An intelligent health management method, system, storage medium and electronic device

Abstract

The application relates to an intelligent health management method and system, a storage medium and an electronic device, and relates to the technical field of health management.The method comprises the following steps: acquiring symptom description information of a target user; determining a symptom entity corresponding to the symptom description information through a preset BERT model; determining medication recommendation information of a target disease corresponding to the symptom entity based on a preset medication knowledge graph; determining commonly used drugs of the target user according to a preset health data platform, and optimizing the medication recommendation information according to the commonly used drugs to obtain target medication recommendation information; determining a target portrait of the target user according to the health data platform, and optimizing the target medication recommendation information according to the target disease and the target portrait to obtain final medication recommendation information corresponding to the target user; and sending the target disease and the final medication recommendation information to a terminal of the target user, and generating a suitable follow-up plan corresponding to the target user.The application has the effect of improving the health management efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of health management, and in particular to an intelligent health management method and system, a storage medium and an electronic device. BACKGROUND

[0002] Health management refers to the whole process of health management of users throughout their life cycle based on disease inquiry and diagnosis-drug intervention-follow-up and other whole links throughout the whole process of health-disease-recovery of users. In addition, health management is an active, whole-process and personalized medical service mode, which helps patients change from "passive treatment" to "active health maintenance" through the integration of medical resources, technical tools and behavior intervention, and ultimately achieves the goal of reducing disease risk, controlling disease progression and improving life quality. In the context of aging and increasing burden of chronic diseases, health management has become an important part of the modern medical system.

[0003] Currently, the way to provide health management services to users is usually as follows: using artificial methods to realize inquiry and diagnosis of users, to recommend targeted medication for users according to the diagnosis, and to develop a targeted follow-up plan for the subsequent rehabilitation process of users. Once the number of users for inquiry is large, the time cost of providing health management services based on artificial methods is high, resulting in low efficiency of health management for users. SUMMARY

[0004] In order to improve the efficiency of health management, the present application provides an intelligent health management method, system, storage medium and electronic device.

[0005] In a first aspect of the present application, an intelligent health management method is provided, which specifically comprises:

[0006] Obtaining symptom description information of a target user;

[0007] Determining a symptom entity corresponding to the symptom description information through a preset BERT model;

[0008] Determining medication recommendation information of a target disease corresponding to the symptom entity based on a preset medication knowledge graph, the medication knowledge graph being a knowledge graph obtained by organizing symptoms, diseases, medication information and mutual relationships in a graph structure;

[0009] Determining commonly used drugs of the target user according to a preset health data platform, and optimizing the medication recommendation information according to the commonly used drugs to obtain target medication recommendation information;

[0010] According to the health data center, the target user's target image is determined, and the target drug recommendation information is optimized according to the target disease and the target image, and the final drug recommendation information corresponding to the target user is obtained.

[0011] After receiving the confirmation instruction of the doctor terminal to the target disease and the final drug recommendation information, the target disease and the final drug recommendation information are sent to the terminal of the target user, and the appropriate follow-up plan corresponding to the target user is generated.

[0012] By adopting the above technical scheme, after obtaining the symptom description information, the symptom description information is analyzed by the BERT model to accurately identify the symptom entity, and then the target disease and the corresponding drug recommendation information associated with the symptom entity are quickly and accurately matched through the drug knowledge graph, so that the target user is efficiently inquired, diagnosed and intervened in drug use. Further, the target drug recommendation information is optimized in combination with the commonly used drugs of the target user, so as to avoid the conflict between the recommended drugs and the commonly used drugs, and then the target drug recommendation information is optimized again in combination with the target image and the target disease, so as to obtain the final drug recommendation information, so that the drug recommendation is more suitable for the needs of the target user. Finally, after the doctor confirms, the target disease and the final drug recommendation information are sent to the terminal of the target user, and the appropriate follow-up plan corresponding to the target user is generated, so as to effectively avoid the artificial health management of the target user and improve the efficiency of health management.

[0013] In an embodiment, before determining the commonly used drugs of the target user according to the preset health data center, the method further comprises:

[0014] The multi-source health data of the target user is integrated by a preset Apache NiFi to obtain integrated data;

[0015] The integrated data is cleaned by a preset Flink to obtain cleaned data;

[0016] The cleaned data is stored in a lake warehouse integrated architecture to obtain a health data center.

