An AI health monitoring and management system based on intelligent diagnosis

Abstract

The present invention discloses an AI health monitoring and management system based on intelligent diagnosis, which relates to the field of data analysis technology. For each type of user health data, the timestamp of the data uploaded to the intelligent diagnosis system within a preset time window is obtained to obtain the data update fluctuation anomaly coefficient and the data update delay coefficient; and the number of uploaded user health data of each type is used to calculate the data upload incomplete coefficient; and the data false coefficient is calculated, and the real state of the user health data is divided into a false state and a temporary real state in combination with a preset threshold; for the user health data in the temporary real state, the data false coefficient sequence is obtained to judge the final real state of the user health data, which can effectively verify the authenticity of the health data; it can identify and correct potential problems of errors or deviations in the transmitted data, ensure the accuracy of the intelligent diagnosis results, and reduce the misleading of users to make inappropriate health decisions or take ineffective health management measures.

Description

Technical Field

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

[0002] AI-powered health monitoring and management systems based on intelligent diagnosis use artificial intelligence (AI) technology to monitor, diagnose, and manage the health status of individuals or groups in real time. These systems typically combine sensors, wearable devices, health data analysis, machine learning, and big data technologies to conduct in-depth analysis of users' health data and, through intelligent algorithms, provide health predictions, risk assessments, diagnostic recommendations, and personalized health management plans. These systems are widely used in a variety of fields, including chronic disease management, elderly health monitoring, sports health management, and mental health monitoring. Their advantage lies in their ability to detect and intervene in health conditions early, improving disease prevention capabilities and increasing the efficiency of medical resource utilization.

[0003] AI health monitoring and management systems based on intelligent diagnosis rely on users' health data transmitted by wearable devices and health monitoring sensors, and often assume that this data is accurate and perform intelligent diagnosis. However, these systems often lack effective verification mechanisms for the authenticity of health data. Since sensors may be affected by various factors, the transmitted data may contain errors or deviations. If the system fails to identify and correct these potential data problems, it may lead to errors in the intelligent diagnosis results, thereby misleading users to make inappropriate health decisions or take ineffective health management measures. Summary of the Invention

[0004] The purpose of the present invention is to solve the above-mentioned problem that may lead to errors in intelligent diagnosis results, thereby misleading users to make inappropriate health decisions or take ineffective health management measures, and to provide an AI health monitoring and management system based on intelligent diagnosis.

[0005] The present invention proposes an AI health monitoring and management system based on intelligent diagnosis, the system comprising:

[0006] Update module: For each type of user health data, obtain the timestamp of its upload to the intelligent diagnosis system within the preset time window, and obtain the data update fluctuation anomaly coefficient and data update delay coefficient based on the timestamp;

[0007] Incomplete module: obtains the number of user health data uploads of each type within a preset time window, and calculates the data upload incomplete coefficient based on the number of user health data uploads of each type;

[0008] False coefficient module: derives the data false coefficient based on the data update fluctuation anomaly coefficient, data update delay coefficient, and data upload incomplete coefficient; and divides the true state of the user's health data into a false state and a temporary true state based on the preset data false coefficient threshold;

[0009] Final judgment module: For the user health data in a temporary real state, continuously obtain the data false coefficient, and sequence it according to time to obtain the data false coefficient sequence, and judge the final real state of the user health data based on the data false coefficient sequence.

[0010] Optionally, obtaining the data update fluctuation anomaly coefficient according to the timestamp includes:

[0011] For each type of user health data, obtain the actual timestamp when it is uploaded to the intelligent diagnosis system within the preset time window, and obtain a timestamp sequence based on time order;

[0012] Calculate the time interval between every two adjacent timestamps in the timestamp sequence, and sequence them to obtain the upload interval sequence;

[0013] The upload interval in the upload interval sequence is marked as E 实 u , u represents the sequence number of the upload interval in the upload interval sequence, u=1, 2, 3, 4, ..., p, p is the total number of upload intervals in the upload interval sequence and p is a positive integer;

[0014] Calculate the mean of the upload interval sequence using the following formula:

[0015] Calculate the data update fluctuation anomaly coefficient using the following formula:

[0016]

[0017] Wherein, YK is the data update fluctuation anomaly coefficient, f is the type number of the user health data, m is the total number of types of the user health data, and m is a positive integer.

