Personal health early warning method and device, medium and product

Abstract

The invention relates to the technical field of health monitoring, in particular to a personal health early warning method and device, a medium and a product. The method comprises the following steps: generating a series of medication prompt information containing timestamps based on diagnosis and treatment records and individual attribute information of a target individual, and sending the medication prompt information to the target individual at the corresponding timestamp through a medicine chest module; obtaining an ideal index range corresponding to the target individual, group index data of patients with the same disease and time sequence index data of the target individual in a medication period; extracting the medication information of the target individual from the diagnosis and treatment record, identifying an abnormal index set from the time sequence index data based on the medication information, the ideal index range and the group index data, and predicting the disease risk of the target individual; and calling the large model to generate a health assessment report for the target individual based on the medication information, the disease risk, the abnormal index set and the time sequence index data. According to the invention, precise health early warning for individuals during medication can be realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of health monitoring, in particular to a personal health early warning method, device, medium and product. BACKGROUND

[0002] With the increasing demand for home health management, personal health monitoring and early warning have gradually become an important direction in the field of health management. Related technologies connect blood pressure meters, blood glucose meters, uric acid meters and other health monitoring devices through a smart medical system, realize data transmission between devices and the system through Bluetooth BLE protocol, and upload data to the background through 4G module and MQTT protocol to complete drug reminder, health data display and other functions, providing basic health management support for nursing homes or homebound elderly. However, the prior art can only realize the collection and simple display of health data, and the basic function of drug reminder, and cannot combine the personalized diagnosis and treatment records, age, gender and other multi-dimensional information of the target individual to deeply analyze the index data during the medication period of the individual, resulting in a lack of forward-looking early warning ability for individual health risks in health management, and difficulty in meeting the needs of users for personalized and precise health management. SUMMARY

[0003] In order to solve the problem that the prior art cannot realize health early warning for individuals during medication, the present application provides a personal health early warning method, device, medium and product.

[0004] In a first aspect, the present application provides a personal health early warning method, which adopts the following technical solution: A personal health early warning method, comprising: Obtaining diagnosis and treatment records and individual attribute information of a target individual, generating a series of medication prompt information containing time stamps based on the diagnosis and treatment records and the individual attribute information, and sending the medication prompt information to the target individual at the corresponding time stamp through a medicine box module; Obtaining ideal index ranges corresponding to the target individual, group index data of patients with the same disease, and time series index data of the target individual during medication; Extracting medication information of the target individual from the diagnosis and treatment records, identifying an abnormal index set from the time series index data based on the medication information, the ideal index range and the group index data, and predicting the risk of the target individual getting sick based on the abnormal index set; Based on the medication information, the risk of getting sick, the abnormal index set and the time series index data, calling a large model to generate a health assessment report for the target individual.

[0005] By adopting the technical scheme, the target individual diagnosis and treatment record and individual attribute information are used to generate a timestamped medication prompt which is accurately pushed through the medicine box module, so as to ensure that the medication person takes medicine on time and in the right amount, reduce missed and wrong medication, lay a foundation for subsequent health monitoring and early warning, and build a multi-dimensional data support system by synchronously acquiring ideal index range, same-disease patient group index data and time series index data during medication, so as to avoid evaluation deviation caused by single data. In combination with medication information, ideal index range and group index data, an abnormal index set is identified and disease risk is predicted, so as to accurately locate health abnormalities from multi-source data correlation analysis and discover potential disease risk in advance. Finally, a health evaluation report is generated by calling a large model to integrate multiple types of key data, personalized content including risk warning, medication suggestion and index interpretation can be output, a comprehensive and instructive health management scheme is provided for the user, and the problem that traditional health monitoring can only collect data but cannot perform deep analysis and accurate early warning is solved.

[0006] In a preferred example, the application can be further configured to: acquire ideal index range corresponding to the target individual and group index data of same-disease patients, including: extract age information and gender information from the individual attribute information, and call ideal index range matched with the age information and the gender information from a medical database; determine disease information of the target individual from the diagnosis and treatment record, the disease information including disease type and disease degree; retrieve index data from multiple same-disease patients matched with the disease information and the individual attribute information from the medical database, the index data of the multiple same-disease patients constituting the group index data.

[0007] By adopting the technical scheme, age and gender are extracted from individual attribute information to call matched ideal index range, so that the ideal index is fitted to individual physiological characteristics, abnormal judgment errors caused by using general range are avoided, and the accuracy of index abnormality identification is improved. Disease type and disease degree are determined from the diagnosis and treatment record, and then matched same-disease patient group index data is retrieved from the medical database in combination with individual attribute information, so as to ensure that the group index data is highly adapted to the health status of the target individual, provide a group benchmark with higher reference value for subsequent abnormal index identification, solve the problem of single dimension in traditional group data screening and low matching degree with individuals, and further guarantee the reliability of abnormal index identification and risk prediction.

[0008] In a preferred example, the application can be further configured to: the abnormal index set is identified from the time series index data based on the medication information, the ideal index range and the group index data, including: screening current index data of a target period before the current time from the timing index data, comparing the current index data with the ideal index range one by one, and screening an initial abnormal index set; calculating a group index statistical value based on the group index data, comparing the group index statistical value with the ideal index range, screening a common abnormal index, and determining a common expected range of the common abnormal index in a diseased state; based on the medication information, identifying a drug-induced influence index and a normal fluctuation range of the drug-induced influence index; labeling an index type for each index in the initial abnormal index set, the index type including: a common abnormal index, a common abnormal index, a drug-induced influence index, and a common drug source superposition index; based on the ideal index range and the normal fluctuation range, determining a single expected range of the drug-induced influence index, and based on the common expected range and the normal fluctuation range, determining a superimposed expected range of the drug-induced influence index; based on the common expected range, the single expected range and the superimposed expected range, screening an abnormal index set from the initial abnormal index set.