[0017] In an embodiment, the target drug recommendation information is optimized according to the target disease and the target image to obtain the final drug recommendation information corresponding to the target user, specifically comprising:

[0018] According to a plurality of historical drugs that have caused side effect reactions in historical users, at least one target drug is determined, the target drug is a historical drug that is prone to cause side effect reactions, and the historical user is a user with the target image of the target disease;

[0019] determining at least one target dose range according to the history of the user, the target dose range being a drug dose range that is prone to cause side effects, the drug dose range being a dose range that is effective for the target disease;

[0020] determining a first weight coefficient of each of the target drugs and a second weight coefficient of each target dose range corresponding to the target drug, the first weight coefficient representing the possibility of the target drug causing side effects, and the second weight coefficient representing the possibility of the drug dose being in the target dose range causing side effects;

[0021] optimizing the target drug recommendation information according to the first weight coefficient and the second weight coefficient to obtain the final drug recommendation information corresponding to the target user.

[0022] In an embodiment, the target drug recommendation information includes at least one actual recommended drug and a corresponding actual recommended dose, and the optimization of the target drug recommendation information according to the first weight coefficient and the second weight coefficient to obtain the final drug recommendation information corresponding to the target user specifically includes:

[0023] if each of the actual recommended drugs is the target drug, and there is a corresponding actual recommended dose in the target dose range corresponding to each of the actual recommended drugs, the corresponding target dose range is determined as a key dose range;

[0024] calculating a first weight coefficient of each of the actual recommended drugs and a second weight coefficient of the first coefficient product of the second weight coefficient of the key dose range, and if the first coefficient product exceeds a preset first threshold, the corresponding actual recommended drug is determined as an adjusted drug;

[0025] calculating a second weight coefficient of the adjusted drug and a second weight coefficient of each target dose range, and selecting a minimum second coefficient product from each of the second coefficient products;

[0026] determining a suitable recommended dose corresponding to the adjusted drug according to the target dose range corresponding to the minimum second coefficient product, and summing the minimum second coefficient product corresponding to the adjusted drug and the first coefficient product corresponding to the remaining recommended drug to obtain a first summation result, the remaining recommended drug being an actual recommended drug other than the adjusted drug;

[0027] if the first summation result does not exceed a preset second threshold, the target drug recommendation information is optimized according to the suitable recommended dose to obtain the final drug recommendation information corresponding to the target user;

[0028] If the first summation result exceeds a preset second threshold value, the actual recommended drug in the target drug recommendation information is adjusted to obtain the final drug recommendation information corresponding to the target user.

[0029] In an implementation, the adjusting the actual recommended drug in the target drug recommendation information to obtain the final drug recommendation information corresponding to the target user specifically includes:

[0030] According to the multiple groups of historical drug combinations for the target disease of the historical user, at least one group of target drug combinations is determined, the target drug combination being a historical drug combination easy to take effect on the target disease;

[0031] A plurality of time length ranges for the target disease of the historical user under a single target drug combination are obtained, and according to each time length range, at least one target time length range is determined, the target time length range being a time length range easy to take effect;

[0032] A first weight value of each target drug combination is determined, and a second weight value of each target time length range corresponding to the target drug combination is determined;

[0033] A first product of the first weight value of each target drug combination and the second weight value of each first time length range corresponding thereto is calculated, a summation of each first product is performed to obtain a second summation result of the corresponding target drug combination, the first time length range being a target time length range within a time length interval from 0 to a preset time length threshold value;

[0034] A maximum second summation result is selected from each second summation result, a target drug combination corresponding to the maximum second summation result is determined as a final drug combination, and the actual recommended drug in the target drug recommendation information is adjusted according to the final drug combination to obtain the final drug recommendation information corresponding to the target user.

[0035] In an implementation, the method further includes:

[0036] According to the multiple groups of historical drug combinations for the target disease of the historical user, at least one group of target drug combinations is determined, the target drug combination being a historical drug combination easy to take effect on the target disease;

[0037] A plurality of time length ranges for the target disease of the historical user under a single target drug combination are obtained, and according to each time length range, at least one target time length range is determined, the target time length range being a time length range easy to take effect;

[0038] A first weight value of each target drug combination is determined, and a second weight value of each target time length range corresponding to the target drug combination is determined;

[0039] The final recommended drugs in the final drug recommendation information are summarized to obtain a final drug combination, when the final drug combination is a target drug combination, a first product of a first weight value of the final drug combination and a second weight value of a corresponding second time length range is calculated, the second time length range is a target time length range in a time length interval from 0 to a preset time length threshold, and a sum of each second product is obtained to obtain a corresponding third sum result;

[0040] If the third sum result exceeds a preset third threshold, each final recommended drug in the final drug recommendation information is verified to be reasonable.