[0018] Optionally, obtaining a data update delay coefficient according to a timestamp includes:

[0019] For each type of user health data, obtain the actual timestamp when it is uploaded to the intelligent diagnosis system within the preset time window, and obtain a timestamp sequence based on time order;

[0020] Calculate the time interval between every two adjacent timestamps in the timestamp sequence, and sequence them to obtain the upload interval sequence; mark the upload interval in the upload interval sequence as E 实 u, u represents the sequence number of the upload interval in the upload interval sequence, u=1, 2, 3, 4, ..., p, p is the total number of upload intervals in the upload interval sequence and p is a positive integer;

[0021] Set the upload interval E in the upload interval sequence 实 u and within the preset upload interval range (T 快 , T 慢 ) is re-marked as Q w , w represents E 实 u Within the preset data upload interval range (T 快 , T 慢 ) is a sequence number of the time outside the time interval, w = 0, 1, 2, 3, 4, ..., d, d is a positive integer; the data update delay coefficient AG is calculated using the following formula: f is the type number of the user health data, m is the total number of types of the user health data, and m is a positive integer.

[0022] Optionally, calculating the data upload incomplete coefficient according to the number of uploaded user health data of each type includes:

[0023] Get the actual number of health data uploaded by each type of user and the corresponding preset health data upload number within the preset time window, and mark the actual data upload number and the corresponding preset data upload number as G respectively. 实 f and G 预 f , f is the type number of the user health data, m is the total number of types of user health data, and m is a positive integer;

[0024] According to G 实 f and G 预 f The data upload incomplete coefficient DF is obtained, and the calculation formula is:

[0025] Optionally, obtaining a data falsehood coefficient based on a data update fluctuation anomaly coefficient, a data update delay coefficient, and a data upload incomplete coefficient includes:

[0026]

[0027] Where Dax is the data false coefficient, YK, AG, and DF are the data update fluctuation anomaly coefficient, data update delay coefficient, and data upload incomplete coefficient, respectively. α, β, and γ are the preset proportional coefficients of YK, AG, and DF, respectively, and α, β, and γ are all greater than 0.

[0028] Optionally, dividing the true state of the user health data into a false state and a temporary true state in combination with a preset data false coefficient threshold includes:

[0029] Compare the data false coefficient with the preset data false coefficient threshold. If the data false coefficient is not less than the preset data false coefficient threshold, it indicates that the true state of the user's health data is false, and issue a warning signal;

[0030] The data false coefficient is compared with the preset data false coefficient threshold. If the data false coefficient is less than the preset data false coefficient threshold, it means that the user health data is temporarily true and needs further analysis.

[0031] Optionally, judging the final true state of the user health data according to the data false coefficient sequence includes:

[0032] Calculate the standard deviation and mean of the data false coefficient sequence, and compare the standard deviation and mean of the data false coefficient sequence with the preset standard deviation and preset mean respectively;

[0033] If the standard deviation of the data false coefficient sequence is less than the preset standard deviation and the mean is less than the preset mean, it means that the final true state of the user health data is true; otherwise, it means that the final true state of the user health data is a false state.