[0009] By adopting the above technical solution, the current index data of the target period is compared with the ideal index range to obtain the initial abnormal index set, the recent data is focused to reduce the interference of historical long-term data, and the potential abnormality is preliminarily locked; the group index statistical value is calculated to determine the common expected range, the drug-induced influence index and the normal fluctuation range are identified based on the medication information, and the group and drug source dimensions are provided as the judgment basis for abnormal index classification; the initial abnormal index is labeled and the single and superimposed expected ranges are determined, which can accurately distinguish different causes of abnormal indexes; finally, the abnormal index set is screened based on multiple expected ranges, which can eliminate accidental errors and false abnormalities caused by non-disease or non-drug source, significantly improve the accuracy of abnormal index identification, and provide high-quality data support for subsequent disease risk prediction.

[0010] In a preferred example, the application can be further configured to: the disease risk of the target individual is predicted based on the abnormal index set, comprising: calling a medical knowledge base, the medical knowledge base pre-storing associated indexes corresponding to various diseases; matching the abnormal index set with the associated indexes to determine a candidate disease set; based on the diagnosis weight of the associated indexes of various diseases in the abnormal index set and the candidate disease set, calculating the risk probability of the target individual suffering from each disease type in the candidate disease set; determine a disease type in the candidate disease set whose risk probability exceeds a preset risk probability threshold as a suspected disease, and the suspected disease and the risk probability of the suspected disease constitute the disease risk.

[0011] By adopting the technical solution, the medical knowledge base of the pre-stored disease correlation indicators is called to provide an authoritative medical basis for matching of the abnormal indicators and diseases, and subjective judgment deviation is avoided. The matching of the abnormal indicator set and the correlation indicators determines the candidate disease set, quickly narrows the range of possible diseases, and improves the risk prediction efficiency. The disease risk probability is calculated in combination with the correlation indicator diagnosis weight, the likelihood of different diseases is quantified, and all potential diseases are avoided to be treated equally. The diseases whose risk probability exceeds the threshold are determined as the suspected diseases to constitute the disease risk, the health risk that needs to be focused on can be determined, the user and the medical personnel can be helped to preferentially deal with the high-risk diseases, and the problems of traditional risk prediction ambiguity and no priority are solved.

[0012] In a preferred example, the application can be further configured to calculate the risk probability of the target individual suffering from each disease type in the candidate disease set based on the diagnosis weight of each disease type in the abnormal indicator set and the candidate disease set, including: obtain the correlation indicator set of the target disease type and the basic diagnosis weight of each correlation indicator from the medical knowledge base, the target disease type being any type in the candidate disease set; determine intersection indicators of the abnormal indicator set and the correlation indicator set, and calculate an indicator proportion of the intersection indicators in the correlation indicator set; for each intersection indicator, calculate an abnormal degree of the intersection indicator, determine an evidence strengthening coefficient based on the abnormal degree, and adjust the basic diagnosis weight of the intersection indicator to obtain a diagnosis weight based on the evidence strengthening coefficient; obtain a comprehensive diagnosis weight of the target disease type based on the diagnosis weight of each intersection indicator; obtain the risk probability of the target disease type by combining the indicator proportion and the comprehensive diagnosis weight.

[0013] By adopting the technical solution, the target disease correlation index set and the basic diagnosis weight are obtained to provide standardized medical parameters for risk probability calculation; the intersection index is determined and the index proportion is calculated to reflect the coverage degree of the abnormal index on disease diagnosis, and the higher the proportion is, the stronger the support of disease diagnosis is; the evidence strengthening coefficient is determined according to the abnormal degree of the intersection index to adjust the diagnosis weight, so that the weight is more suitable for the actual abnormal situation of the individual index, and the quantitative deviation caused by the fixed weight is avoided; the risk probability is obtained by comprehensively considering the index proportion and the comprehensive diagnosis weight, the refinement and personalization of risk probability calculation are realized, compared with the simple weighted calculation, the real risk of the individual suffering from the target disease can be more accurately reflected, and the scientificity and credibility of the prediction of the risk of suffering from the disease are further improved.

[0014] In a preferred example, the application can be further configured to: based on the medication information, the risk of suffering from the disease, the set of abnormal indexes and the time series index data, calling a large model to generate a health assessment report for the target individual, including: analyzing the time series index data by using a moving average analysis method to determine the index change trend; substituting the medication information, the risk of suffering from the disease, the set of abnormal indexes and the index change trend into a preset prompt word template to integrate into a structured prompt word; inputting the structured prompt word into the large model, and receiving the health assessment report for the target individual generated by the large model, the health assessment report including disease risk warning, medication adjustment suggestion and index interpretation report.

[0015] By adopting the technical solution, the time series index data is analyzed by using a moving average analysis method to determine the change trend, which can smooth the short-term data fluctuation, clearly present the long-term change rule of the index, and help identify potential health deterioration or improvement trend; the medication information, the risk of suffering from the disease, the set of abnormal indexes and the index change trend are integrated into a structured prompt word to ensure that the large model obtains comprehensive and orderly input data, and avoid the decline of report quality caused by information loss or confusion; the health assessment report including disease risk warning, medication adjustment suggestion and index interpretation report is generated by inputting the large model, which can convert multi-dimensional data into health guidance content that is easy to understand and practical, so that the target individual can clearly understand their own health status, risk points and countermeasures, solve the problem that the traditional data presentation is professional and obscure, and the user is difficult to understand and apply, and improve the practicality of health management.