[0041] In an embodiment, the method further comprises:

[0042] An intersection operation is performed on each target drug combination to obtain at least one common drug, and each target drug combination containing a single common drug is determined as an associated drug combination;

[0043] A third product of a first weight value of each associated drug combination and a second weight value of a corresponding third time length range is calculated, the third time length range is a target time length range in a time length interval from 0 to a preset time length threshold, and a sum of each third product is obtained to obtain a fourth sum result of the corresponding associated drug combination;

[0044] If the fourth sum result exceeds the third threshold, the corresponding associated drug combination is determined as a key drug combination, and a first number of the corresponding associated drug combination of a single common drug and a second number of the corresponding key drug combination are counted;

[0045] A ratio of the second number to the first number corresponding to the same common drug is calculated, when each final recommended drug in the final drug recommendation information is the common drug, a medication reminder frequency of the corresponding final recommended drug is determined according to the ratio, and the larger the ratio is, the more the medication reminder frequency of the corresponding final recommended drug is.

[0046] In a second aspect of the present application, an intelligent health management system is provided, specifically comprising:

[0047] An information acquisition module is configured to acquire symptom description information of a target user;

[0048] A symptom recognition module is configured to determine a symptom entity corresponding to the symptom description information by using a preset BERT model;

[0049] The medication recommendation module is configured to determine medication recommendation information of the target disease corresponding to the symptom entity based on a preset medication knowledge graph, wherein the medication knowledge graph is a knowledge graph obtained by organizing symptoms, diseases, medication information, and mutual relationships in a graph structure.

[0050] The first optimization module is configured to determine commonly used medication of the target user according to the preset health data platform, and optimize the medication recommendation information based on the commonly used medication to obtain target medication recommendation information.

[0051] The second optimization module is configured to determine a target portrait of the target user according to the health data platform, and optimize the target medication recommendation information based on the target disease and the target portrait to obtain final medication recommendation information corresponding to the target user.

[0052] The follow-up determination module is configured to send the target disease and the final medication recommendation information to a terminal of the target user and generate a suitable follow-up plan corresponding to the target user after receiving a confirmation instruction of the target disease and the final medication recommendation information from a doctor terminal.

[0053] By using the above technical solutions, the information acquisition module acquires symptom description information of a target user, the symptom recognition module determines a symptom entity corresponding to the symptom description information, then the medication recommendation module determines medication recommendation information of a target disease corresponding to the symptom entity based on a medication knowledge graph, then the first optimization module optimizes the medication recommendation information to obtain target medication recommendation information, then the second optimization module optimizes the target medication recommendation information to obtain final medication recommendation information corresponding to the target user, and finally, the target disease and the final medication recommendation information are sent to a terminal of the target user, and a suitable follow-up plan corresponding to the target user is generated.

[0054] In a third aspect of the present application, a computer readable storage medium is provided, wherein the computer readable storage medium stores a computer program, and when the computer program is loaded and executed by a processor, the method steps of any one of the first aspect are executed.

[0055] In a fourth aspect of the present application, an electronic device is provided, specifically comprising:

[0056] A processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein the processor is configured to load and execute the computer program stored in the memory, so that the electronic device executes the method of any one of the first aspect.

[0057] To sum up, the present application comprises at least one of the following beneficial technical effects: after obtaining the symptom description information, the symptom description information is analyzed by the BERT model to accurately identify the symptom entity, and then the symptom entity related target disease and corresponding drug recommendation information are quickly and accurately matched through the drug knowledge graph, so as to realize efficient inquiry, diagnosis and drug intervention for the target user. Further, the target drug recommendation information is optimized in combination with the commonly used drugs of the target user, so as to avoid the conflict between the recommended drugs and the commonly used drugs, and then the target drug recommendation information is optimized again in combination with the target portrait and the target disease, so as to obtain the final drug recommendation information, so that the drug recommendation is more suitable for the needs of the target user. Finally, after the doctor confirms, the target disease and the final drug recommendation information are sent to the terminal of the target user, and the appropriate follow-up plan corresponding to the target user is generated, so as to effectively avoid the manual health management of the target user and improve the efficiency of health management. BRIEF DESCRIPTION OF DRAWINGS