[0034] Beneficial effects of the present invention:

[0035] The present invention proposes an AI health monitoring and management system based on intelligent diagnosis. For each type of user health data, the timestamp of its upload to the intelligent diagnosis system within a preset time window is obtained to obtain the data update fluctuation anomaly coefficient and the data update delay coefficient; and the data upload incomplete coefficient is calculated according to the number of uploaded user health data of each type; the data false coefficient is obtained according to the data update fluctuation anomaly coefficient, the data update delay coefficient and the data upload incomplete coefficient; and the real state of the user health data is divided into a false state and a temporary real state in combination with a preset data false coefficient threshold; for user health data in a temporarily real state, the data false coefficient sequence is obtained to judge the final real state of the user health data, which can effectively verify the authenticity of the health data; the system can identify and correct potential problems of errors or deviations in the transmitted data, ensure the accuracy of the intelligent diagnosis results, and reduce the misleading of users to make inappropriate health decisions or take ineffective health management measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The present invention will be further described below with reference to the accompanying drawings.

[0037] Figure 1This is a framework diagram of an AI health monitoring and management system based on intelligent diagnosis. DETAILED DESCRIPTION

[0038] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0039] Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making any creative work shall fall within the scope of protection of the present invention.

[0040] The embodiment of the present invention provides an AI health monitoring and management system based on intelligent diagnosis. Figure 1 , Figure 1 This is a framework diagram of an AI health monitoring and management system based on intelligent diagnosis provided by an embodiment of the present invention. The system includes:

[0041] Update module: For each type of user health data, obtain the timestamp of its upload to the intelligent diagnosis system within the preset time window, and obtain the data update fluctuation anomaly coefficient and data update delay coefficient based on the timestamp;

[0042] Incomplete module: obtains the number of user health data uploads of each type within a preset time window, and calculates the data upload incomplete coefficient based on the number of user health data uploads of each type;

[0043] False coefficient module: derives the data false coefficient based on the data update fluctuation anomaly coefficient, data update delay coefficient, and data upload incomplete coefficient; and divides the true state of the user's health data into a false state and a temporary true state based on the preset data false coefficient threshold;

[0044] Final judgment module: For the user health data in a temporary real state, continuously obtain the data false coefficient, and sequence it according to time to obtain the data false coefficient sequence, and judge the final real state of the user health data based on the data false coefficient sequence.

[0045] An AI health monitoring and management system based on intelligent diagnosis provided in an embodiment of the present invention can effectively verify the authenticity of health data; the system can identify and correct potential problems of errors or deviations in the transmitted data, ensure the accuracy of the intelligent diagnosis results, and reduce the misleading of users into making inappropriate health decisions or taking ineffective health management measures.

[0046] In one embodiment, obtaining the data update fluctuation anomaly coefficient according to the timestamp includes:

[0047] For each type of user health data, obtain the actual timestamp when it is uploaded to the intelligent diagnosis system within the preset time window, and obtain a timestamp sequence based on time order;

[0048] Calculate the time interval between every two adjacent timestamps in the timestamp sequence, and sequence them to obtain the upload interval sequence;

[0049] The upload interval in the upload interval sequence is marked as E 实 u , u represents the sequence number of the upload interval in the upload interval sequence, u=1, 2, 3, 4, ..., p, p is the total number of upload intervals in the upload interval sequence and p is a positive integer;

[0050] Calculate the mean of the upload interval sequence using the following formula:

[0051] Calculate the data update fluctuation anomaly coefficient using the following formula:

[0052]

[0053] Wherein, YK is the data update fluctuation anomaly coefficient, f is the type number of the user health data, m is the total number of types of the user health data, and m is a positive integer.

[0054] It should be noted that the preset time window is set by professionals based on actual circumstances and is not specifically limited or elaborated upon.

[0055] It should be noted that each type of user health data refers to various physiological health indicators collected by wearable devices or health monitoring sensors, such as heart rate, blood pressure, body temperature, blood oxygen saturation, number of steps, sleep quality, and weight. This health data typically changes over time and is uploaded to the intelligent diagnostic system by the device in real time or periodically. This data is typically acquired through real-time connection and synchronization with wearable devices (such as smartwatches and fitness bands) or other health sensors (such as blood pressure monitors and blood glucose meters). The device regularly measures the user's physiological parameters and transmits the data to the system via the network. This uploaded health data is timestamped to record the specific time the data was uploaded. The system analyzes these timestamp sequences, calculates the upload interval and its fluctuations, and thus assesses the stability and authenticity of the data. Therefore, the acquisition of health data depends not only on the measurement accuracy of the device but also on the reliability and stability of data transmission to ensure that the acquired health data is representative and conforms to expected regularity.