[0016] In a second aspect, the application provides an electronic device, which adopts the following technical solution: at least one processor; a memory; at least one application program, wherein the at least one application program is stored in the memory and configured to be executed by the at least one processor, the at least one application program being configured to perform the personal health early warning method according to any one of the first aspect.

[0017] In a third aspect, the present application provides a computer-readable storage medium, which adopts the technical scheme as follows: A computer-readable storage medium, which stores a computer program, when the computer program is executed in a computer, the computer is caused to perform the personal health early warning method according to any one of the first aspect.

[0018] In a fourth aspect, the present application provides a computer program product, which adopts the technical scheme as follows: A computer program product, comprising a computer program, when the computer program is executed by a processor, the personal health early warning method according to any one of the first aspect is realized.

[0019] In summary, the present application has the following beneficial technical effects: The present application can ensure that the medication is taken on time and in the right amount, reduce missed or wrong medication, and lay a foundation for subsequent health monitoring and early warning by generating a timestamped medication prompt based on the target individual's diagnosis and treatment records and individual attribute information and accurately pushing it through the medicine box module; The ideal index range, patient group index data, and time series index data during medication are obtained synchronously to build a multi-dimensional data support system and avoid evaluation bias caused by single data; The abnormal index set is identified and the risk of disease is predicted by combining medication information, ideal index range, and group index data, which can accurately locate health abnormalities from multi-source data correlation analysis and discover potential disease risks in advance; Finally, a health assessment report is generated by calling a large model to integrate multiple types of key data, which can output personalized content including risk early warning, medication suggestions, and index interpretation to provide users with comprehensive and instructive health management solutions, solving the problem that traditional health monitoring can only collect data but cannot perform in-depth analysis and accurate early warning. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 is a flowchart of a personal health early warning method provided by an embodiment of the present application; Figure 2 is a structural schematic diagram of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0021] The following will be described in detail with reference to the accompanying drawings. Figure 1 to the accompanying drawings Figure 2 The present application will be further described in detail.

[0022] The specific embodiments are merely explanatory of this application, and are not intended to limit this application, and those skilled in the art can make modifications to the embodiments without creative contribution after reading the specification, as long as the modifications are within the scope of the claims of this application.

[0023] To make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative contribution fall within the scope of protection of the present application.

[0024] In addition, the term "and / or" herein merely describes the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B can represent the three cases of A alone, A and B together, and B alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects, unless otherwise specified.

[0025] It should be noted that in the optional embodiments of the present application, the object information and other related data involved in the embodiments of the present application when applied to specific products or technologies need to obtain the permission or consent of the object, and the collection, use and processing of related data need to comply with relevant laws, regulations and standards of countries and regions. That is, if the embodiments of the present application involve data related to the object, the data needs to be obtained with the authorization and consent of the object, the authorization and consent of the relevant department, and in compliance with the relevant laws, regulations and standards of the country and region. If the embodiments involve personal information, the consent of the individual is required for obtaining all personal information, and the individual's separate consent is required for sensitive information. The embodiments also need to be implemented with the authorization and consent of the object.

[0026] The embodiments of the present application provide a personal health warning method, as shown in Figure 1 The method provided in the embodiments of the present application is executed by an electronic device, which can be a server or a terminal device. The server can be a physical server, a server cluster composed of multiple physical servers, or a distributed system, and can also be a cloud server providing cloud computing services. The terminal device can be a smart phone, a tablet computer, a notebook computer, a desktop computer, etc., but is not limited thereto. The terminal device and the server can be directly or indirectly connected through wired or wireless communication, and the embodiments of the present application do not limit this. The method includes steps S101-S104, wherein: S101, obtain the diagnosis and treatment records and individual attribute information of the target individual, generate a series of medication prompt information containing timestamps based on the diagnosis and treatment records and individual attribute information, and send the medication prompt information to the target individual at the corresponding timestamps through the medicine box module.

[0027] Specifically, the diagnosis and treatment records include structured electronic data of the individual in the medical institution, including diagnosis results and medication information, wherein the diagnosis results include disease information, and the medication information includes drug types, medication amounts, and medication frequencies. The individual attribute information includes age information, gender information, and individual daily habit information. The individual daily habit information can be input through an electronic device and includes meal times for breakfast, lunch, and dinner, and a series of timestamps are generated in combination with the individual attribute information. The medicine box module can be a smart hardware device, which is used to connect various health monitoring devices through Bluetooth and forward the time series index data of the target individual collected by the health monitoring devices to the electronic device using the MQTT protocol. The health monitoring devices can include a sphygmomanometer, a blood glucose meter, a uric acid meter, a blood oxygen meter, and an electrocardiograph. The target individual regularly collects physical index data through the health monitoring devices, and the collected data is arranged in time to obtain the time series index data.

[0028] The medication information is extracted from the diagnosis and treatment information. If the medication information includes medication amounts and medication frequencies, a series of medication prompt information is generated in combination with the individual daily habit information. If the medication information only includes drug types, the medication amounts and medication frequencies matched with the age group are determined based on the individual attribute information, and a series of medication prompt information is generated in combination with the individual daily habit information. For example, the medication amount is three tablets at a time, three times a day, and taken after meals. Based on the individual daily habit information, three meal times are determined, and medication prompt information with corresponding timestamps is generated.

