to claim 2 is characterized in that: The medical characteristic data collection module (2) comprises a cardiovascular report analysis unit (21), a respiratory system report analysis unit (22) and a sleep report analysis unit (23); The cardiovascular report analysis unit (21) is used to obtain the user's cardiovascular health data including standard cardiovascular data including standard heart rate ZA, standard blood oxygen saturation ZB and standard heart rate variability ZC; The respiratory system report analysis unit (22) is used to obtain the user's respiratory system data including standard respiratory frequency ZD, standard respiratory depth ZE and standard respiratory efficiency ZF; The sleep report analysis unit (23) is used to obtain the user's sleep health data including the standard deep sleep ratio ZG, the standard sleep heart rate ZH and the standard sleep stability ZI.

4. The AI ​​intelligent health monitoring and management system based on big data according to claim 3 is characterized by: The cardiovascular health analysis module (3) comprises a cardiovascular data calculation unit (31) and a cardiovascular data comparison unit (32); The cardiovascular data calculation unit (31) is used to integrate and calculate the user's real-time cardiovascular health data and the acquired cardiovascular health data of the visit, so as to obtain the abnormal reference coefficient of the user's cardiovascular health data, including a first abnormal reference coefficient of cardiovascular health data XYZ1 and a second abnormal reference coefficient of cardiovascular health data XYZ2. When any parameter under the user's real-time cardiovascular health data is greater than 40% of any parameter under the visit cardiovascular health data or any parameter under the user's real-time cardiovascular health data is less than 40% of any parameter under the visit cardiovascular health data, the AI ​​automatically enables the second abnormal reference coefficient of cardiovascular health data XYZ2. If the condition is not met, the AI ​​defaults to using the first abnormal reference coefficient of cardiovascular health data XYZ1. The first cardiovascular health data abnormal reference coefficient XYZ1 is obtained by first integrating and calculating the user's real-time cardiovascular health data to obtain a window mean, and then integrating and calculating the window mean with the visit cardiovascular health data; The second cardiovascular health data abnormal reference coefficient XYZ2 is obtained by integrating and calculating the user's real-time cardiovascular health data and the obtained cardiovascular health data of the visit; The cardiovascular data comparison unit (32) is used to compare the first cardiovascular health data abnormality reference coefficient XYZ1 and the second cardiovascular health data abnormality reference coefficient XYZ2 with a preset cardiovascular health data abnormality threshold respectively, so as to generate a first comparison result. The specific method is as follows: When XYZ1≥Y1, it means that the current user does not have abnormal cardiovascular health data; When XYZ1<Y1, it means that the current user has abnormal cardiovascular health data; When XYZ2>Y1×120%, it means that the current user does not have abnormal cardiovascular health data; When XYZ2≤Y2×120%, it means that the current user has abnormal cardiovascular health data.

5. The AI ​​intelligent health monitoring and management system based on big data according to claim 4 is characterized in that: The first cardiovascular health data abnormality reference coefficient XYZ1 and the second cardiovascular health data abnormality reference coefficient XYZ2 are respectively obtained by the following formulas: In the formula: ZA is the standard heart rate, ZB is the standard blood oxygen saturation, ZC is the standard heart rate variability, PA is the average value of the user's heart rate window, PB is the average value of the user's blood oxygen saturation window, PC is the average value of the user's heart rate variability window, An is the user's real-time heart rate at the nth timestamp, Bn is the user's real-time blood oxygen saturation at the nth timestamp, Cn is the user's real-time heart rate variability at the nth timestamp; n is the total number of timestamps, m is the window size, k is the kth data point in the window, a1, a2 and a3 are weight values, and the values of a1, a2 and a3 are adjusted and set by the user.

6. The AI ​​intelligent health monitoring and management system based on big data according to claim 5 is characterized by: The respiratory system analysis module (4) includes a respiratory system data calculation unit (41) and a respiratory system data comparison unit (42); The respiratory system data calculation unit (41) is used to integrate and calculate the user's real-time respiratory system data and the obtained medical respiratory system data, so as to obtain the respiratory system data abnormality reference coefficient of the user, including the first respiratory system data abnormality reference coefficient HYZ1 and the second respiratory system data abnormality reference coefficient HYZ2. When any data in the user's real-time respiratory system data is greater than 40% of any data in the medical respiratory system data or any data in the user's real-time respiratory system data is less than 40% of any data in the medical respiratory system data, AI automatically enables the second respiratory system data abnormality reference coefficient HYZ2. If the condition is not met, AI defaults to using the first respiratory system data abnormality reference coefficient HYZ1; Among them, the first respiratory system data abnormality reference coefficient HYZ1 is obtained by first integrating and calculating the window average value of the user's real-time respiratory system data, and then integrating and calculating the window average value with the medical respiratory system data; The second respiratory system data abnormality reference coefficient HYZ2 is obtained by integrating and calculating the user's real-time respiratory system data and the obtained medical respiratory system data; The respiratory system data comparison unit (42) is used to compare the first respiratory system data abnormality reference coefficient HYZ1 and the second respiratory system data abnormality reference coefficient HYZ2 with a preset respiratory system data abnormality threshold respectively, so as to generate a second comparison result. The specific method is as follows: When HYZ1≥Y2, it means that the current user does not have abnormal respiratory system data; When HYZ1<Y2, it means that the current user has abnormal respiratory system data; When HYZ2>Y2×120%, it means that the current user does not have abnormal respiratory system data; When HYZ2≤Y2×120%, it means that the current user has abnormal respiratory system data.