[0056] It should be noted that the data update fluctuation anomaly coefficient is an indicator used to measure the fluctuation in health data upload intervals. It assesses the stability of uploaded data by calculating the ratio of the standard deviation of the upload intervals to the mean in the timestamp sequence. When health data upload intervals fluctuate significantly, this may indicate instability during data collection or transmission, such as device failure, network latency, users not wearing their devices promptly, or inaccurate measurements. A larger fluctuation anomaly coefficient indicates more drastic changes in the intervals between data uploads and greater data instability. Such fluctuations often make it difficult for the system to accurately obtain continuous and reliable health information. Therefore, an increase in the data update fluctuation anomaly coefficient raises questions about the authenticity of the data. The system may be unable to make accurate diagnoses or predictions based on this unstable data, increasing the likelihood of errors in intelligent diagnostic results. A higher fluctuation anomaly coefficient may indicate poor sensor data quality, which in turn affects the accuracy of health monitoring and may even mislead users into making incorrect health decisions.

[0057] In one implementation, analyzing the data update fluctuation anomaly coefficient for determining the authenticity of user health data provides a means of quantifying the regularity of data uploads. By monitoring fluctuations in the intervals between health data uploads, the system can promptly identify abnormal upload patterns, such as excessively long or short upload intervals, or frequent fluctuations. These can signal potential problems in data collection or transmission. For example, if the intervals between health data uploads exhibit unusually large fluctuations, this could indicate device failure, network instability, or inconsistent user wearing behavior. These issues can distort data and affect the accuracy of intelligent diagnosis. By calculating the fluctuation anomaly coefficient, the system can identify these abnormal fluctuations, providing early warning of potential data authenticity issues and preventing erroneous health judgments based on unreliable data. Furthermore, this coefficient helps the system distinguish between data within the normal fluctuation range and abnormalities likely caused by external factors or device issues, enabling it to take appropriate remedial measures, such as recollecting data or notifying the user to check their device, ensuring more accurate and reliable diagnostic results.

[0058] In one embodiment, obtaining a data update delay coefficient according to a timestamp includes:

[0059] For each type of user health data, obtain the actual timestamp when it is uploaded to the intelligent diagnosis system within the preset time window, and obtain a timestamp sequence based on time order;

[0060] Calculate the time interval between every two adjacent timestamps in the timestamp sequence, and sequence them to obtain the upload interval sequence; mark the upload interval in the upload interval sequence as E 实 u, u represents the sequence number of the upload interval in the upload interval sequence, u=1, 2, 3, 4, ..., p, p is the total number of upload intervals in the upload interval sequence and p is a positive integer;

[0061] Set the upload interval E in the upload interval sequence 实 u and within the preset upload interval range (T 快 , T 慢 ) is re-marked as Q w , w represents E 实 u Within the preset data upload interval range (T 快 , T 慢 ) is a sequence number of the time outside the time interval, w = 0, 1, 2, 3, 4, ..., d, d is a positive integer; the data update delay coefficient AG is calculated using the following formula: f is the type number of the user health data, m is the total number of types of the user health data, and m is a positive integer.

[0062] It should be noted that the preset upload interval range is set by professionals based on experience and data analysis according to the actual application scenario and the user's health monitoring needs. This range usually takes into account the collection frequency of different types of health data and the user's daily activity patterns. Setting a reasonable upload interval range helps ensure the timeliness and integrity of the data, while avoiding data loss or inconsistency due to too frequent or too long uploads. Through this customized upload interval setting, the system can more accurately evaluate the normality of data updates and promptly identify abnormal upload behavior, thereby ensuring the accuracy and reliability of intelligent diagnostic results.