[0029] S102, obtain the ideal index range corresponding to the target individual, the group index data of patients with the same disease, and the time series index data of the target individual during medication.

[0030] Specifically, the ideal index range represents a healthy physiological index interval; the group index data represents a set of physiological index monitoring data of a plurality of patients with the same disease type and the same disease severity as the target individual, and the individual index is compared and analyzed to determine whether there is a common abnormality related to the disease. The time series index data refers to the physiological index data collected by the health monitoring devices (sphygmomanometer, blood glucose meter, uric acid meter, blood oxygen meter, electrocardiograph) in time sequence during medication, each data including: collection time, index type, and index value, such as 2025-11-09 08:00, systolic pressure 125 mmHg; 2025-11-09 08:00, diastolic pressure 80 mmHg.

[0031] S103, extract medication information of the target individual from the medical record, identify an abnormal index set from the time series index data based on the medication information, ideal index range and population index data, and predict the risk of the target individual based on the abnormal index set.

[0032] The medication information refers to the medication details of the target individual extracted from the medical record, including: drug name, drug type, medication amount, medication frequency, and medication side effects. The extraction method of medication information can be a text analysis algorithm, such as OCR recognition and keyword matching. The risk of illness refers to the type of disease that the target individual may suffer from and the corresponding probability predicted based on the abnormal index combined with the medical knowledge base.

[0033] Specifically, the process of identifying the abnormal index set includes: comparing the time series index data and the ideal index range to obtain an initial abnormal index set. Compare the common abnormal index and the ideal index range to filter the common abnormal index, and determine the common expected range of the common abnormal index under the disease state. Based on the medication information, identify the drug-induced influence index and the normal fluctuation range of the drug-induced influence index. Label each index in the initial abnormal index set with an index type, including: normal abnormal index, common abnormal index, drug-induced influence index, and common drug source superposition index. Based on the common expected range, the ideal index, and the normal fluctuation range, filter the abnormal index set from the initial abnormal index set.

[0034] Call the medical knowledge base, construct a candidate disease set based on the medical knowledge base and the abnormal index set, and calculate the risk probability of each disease type in the candidate disease set. Determine the disease type in the candidate disease set whose risk probability exceeds the preset risk probability threshold as the suspected disease, and the suspected disease and the risk probability of the suspected disease constitute the risk of illness. The risk probability threshold is a preset value, which can be adjusted based on the severity of the disease, such as a cardiovascular and cerebrovascular disease threshold of 40% and a general chronic disease threshold of 60%.

[0035] S104, based on the medication information, the risk of illness, the abnormal index set, and the time series index data, call the large model to generate a health assessment report for the target individual.

[0036] Specifically, the moving average analysis method is used to analyze the time series index data to determine the index change trend. The medication information, risk of illness, abnormal index, and index change trend are substituted into the preset prompt word template to integrate into a structured prompt word. The structured prompt word is input into the large model, and the health assessment report of the target individual is received, including disease risk warning, medication adjustment suggestion, and index interpretation report.

[0037] The embodiment generates a timestamped medication prompt by obtaining the diagnosis and treatment records of the target individual and the individual attribute information, and accurately pushes the medication prompt through the medicine box module, which can ensure that the medication person takes the medication on time and in the right amount, reduces missed or wrong medication, and lays a foundation for subsequent health monitoring and early warning; simultaneously obtains ideal index range, patient group index data, and time series index data during medication, constructs a multi-dimensional data support system, and avoids evaluation bias caused by single data; in combination with medication information, ideal index range, and group index data, an abnormal index set is identified and disease risk is predicted, which can accurately locate health abnormalities from multi-source data correlation analysis and discover potential disease risks in advance; finally, a health assessment report is generated by calling a large model to integrate multiple types of key data, which can output personalized content including risk warning, medication suggestion, and index interpretation, and provide users with comprehensive and instructive health management solutions, solving the problem that traditional health monitoring can only collect data but cannot perform in-depth analysis and accurate early warning.

[0038] In one possible implementation of the embodiment of the present application, the ideal index range corresponding to the target individual and the group index data of patients with the same disease are obtained, including: extracting age information and gender information from the individual attribute information, and retrieving the ideal index range matching the age information and gender information from the medical database; determining the disease information of the target individual from the diagnosis and treatment records, the disease information including the disease type and the disease severity; retrieving the index data from multiple patients with the same disease matching the disease information and the individual attribute information from the medical database, the index data of the multiple patients with the same disease constituting the group index data.

[0039] In the embodiment, the individual attribute information is pre-stored in the electronic device, and the core includes age information and gender information. Because there are differences in physiological skills among different ages and genders, the definition of the normal range of health indicators in the medical guidelines is different. For example, the ideal range of uric acid for adult women is lower than that for adult men. The medical database refers to a structured database built in the electronic device, which integrates medical guidelines and clinical standards. The ideal range of various physiological indicators and the index data of different patients are stored according to the age+gender classification. The index data of the patients can be collected and integrated from the database of the medical institution. The ideal index range is the ideal range of each index extracted from the medical database based on the age and gender of the target individual.

[0040] Retrieving multiple patients with the same disease from the medical database that matches the disease type, disease severity (early, medium, late, mild, moderate, severe), age information (within ±5 years of the target individual's age), and gender, and integrating the index data of the patients with the same disease into group index data.