7. The AI ​​intelligent health monitoring and management system based on big data according to claim 6 is characterized by: The first respiratory system data abnormality reference coefficient HYZ1 and the second respiratory system data abnormality reference coefficient HYZ2 are respectively calculated and obtained through the following formulas: In the formula: ZD is the standard respiratory rate, ZE is the standard respiratory depth, ZF is the standard respiratory efficiency, PD is the average value of the user's respiratory rate window, PE is the average value of the user's respiratory depth window, PF is the average value of the user's respiratory efficiency window, Dn is the user's real-time respiratory rate at the nth timestamp, En is the user's real-time respiratory depth at the nth timestamp, Fn is the user's real-time respiratory efficiency at the nth timestamp; n is the total number of timestamps, m is the window size, k is the kth data point within the window, b1, b2, and b3 are weight values, and the values of b1, b2, and b3 are adjusted and set by the user.

8. The AI ​​intelligent health monitoring and management system based on big data according to claim 7 is characterized by: The sleep health analysis module (5) includes a sleep health data calculation unit (51) and a sleep health data comparison unit (52); The sleep health data calculation unit (51) is used to integrate and calculate the user's real-time sleep health data and the obtained medical sleep health data, so as to obtain the sleep health data abnormality reference coefficient of the user, including the first sleep health data abnormality reference coefficient SYZ1 and the second sleep health data abnormality reference coefficient SYZ2. When any parameter in the user's real-time sleep health data is greater than 40% of any parameter in the medical sleep health data or any parameter in the user's real-time sleep health data is less than 40% of any parameter in the medical sleep health data, AI automatically enables the second sleep health data abnormality reference coefficient SYZ2. If the condition is not met, AI defaults to using the first sleep health data abnormality reference coefficient SYZ1; Among them, the first sleep health data abnormality reference coefficient SYZ1 is obtained by first integrating and calculating the window average value of the user's real-time sleep health data, and then integrating and calculating the window average value with the medical sleep health data; The second sleep health data abnormality reference coefficient SYZ2 is obtained by integrating and calculating the user's real-time sleep health data and the obtained medical sleep health data. The sleep health data comparison unit (52) is used to compare the first sleep health data anomaly reference coefficient SYZ1 and the second sleep health data anomaly reference coefficient SYZ2 with a preset sleep health data anomaly threshold respectively, so as to generate a third comparison result. The specific method is as follows: When SYZ1≥Y3, it means that there is no sleep health data anomaly for the current user; When SYZ1<Y3, it means that the current user has a sleep health data anomaly; When SYZ2>Y3×120%, it means that there is no sleep health data anomaly for the current user; When SYZ2≤Y3×120%, it means that the current user has a sleep health data anomaly.

9. The AI ​​intelligent health monitoring and management system based on big data according to claim 8 is characterized by: The first sleep health data anomaly reference coefficient SYZ1 and the second sleep health data anomaly reference coefficient SYZ2 are respectively calculated and obtained through the following formulas: In the formula: ZG is the standard deep sleep ratio, ZH is the standard sleep heart rate, ZI is the standard sleep stability, PG is the window mean value of the user's deep sleep ratio, PH is the window mean value of the user's sleep heart rate, PI is the window mean value of the user's sleep stability, Gn is the real-time deep sleep ratio of the user at the nth time stamp, Hn is the real-time sleep heart rate of the user at the nth time stamp, In is the real-time sleep stability of the user at the nth time stamp; n is the total number of time stamps, m is the window size, k is the kth data point within the window, c1, c2, and c3 are weight values, and the values of c1, c2, and c3 are adjusted and set by the user.

10. The AI ​​intelligent health monitoring and management system based on big data according to claim 9 is characterized in that: The intelligent alarm module (6) intelligently identifies the current time period through AI and performs intelligent alarm according to the time period. The specific method is as follows: When it is not night time, AI identifies the first comparison result and the second comparison result. When the first cardiovascular health data anomaly reference coefficient XYZ1 and the first respiratory system data anomaly reference coefficient HYZ1 are enabled, if both the first comparison result and the second comparison result are in data anomaly, AI adjusts the color of the bracelet LED screen to yellow and gives a sound prompt at a frequency of once every two seconds. If either the first comparison result or the second comparison result is in data anomaly, AI does not give an alarm. When the second cardiovascular health data anomaly reference coefficient XYZ2 and the second respiratory system data anomaly reference coefficient HYZ2 are enabled, if either is in data anomaly, AI adjusts the color of the bracelet LED screen to red and gives a sound prompt twice every second, and sends information to the emergency contact and makes an intelligent call; When it is at night, AI identifies the second comparison result and the third comparison result. When the first respiratory system data abnormal reference coefficient HYZ1 and the first sleep health data abnormal reference coefficient SYZ1 are enabled, if the second comparison result and the third comparison result are both in, AI adjusts the bracelet LED screen color to yellow, and makes a sound prompt at a frequency of once every two seconds, and the prompt tone is increased by 30% compared to non-night time. If either the second comparison result or the third comparison result is in data abnormality, AI does not alarm. When the second respiratory system data abnormal reference coefficient HYZ2 and the second sleep health data abnormal reference coefficient SYZ2 are enabled, if either the second comparison result or the third comparison result is in, AI adjusts the bracelet LED screen color to red, and makes a sound prompt twice every second, and the prompt tone is increased by 150% compared to non-night time, and calls the emergency contact on a smart phone.

Images