[0063] It should be noted that the data update delay coefficient is a metric used to measure the degree of delay in uploading user health data. It assesses the timeliness of data updates by analyzing the intervals between health data uploads, particularly those that exceed the preset upload interval. When the upload intervals are excessively long or irregular, the data update delay coefficient increases, indicating possible delays or anomalies in the data collection and upload process. This delay typically reflects the system's inability to obtain real-time health information, which may be caused by device failure, network issues, or the user not wearing the device on time. A higher data update delay coefficient indicates poorer data update frequency and timeliness, resulting in the system being unable to obtain the user's latest health status in real time. This delay may cause the intelligent diagnostic system to rely on outdated or incomplete data when performing health analysis, thereby affecting the accuracy of diagnostic results. In particular, in health monitoring scenarios where timeliness is crucial (such as heart disease monitoring and blood sugar monitoring), data delays may prevent the system from responding to sudden health issues in a timely manner, increasing the risk of diagnostic errors. Therefore, a higher delay coefficient generally indicates lower data authenticity and a higher probability of diagnostic errors, potentially misleading users into making inappropriate health decisions.

[0064] In one implementation, analyzing the data update delay coefficient for determining the authenticity of user health data provides a means of quantifying the regularity of data uploads. By monitoring fluctuations in the intervals between health data uploads, the system can promptly identify abnormal upload patterns, such as excessively long or short upload intervals, or frequent fluctuations. These can signal potential problems in data collection or transmission. For example, if the intervals between health data uploads exhibit unusually large fluctuations, this could indicate device failure, network instability, or inconsistent user wearing behavior. These issues can lead to data distortion and affect the accuracy of intelligent diagnosis. By calculating the fluctuation anomaly coefficient, the system can identify these abnormal fluctuations, providing early warning of potential data authenticity issues and preventing erroneous health judgments based on unreliable data. Furthermore, this coefficient helps the system distinguish between data within the normal fluctuation range and abnormalities likely caused by external factors or device issues, enabling it to take appropriate remedial measures, such as recollecting data or notifying the user to inspect their device, ensuring a more accurate and reliable diagnosis.

[0065] In one embodiment, obtaining the number of user health data uploads of each type within a preset time window and calculating the data upload incompleteness coefficient based on the number of user health data uploads of each type includes:

[0066] Get the actual number of health data uploaded by each type of user and the corresponding preset health data upload number within the preset time window, and mark the actual data upload number and the corresponding preset data upload number as G respectively. 实 f and G 预 f , f is the type number of the user health data, m is the total number of types of user health data, and m is a positive integer;

[0067] According to G 实 f and G 预 f The data upload incomplete coefficient DF is obtained, and the calculation formula is:

[0068] It should be noted that the preset data upload quantity is set by professionals based on actual circumstances and is not limited or elaborated on in detail.

[0069] It should be noted that the data upload incompleteness coefficient is an indicator used to measure the difference between the actual amount of user health data uploaded within a preset time window and the preset upload amount. Specifically, it reflects whether data uploads are occurring at the expected frequency. When the actual upload amount is less than the preset upload amount, the data upload incompleteness coefficient increases, indicating that health data uploaded within that time window was missing or untimely. A larger data upload incompleteness coefficient indicates that the health data received by the system is incomplete or discontinuous, which may affect the timeliness and accuracy of the data. This directly affects the judgment capabilities of the intelligent diagnosis system, as the system relies on this health data for health analysis and prediction. If the uploaded data is incomplete, the intelligent diagnosis system may not obtain sufficient and accurate health information, resulting in incorrect judgments. For example, if a certain type of health data (such as blood pressure or heart rate) is missing or the interval is too long, the system may not be able to detect health changes in a timely manner, missing early warning opportunities, or even leading to misdiagnosis. Therefore, a larger data upload incompleteness coefficient indicates a lower accuracy and reliability of intelligent diagnosis, and a greater risk of incorrect diagnosis results and health recommendations, which may mislead users into making inappropriate health decisions.