[0041] The embodiment extracts age and gender from individual attribute information to retrieve a matched ideal index range, so that the ideal index is fitted to the physiological characteristics of the individual, abnormal judgment errors caused by the use of a general range are avoided, and the accuracy of index abnormality identification is improved; the disease type and disease severity are determined from the diagnosis and treatment records, and then the individual attribute information is combined to retrieve matched index data of the same disease patient group from the medical database, so as to ensure that the group index data is highly adapted to the health status of the target individual, provide a group benchmark with more reference value for subsequent abnormal index identification, solve the problem of single screening dimension and low matching degree of traditional group data, and further ensure the reliability of abnormal index identification and risk prediction.

[0042] In one possible implementation of the embodiment of the present application, abnormal index set is identified from time sequence index data based on medication information, ideal index range and group index data, including: The current index data of the target period before the current time is filtered from the time sequence index data, the current index data is compared with the ideal index range one by one, and the initial abnormal index set is filtered out; The group index statistical value is calculated based on the group index data, the group index statistical value is compared with the ideal index range, the common abnormal index is filtered out, and the common expected range of the common abnormal index under the disease state is determined; Based on the medication information, the drug-induced influence index and the normal fluctuation range of the drug-induced influence index are identified; Each index in the initial abnormal index set is labeled with an index type, and the index type includes: normal abnormal index, common abnormal index, drug-induced influence index and common drug source superposition index; Based on the ideal index range and the normal fluctuation range, the single expected range of the drug-induced influence index is determined, and based on the common expected range and the normal fluctuation range, the superposition expected range of the drug-induced influence index is determined; Based on the common expected range, the single expected range and the superposition expected range, the abnormal index set is filtered out from the initial abnormal index set.

[0043] In the embodiment, the target period before the current time is a configurable time window covering multiple index collection times. Optionally, the target period is 3 days or 7 days, and screening the data in the period can focus on the recent health status and avoid historical long-term data interfering with the current abnormality judgment. Each index value in the current index data is compared with the corresponding ideal range. If the value exceeds the ideal range, the data is marked as single abnormal data, and the abnormal direction (exceeding or being lower than the ideal range) is recorded. For the current index data of the same index type, if the proportion of single abnormal data exceeds the preset proportion threshold, the index type is included in the abnormal index set. If the proportion does not exceed the preset proportion threshold, it is determined that it is an occasional abnormality and is not included in the abnormal index set to avoid single error. The preset proportion threshold can be set according to actual needs. Optionally, the preset proportion threshold can be 40%, or the corresponding proportion threshold can be set according to the individualization of the index type. All index types that meet the condition that the proportion of single abnormal data exceeds the preset proportion threshold are summarized to obtain an initial abnormal index set.

[0044] The population index statistical value can be a population index average value or a 95% confidence interval, and the 95% confidence interval = average value ± 1.96 x standard deviation, reflecting the index fluctuation range of 95% of patients in the population. The common abnormal index refers to an index whose population index average value exceeds the corresponding ideal index range, or an index whose 95% confidence interval overlaps with the ideal index range, i.e., the value of the 95% confidence interval of the same disease patient population exceeds the ideal index range, indicating that the index abnormality is a common phenomenon under the disease state, such as the systolic blood pressure of patients with mild neck and shoulder syndrome due to pain stress, and the 95% confidence interval of the population is 112-144 mmHg, part of which exceeds the upper limit of the ideal range 139 mmHg. Systolic blood pressure is a common abnormal index. The common expected range refers to the normal fluctuation boundary of the index in the same disease patient population under the disease state, i.e., the 95% confidence interval of the population index statistical value, which is used to judge whether the index abnormality of the target individual belongs to the disease-related population common abnormality.

[0045] The drug-induced influence index refers to a physiological index directly or indirectly affected by the current drug taken by the target individual, such as Xiaoshi Jianwei Tablets, which can cause a slight increase in blood pressure, and the corresponding drug-induced influence index is systolic blood pressure and diastolic blood pressure. Acyclovir ointment has no significant index influence, and has no corresponding drug-induced influence index. A drug index influence database can be constructed in advance and stored in an electronic device, storing the mapping relationship of drug name-drug-induced influence index-influence direction (increase / decrease / no influence) sorted from domestic and foreign drug instructions and clinical research literature.

[0046] extracting a drug type currently being polypharmacy from medication information of the target individual, calling a drug index influence database to determine a drug source influence index corresponding to the drug type polypharmacy of the target individual and a normal fluctuation range of the drug source influence index. For each drug source influence index, it is determined whether the drug source influence index is a common drug source superposition index (both a common index and a drug source influence index): if so, superimposing the common expected range and the normal fluctuation range to obtain a superimposed expected range, the lower limit of the superimposed expected range = the lower limit of the common expected range - the lower limit of the normal fluctuation range, the upper limit of the superimposed expected range = the upper limit of the common expected range + the upper limit of the normal fluctuation range; if the drug source influence index is not a common abnormal index, superimposing the ideal index range and the normal fluctuation range to obtain a single expected range, the lower limit of the single expected range = the lower limit of the ideal index range - the lower limit of the normal fluctuation range, the single expected range = the upper limit of the ideal index range + the upper limit of the normal fluctuation range.

[0047] For abnormal indexes that are neither common indexes nor drug source influence indexes, mark them as ordinary abnormal indexes in the initial abnormal index set; mark indexes that are only common abnormal indexes and non-drug source influence indexes as common abnormal indexes; mark indexes that are only drug source influence indexes and non-common abnormal indexes as drug source influence indexes; mark indexes that are both common abnormal indexes and drug source influence indexes as common drug source superposition indexes.

[0048] Incorporate the indexes marked as ordinary abnormal indexes in the initial abnormal index set into the abnormal index set. For each common abnormal index in the initial abnormal index set, compare the index value of the target individual with the common expected range of the index: if the index value of the target individual exceeds the common expected range, determine that the index is an abnormal index and incorporate it into the abnormal index set. For each drug source influence index in the initial abnormal index set, compare the index value of the target individual with the single expected range of the index: if the index value of the target individual exceeds the single expected range, determine that the index is an abnormal index and incorporate it into the abnormal index set. For each common drug source superposition index in the initial abnormal index set, compare the index value of the target individual with the superimposed expected range of the index: if the index value of the target individual exceeds the superimposed expected range, determine that the index is an abnormal index and incorporate it into the abnormal index set.

[0049] wherein any range of the common expected range, the single expected range and the superimposed expected range is taken as a target range, any index is taken as a target index, for comparison of the index value of the target index with the target range, determine the total number of data of the target index in the current index data, and the number of data exceeding the target range, calculate the proportion of the number of data exceeding the target range in the total number of data, when the proportion exceeds the preset proportion threshold of the target index, determine that the target index exceeds the corresponding target range.

[0050] The embodiment obtains an initial abnormal index set by screening the current index data of the target period and comparing it with the ideal index range, focuses on recent data to reduce the interference of historical long-term data, and preliminarily locks the potential abnormalities; calculates group index statistics to determine the common expected range, identifies the drug-induced influence index and normal fluctuation range based on the medication information, and provides the judgment basis of the group and drug source dimensions for the classification of abnormal indexes; labels the types of the initial abnormal indexes and determines the single and superimposed expected range, which can accurately distinguish different causes of abnormal indexes; and finally screens the abnormal index set based on the multi-class expected range, which can eliminate accidental errors and false abnormalities caused by non-disease or non-drug sources, significantly improve the accuracy of abnormal index identification, and provide high-quality data support for subsequent disease risk prediction.

[0051] In one possible implementation of the embodiment of the present application, the disease risk of a target individual is predicted based on the abnormal index set, which includes: A medical knowledge base is called, in which the correlation indexes corresponding to various diseases are pre-stored; The abnormal index set is matched with the correlation indexes to determine a candidate disease set; Based on the diagnostic weights of the correlation indexes of various diseases in the abnormal index set and the candidate disease set, the risk probability of the target individual suffering from each disease type in the candidate disease set is calculated; The disease types in the candidate disease set whose risk probability exceeds a preset risk probability threshold are determined as suspected diseases, and the suspected diseases and the risk probability of the suspected diseases constitute the disease risk.

[0052] In the embodiment, the medical knowledge base needs to integrate domestic and foreign authoritative medical guidelines, clinical diagnosis and treatment paths, and medical literature data, store disease types and corresponding correlation indexes, and the correlation indexes are indexes that may be affected by the disease types. The abnormal index set and the correlation indexes are matched, for each abnormal index, all disease types corresponding to the abnormal index in the medical knowledge base are determined, and all disease types corresponding to all abnormal indexes in the abnormal index set are summarized to obtain the candidate disease set. The preset risk probability threshold can be set by medical experts according to actual experience, which is not limited in the embodiment.

[0053] The embodiment calls the pre-stored medical knowledge base of disease correlation indexes, provides authoritative medical basis for matching of abnormal indexes and diseases, and avoids subjective judgment deviation; matches the abnormal index set with the correlation indexes to determine the candidate disease set, quickly narrows down the possible disease range, and improves the risk prediction efficiency; combines the correlation index diagnostic weight to calculate the disease risk probability, which can quantify the likelihood of suffering from different diseases and avoid treating all potential diseases equally; the diseases with risk probability exceeding the threshold are determined as suspected diseases to constitute the disease risk, which can clearly focus on the health risks, help users and medical personnel to prioritize high-risk diseases, and solve the problems of traditional risk prediction ambiguity and no priority.

[0054] In a possible implementation of the embodiment of the application, based on the diagnostic weights of the correlation indicators of each disease type in the abnormal indicator set and the alternative disease set, the risk probability of the target individual suffering from each disease type in the alternative disease set is calculated, including: Obtaining the correlation indicator set of the target disease type and the basic diagnostic weight of each correlation indicator from the medical knowledge base, the target disease type being any type in the alternative disease set; Determining the intersection indicators of the abnormal indicator set and the correlation indicator set, and calculating the indicator proportion of the intersection indicators in the correlation indicator set; For each intersection indicator, the abnormal degree of the intersection indicator is calculated, the evidence strengthening coefficient is determined based on the abnormal degree, and the diagnostic weight of the intersection indicator is adjusted based on the evidence strengthening coefficient to obtain the diagnostic weight; Based on the diagnostic weights of the intersection indicators, the comprehensive diagnostic weight of the target disease type is obtained; The risk probability of the target disease type is obtained by combining the indicator proportion and the comprehensive diagnostic weight.

[0055] In the embodiment, for each correlation indicator of the target disease type, the indicator diagnostic weight is the proportion of the samples with the abnormal indicator in the total samples of the patients suffering from the target disease type. For each intersection indicator, the average value of the intersection indicator in the current indicator data is calculated, and the corresponding range is determined according to the indicator type labeled by the intersection indicator in the abnormal indicator set. The range corresponding to the ordinary abnormal indicator is the ideal indicator range, the common abnormal indicator corresponds to the common expected range, the drug-induced influence indicator corresponds to the single expected range, and the common drug-induced superposition indicator corresponds to the superposition expected range.