[0070] In one implementation, analyzing the data upload incompleteness coefficient for determining the authenticity of user health data provides the following benefits: it helps the system promptly identify missing or irregular data uploads, thereby ensuring the integrity and representativeness of the health data underlying the intelligent diagnostic system. By calculating this coefficient, the system can effectively detect which health data is being uploaded infrequently or interrupted, and thus determine whether this data truly reflects the user's health status. A larger data upload incompleteness coefficient indicates a more severe degree of data loss, potentially leading to inaccurate or untimely diagnostic results. For example, for some important health indicators (such as heart rate and blood sugar), incomplete uploads may miss important health changes, impacting the system's early warning and decision-making capabilities. By analyzing and monitoring this coefficient, the system can proactively identify potential health data issues and take timely remedial measures (such as reminding users to upload data or resynchronizing devices), ensuring that the intelligent diagnostic system relies on accurate and comprehensive data when making health assessments. This improves the accuracy and reliability of diagnoses and reduces the risk of misleading users.

[0071] In one embodiment, obtaining a data false coefficient based on the data update fluctuation anomaly coefficient, the data update delay coefficient, and the data upload incomplete coefficient includes:

[0072]

[0073] Where Dax is the data false coefficient, YK, AG, and DF are the data update fluctuation anomaly coefficient, data update delay coefficient, and data upload incomplete coefficient, respectively. α, β, and γ are the preset proportional coefficients of YK, AG, and DF, respectively, and α, β, and γ are all greater than 0.

[0074] In one embodiment, dividing the true state of the user's health data into a false state and a temporary true state in combination with a preset data false coefficient threshold includes:

[0075] Compare the data false coefficient with the preset data false coefficient threshold. If the data false coefficient is not less than the preset data false coefficient threshold, it indicates that the true state of the user's health data is false, and issue a warning signal;

[0076] The data false coefficient is compared with the preset data false coefficient threshold. If the data false coefficient is less than the preset data false coefficient threshold, it means that the user health data is temporarily true and needs further analysis.

[0077] It should be noted that the preset data false coefficient threshold is set by professionals based on actual circumstances and is not limited or elaborated on in detail.

[0078] It should be noted that if the data falsity coefficient is at least a preset threshold, this indicates significant discrepancies or issues with the health data, potentially resulting in inaccurate or incomplete health data due to equipment failure, data loss, or upload delays. The system will issue an early warning signal, alerting the user or healthcare administrator that the data may be falsified and requires immediate verification or correction. In this case, the system may cease relying on this data for diagnosis or decision-making to avoid misleading users into inappropriate health management practices. If the data falsity coefficient is less than a preset threshold, this means the user's health data is temporarily authentic, meaning the integrity and timeliness of the uploaded data are within reasonable limits. However, some uncertainty remains, requiring further analysis. The system will continue to monitor data updates, collect more data, and further verify the data's authenticity through methods such as time series analysis. If the data falsity coefficient remains consistently low, the data can be considered increasingly authentic, and the system can rely on this data for more accurate health assessments and diagnoses.

[0079] In one implementation, this division method based on data false coefficients and thresholds can provide a real-time health data quality control mechanism to ensure that the judgments made by the intelligent diagnosis system are based on reliable and accurate data. When the system detects that the data false coefficient is too high, it can effectively avoid misdiagnosis and protect users from health misleading caused by inaccurate data. At the same time, when the data is judged to be temporarily true, the system can continue to perform detailed analysis of the data to improve the accuracy of health monitoring. Through this dynamic evaluation and timely early warning mechanism, users can not only obtain more reliable health diagnoses, but also take timely measures when potential problems arise, such as checking equipment, re-uploading data, etc., thereby improving the overall effect of health management and the reliability of the intelligent diagnosis system.