[0056] If the average value is within the corresponding range, the evidence strengthening coefficient of the intersection indicator is determined as 1, if not, the difference between the average value and the range is calculated. If the average value exceeds the upper limit of the range, the difference between the average value and the upper limit is calculated, if the average value is lower than the lower limit of the range, the difference between the lower limit and the average value is calculated, and the ratio of the difference value and the average value is calculated as the abnormal degree. The sum of the abnormal degree and 1 is calculated as the evidence strengthening coefficient.

[0057] The product of the basic diagnostic weight and the evidence strengthening coefficient is calculated to obtain the diagnostic weight. The average value of the diagnostic weights of the intersection indicators is calculated as the comprehensive diagnostic weight of the target disease type. Finally, the product of the indicator proportion and the comprehensive diagnostic weight is calculated to obtain the risk probability of the target disease type.

[0058] The embodiment obtains a target disease correlation index set and a basic diagnosis weight to provide standardized medical parameters for risk probability calculation; determines an intersection index and calculates an index proportion to reflect the coverage of abnormal indexes on disease diagnosis, and the higher the proportion, the stronger the support of disease diagnosis; combines the abnormal degree of the intersection index to determine an evidence strengthening coefficient to adjust the diagnosis weight, so that the weight is more suitable for the actual abnormality of individual indexes, and quantitative deviation caused by fixed weight is avoided; the risk probability is obtained by comprehensively considering the index proportion and the comprehensive diagnosis weight, the refinement and personalization of risk probability calculation are realized, and compared with simple weighted calculation, the real risk of an individual suffering from a target disease can be more accurately reflected, and the scientificity and reliability of disease risk prediction are further improved.

[0059] In one possible implementation of the embodiment of the present application, based on the medication information, the disease risk, the abnormal index set and the time series index data, a large model is called to generate a health assessment report for the target individual, which includes: The time series index data is analyzed by using the moving average analysis method to determine the index change trend; The medication information, the disease risk, the abnormal index set and the index change trend are substituted into a preset prompt word template to integrate into a structured prompt word; The structured prompt word is input into the large model, and a health assessment report for the target individual is received, which includes disease risk warning, medication adjustment suggestion and index interpretation report.

[0060] The embodiment uses the moving average analysis method to analyze the time series index data to determine the change trend, which can smooth short-term data fluctuations and clearly present the long-term change rule of the index, helping to identify potential health deterioration or improvement trends; the medication information, the disease risk, the abnormal index set and the index change trend are integrated into a structured prompt word to ensure that the large model obtains comprehensive and organized input data, avoiding the decline of report quality caused by information missing or confusion; the health assessment report including disease risk warning, medication adjustment suggestion and index interpretation report is generated by inputting the large model, which can convert multi-dimensional data into health guidance content that is easy to understand and practical, so that the target individual can clearly understand their own health status, risk points and countermeasures, solve the problem that traditional data presentation is professional and obscure, and users are difficult to understand and apply, and improve the practicality of health management.

[0061] In the embodiment, the preset prompt word template can be flexibly set according to actual needs. For example, the preset prompt word template includes: generating a disease risk warning according to the suspected disease type and the suspected probability in the disease risk; generating an index interpretation report according to the abnormal index set and the index change trend; and generating a medication adjustment suggestion according to the medication information and the drug-induced influence index and the common drug-induced superposition index in the abnormal index set.

[0062] An electronic device is provided in embodiments of the present application, such as Figure 2 Figure 2 The electronic device 200 shown in FIG. 2 includes a processor 201 and a memory 203. The processor 201 and the memory 203 are connected, such as through a bus 202. Optionally, the electronic device 200 can also include a transceiver 204. It should be noted that the transceiver 204 is not limited to one in actual applications, and the structure of the electronic device 200 does not constitute a limitation on the embodiments of the present application.

[0063] The processor 201 can be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute various exemplary logical blocks, modules and circuits described in combination with the disclosure of the present application. The processor 201 can also be a combination of computing functions, such as one or more microprocessor combinations, combinations of DSP and microprocessor, etc.

[0064] The bus 202 can include a path for transmitting information between the above-mentioned components. The bus 202 can be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. The bus 202 can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 2 In the figure, only one thick line is used, but it does not mean that there is only one bus or one type of bus.

[0065] ​The memory 203 can be a ROM (Read Only Memory) or other type of static storage device that can store static information and instructions, a RAM (Random Access Memory) or other type of dynamic storage device that can store information and instructions, an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer, but is not limited thereto.

[0066] The memory 203 is configured to store application program codes for implementing the solutions of the present application, and the processor 201 is configured to control the execution. The processor 201 is configured to execute the application program codes stored in the memory 203 to implement the content shown in the foregoing personal health warning method embodiments.

[0067] Figure 2 The electronic device shown is merely an example, and should not impose any limitation on the functions and use range of the embodiments of the present application.

[0068] The embodiments of the present application provide a computer readable storage medium, which stores a computer program. When the computer program is run on a computer, the computer can execute the content shown in the foregoing personal health warning method embodiments.

[0069] It should be understood that, although each step in the flowchart of the accompanying drawings is shown in sequence according to the direction of the arrow, these steps are not necessarily executed in sequence according to the direction of the arrow. Unless otherwise specified herein, the execution of these steps is not strictly limited in sequence, and they can be executed in other sequences. Moreover, at least part of the steps in the flowchart of the accompanying drawings can include multiple sub-steps or multiple stages, which are not necessarily executed at the same time, but can be executed at different times, and the execution sequence is not necessarily sequential, but can be executed in rotation or alternation with at least part of other steps or sub-steps or stages of other steps.