[0080] In one embodiment, determining the final true state of the user's health data based on the data false coefficient sequence includes:

[0081] Calculate the standard deviation and mean of the data false coefficient sequence, and compare the standard deviation and mean of the data false coefficient sequence with the preset standard deviation and preset mean respectively;

[0082] If the standard deviation of the data false coefficient sequence is less than the preset standard deviation and the mean is less than the preset mean, it means that the final true state of the user health data is true; otherwise, it means that the final true state of the user health data is a false state.

[0083] It should be noted that the preset standard deviation and preset mean are set by professionals based on actual conditions and are not specifically limited or elaborated.

[0084] It should be noted that the mean of the data false coefficient sequence represents the overall reliability of the health data within the time window. If the mean is small, it means that the health data is relatively true as a whole; if the mean is large, it may indicate that the data is generally inaccurate; the standard deviation reflects the degree of fluctuation of the data false coefficient. If the standard deviation is small, it means that the data changes relatively smoothly and the fluctuation is small, indicating that the data quality is relatively stable; if the standard deviation is large, it may indicate that there are large inconsistencies in the update and upload of health data, and the fluctuation is large, which may affect the authenticity of the data. Compare the calculated standard deviation and mean of the data false coefficient sequence with the preset standard deviation and mean: if the standard deviation of the data false coefficient sequence is smaller than the preset standard deviation, it means that the data fluctuation is small and the stability is strong. The system can assume that the quality of the health data is high and the credibility is strong; if the mean of the data false coefficient sequence is smaller than the preset mean, it means that the falsity of the data is low. The system can infer that the data is closer to the truth and the data accuracy is high. Through the above comparison, if the standard deviation and mean of the data false coefficient sequence meet the preset standards (i.e., the standard deviation is less than the preset standard deviation and the mean is less than the preset mean), the system considers that the final true state of the user health data of the segment is true, indicating that the data is credible, and intelligent diagnosis can rely on this data for health analysis and decision-making. Conversely, if the standard deviation and mean of the data false coefficient do not meet the preset standards (i.e., the standard deviation is greater than the preset standard deviation or the mean is greater than the preset mean), the system determines that the health data is false, indicating that the accuracy and credibility of the data are low, and the intelligent diagnosis results may have large deviations, requiring further verification or correction measures.

[0085] In one implementation, this method of judging the final true state of health data by the standard deviation and mean of the data's false coefficient sequence has great advantages. It not only helps the system monitor the stability and credibility of the data in real time, but also evaluates its authenticity based on the long-term trend of the data, rather than relying on a single data verification. By dynamically adjusting the preset thresholds of the standard deviation and mean, the system can flexibly perform data quality control based on the characteristics of health data for different users and in different environments. If the system detects false data, it can quickly issue a warning and provide solutions or optimization measures to ensure the accuracy of health monitoring and diagnosis. Ultimately, this approach improves the reliability of health data and the accuracy of intelligent diagnosis, reduces health misdiagnosis and decision-making errors caused by data problems, and enhances user trust and security in the system.