[0070] The embodiments of the present application provide a computer program product, which includes a computer program. When the computer program is executed by a processor, the content shown in the foregoing personal health warning method embodiments is implemented.

[0071] The above merely preferred embodiments of the present application and it should be noted that for those of ordinary skill in the art, without departing from the principles of the present application, can make a number of improvements and refinements, these improvements and refinements should also be considered as the scope of protection of the present application.

Claims

1. A personal health alert method, characterized by, The method comprises the following steps: obtaining the medical records and individual attribute information of a target individual, generating a series of medication prompt information containing timestamps based on the medical records and individual attribute information, and sending the medication prompt information to the target individual at the corresponding timestamp through a medicine cabinet module; obtaining the ideal index range corresponding to the target individual, the group index data of patients with the same disease, and the time sequence index data of the target individual during medication; extracting the medication information of the target individual from the medical records, identifying an abnormal index set from the time sequence index data based on the medication information, the ideal index range, and the group index data, and predicting the risk of the target individual based on the abnormal index set; based on the medication information, the risk of the target individual, the abnormal index set, and the time sequence index data, calling a large model to generate a health assessment report for the target individual.

2. The personal health alert method of claim 1, wherein, obtaining the ideal index range corresponding to the target individual and the group index data of patients with the same disease comprises: extracting age information and gender information from the individual attribute information, and calling the ideal index range matching the age information and the gender information from a medical database; determining the disease information of the target individual from the medical records, the disease information including disease type and disease severity; retrieving index data from a plurality of patients with the same disease from the medical database that matches the disease information and the individual attribute information, the index data of the plurality of patients with the same disease constituting the group index data.

3. The personal health alert method of claim 1, wherein, The abnormal index set is identified from the time sequence index data based on the medication information, the ideal index range, and the group index data, comprising: screening the current index data of the target period before the current time from the time sequence index data, comparing the current index data with the ideal index range one by one, and screening an initial abnormal index set; calculating the group index statistical value based on the group index data, comparing the group index statistical value with the ideal index range, screening common abnormal indexes, and determining the common expected range of the common abnormal indexes in the disease state; based on the medication information, identifying drug-induced impact indexes and normal fluctuation ranges of the drug-induced impact indexes; labeling each index in the initial abnormal index set with an index type, the index type including: common abnormal index, common abnormal index, drug-induced impact index, and common drug-induced superposition index; based on the ideal index range and the normal fluctuation range, determining the single expected range of the drug-induced impact index, and based on the common expected range and the normal fluctuation range, determining the superposition expected range of the drug-induced impact index; based on the common expected range, the single expected range, and the superposition expected range, screening an abnormal index set from the initial abnormal index set.

4. The personal health alert method of claim 1, wherein, The method comprises the following steps: calling a medical knowledge base, the medical knowledge base pre-storing associated indexes corresponding to various diseases; match the abnormal indicator set with the correlation indicator to determine a candidate disease set; calculate a risk probability of the target individual suffering from each disease type in the candidate disease set based on the abnormal indicator set and a diagnostic weight of the correlation indicator of each disease type in the candidate disease set; determine a disease type in the candidate disease set with a risk probability exceeding a preset risk probability threshold as a suspected disease, and the suspected disease and the risk probability of the suspected disease constitute the disease risk.

5. The personal health alert method of claim 4, wherein, The calculating of the risk probability of the target individual suffering from each disease type in the candidate disease set based on the abnormal indicator set and the diagnostic weight of the correlation indicator of each disease type in the candidate disease set comprises: obtaining a correlation indicator set of a target disease type and a basic diagnostic weight of each correlation indicator from the medical knowledge base, the target disease type being any type in the candidate disease set; determining an intersection indicator of the abnormal indicator set and the correlation indicator set, and calculating an indicator proportion of the intersection indicator in the correlation indicator set; for each intersection indicator, calculating an abnormal degree of the intersection indicator, determining an evidence strengthening coefficient based on the abnormal degree, and adjusting the basic diagnostic weight of the intersection indicator to obtain a diagnostic weight based on the evidence strengthening coefficient; obtaining a comprehensive diagnostic weight of the target disease type based on the diagnostic weight of each intersection indicator; obtaining the risk probability of the target disease type by integrating the indicator proportion and the comprehensive diagnostic weight.

6. The personal health alert method of claim 1, wherein, The calling of a large model to generate a health assessment report for the target individual based on the medication information, the disease risk, the abnormal indicator set and the time series indicator data comprises: determining an indicator change trend by analyzing the time series indicator data using a moving average analysis method; integrating the medication information, the disease risk, the abnormal indicator set and the indicator change trend into a structured prompt word by substituting them into a preset prompt word template; inputting the structured prompt word into the large model, and receiving the health assessment report for the target individual from the large model, the health assessment report including a disease risk warning, a medication adjustment suggestion and an indicator interpretation report.

7. An electronic device, comprising: comprise: at least one processor; a memory; at least one application program, wherein the at least one application program is stored in the memory and configured to be executed by the at least one processor, and the at least one application program is configured to execute the personal health warning method of any one of claims 1-6.

8. A computer-readable storage medium having stored thereon a computer program, characterized in that, When the computer program is executed in the computer, the computer executes the personal health warning method of any one of claims 1-6.

9. A computer program product, characterised in that, comprise a computer program, which, when executed by a processor, implements the steps of the personal health warning method of any one of claims 1-6.

Images