[0086] The above is a detailed description of an embodiment of the present invention. However, the content is only a preferred embodiment of the present invention and should not be used to artificially limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. An AI health monitoring and management system based on intelligent diagnosis, characterized in that: The system comprises: Update module: For each type of user health data, obtain the timestamp of its upload to the intelligent diagnosis system within the preset time window, and obtain the data update fluctuation anomaly coefficient and data update delay coefficient based on the timestamp; Incomplete module: obtains the number of user health data uploads of each type within a preset time window, and calculates the data upload incomplete coefficient based on the number of user health data uploads of each type; False coefficient module: derives the data false coefficient based on the data update fluctuation anomaly coefficient, data update delay coefficient, and data upload incomplete coefficient; and divides the true state of the user's health data into a false state and a temporary true state based on the preset data false coefficient threshold; Final judgment module: For the user health data in the temporary real state, continuously obtain the data false coefficient, and sequence it according to time to obtain the data false coefficient sequence, and judge the final real state of the user health data based on the data false coefficient sequence; The data update fluctuation anomaly coefficients obtained based on the timestamp include: For each type of user health data, obtain the actual timestamp when it is uploaded to the intelligent diagnosis system within the preset time window, and obtain a timestamp sequence based on time order; Calculate the time interval between every two adjacent timestamps in the timestamp sequence, and sequence them to obtain the upload interval sequence; The upload interval in the upload interval sequence is marked as E 实 u , u represents the sequence number of the upload interval in the upload interval sequence, u=1, 2, 3, 4, ..., p, p is the total number of upload intervals in the upload interval sequence and p is a positive integer; Calculate the mean of the upload interval sequence using the following formula: Calculate the data update fluctuation anomaly coefficient using the following formula: Where YK is the data update fluctuation anomaly coefficient, f is the type number of the user health data, m is the total number of types of user health data, and m is a positive integer; The data update delay coefficient obtained based on the timestamp includes: For each type of user health data, obtain the actual timestamp when it is uploaded to the intelligent diagnosis system within the preset time window, and obtain a timestamp sequence based on time order; Calculate the time interval between every two adjacent timestamps in the timestamp sequence, and sequence them to obtain the upload interval sequence; mark the upload interval in the upload interval sequence as E 实 u , u represents the sequence number of the upload interval in the upload interval sequence, u=1, 2, 3, 4, ..., p, p is the total number of upload intervals in the upload interval sequence and p is a positive integer; Set the upload interval E in the upload interval sequence 实 u and within the preset upload interval range (T 快 , T 慢 ) is re-marked as Q w , w represents E 实 u Within the preset data upload interval range (T 快 , T 慢 ) is a sequence number of the time outside the time interval, w = 0, 1, 2, 3, 4, ..., d, d is a positive integer; the data update delay coefficient AG is calculated using the following formula: f is the type number of the user health data, m is the total number of types of the user health data, and m is a positive integer; The data upload incomplete coefficient is calculated based on the number of uploaded user health data of each type, including: Get the actual number of health data uploaded by each type of user and the corresponding preset health data upload number within the preset time window, and mark the actual data upload number and the corresponding preset data upload number as G respectively. 实 f and G 预 f , f is the type number of the user health data, m is the total number of types of user health data, and m is a positive integer; According to G 实 f and G 预 f The data upload incomplete coefficient DF is obtained, and the calculation formula is: The data false coefficients obtained based on the data update fluctuation anomaly coefficient, data update delay coefficient, and data upload incomplete coefficient include: Where Dax is the data false coefficient, YK, AG, and DF are the data update fluctuation anomaly coefficient, data update delay coefficient, and data upload incomplete coefficient, respectively. α, β, and γ are the preset proportional coefficients of YK, AG, and DF, respectively, and α, β, and γ are all greater than 0.

2. The AI health monitoring and management system based on intelligent diagnosis according to claim 1 is characterized in that: The true state of the user's health data is divided into a false state and a temporary true state in combination with a preset data false coefficient threshold, including: Compare the data false coefficient with the preset data false coefficient threshold. If the data false coefficient is not less than the preset data false coefficient threshold, it indicates that the true state of the user's health data is false, and issue a warning signal; The data false coefficient is compared with the preset data false coefficient threshold. If the data false coefficient is less than the preset data false coefficient threshold, it means that the user health data is temporarily true and needs further analysis.

3. The AI health monitoring and management system based on intelligent diagnosis according to claim 1 is characterized in that: Judging the final true state of user health data based on the data false coefficient sequence includes: Calculate the standard deviation and mean of the data false coefficient sequence, and compare the standard deviation and mean of the data false coefficient sequence with the preset standard deviation and preset mean respectively; If the standard deviation of the data false coefficient sequence is less than the preset standard deviation and the mean is less than the preset mean, it means that the final true state of the user health data is true; otherwise, it means that the final true state of the user health data is a false state.

Images