[0058] Figure 1 FIG. 1 is a flow diagram of an intelligent health management method provided by an embodiment of the present application;

[0059] Figure 2 FIG. 2 is a system architecture diagram of an intelligent health management method provided by an embodiment of the present application;

[0060] Figure 3 FIG. 3 is a structure diagram of an intelligent health management system provided by an embodiment of the present application;

[0061] Figure 4 FIG. 4 is a structure diagram of another intelligent health management system provided by an embodiment of the present application.

[0062] Reference signs: 11, information acquisition module; 12, symptom identification module; 13, drug recommendation module; 14, first optimization module; 15, second optimization module; 16, follow-up determination module; 17, middle platform construction module; 18, drug verification module; 19, drug reminder module. DETAILED DESCRIPTION

[0063] In order to enable personnel in the technical field to better understand the technical solutions in the present specification, the technical solutions in the present specification will be described clearly and completely in combination with the drawings in the embodiments of the present specification. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments.

[0064] In the description of the embodiments of the present application, the words "exemplary", "for example", or "e.g." are used to mean serving as an example, instance, or illustration. Any embodiment or design solution described as "exemplary", "for example", or "e.g." in the embodiments of the present application should not be construed as being more preferred or advantageous than other embodiments or design solutions. In fact, the use of the words "exemplary", "for example", or "e.g." is intended to present related concepts in a specific manner.

[0065] In the description of the embodiments of the present application, the term "and / or" is merely a description of the association relationship of the associated objects, which means that there can be three relationships, for example, A and / or B, which can represent the three cases of A alone, B alone, and A and B together. In addition, unless otherwise specified, the term "multiple" means two or more. For example, multiple systems refer to two or more systems, and multiple screen terminals refer to two or more screen terminals. In addition, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly indicating the indicated technical features. Therefore, the features defined with "first" and "second" can explicitly or implicitly include one or more features. The terms "include", "contain", "have" and their variants mean "include but are not limited to", unless otherwise specifically emphasized.

[0066] Referring to Figure 1 The embodiments of the present application disclose a flowchart of an intelligent health management method, which can be implemented by relying on a computer program and can also be run on an intelligent health management system based on the von Neumann system. The computer program can be integrated in an application or run as an independent tool application. Specifically, the computer program comprises the following steps:

[0067] S101: Obtain symptom description information of a target user.

[0068] Specifically, in the embodiments of the present application, the target user is a user who performs online consultation for physical discomfort, and the symptom description information is a description of the target user for the physical discomfort symptoms. The symptom description information can be description information obtained based on the oral description of the target user for the physical discomfort symptoms. The oral description can be oral description in the form of Mandarin, or oral description in the form of the dialect of the target user. In other embodiments, the symptom description information can also be a textual description of the physical discomfort symptoms.

[0069] Further, the execution subject of the intelligent health management method disclosed in the application is a server, the server is wirelessly connected with a terminal, the terminal is a smart phone or a personal computer of a target user, the terminal is installed with a health management related client, the server is a background server of the client, and specifically can be an independent physical server or a cluster composed of multiple physical servers. An implementation scenario is as follows: when the target user feels unwell and needs to perform online consultation, the online consultation function of the client in the terminal is opened, and the discomfort symptoms of the target user are described, finally the server collects the voice information described by the target user through a pre-installed microphone component in the terminal, and then the voice information is analyzed through a natural language processing (NLP) technology to obtain the symptom description information of the target user. In other embodiments, the target user can also send the text description of the discomfort symptoms to the server through the client. For details, see Figure 2 .

[0070] S102: Determine the symptom entity corresponding to the symptom description information through a pre-set BERT model.

[0071] Specifically, the BERT model is a pre-training language model based on the Transformer architecture, which can efficiently learn the deep semantic representation of language and perform outstandingly in many tasks such as question answering, text classification and sentiment analysis. After the symptom description information is determined, the symptom description information is input into the BERT model, the symptom text in the symptom description information is analyzed through the BERT model, and finally the symptom entity is obtained. The symptom entity is a specific abnormal performance described by the target user subjectively or found by objective examination, which has quantifiable attributes and context dependence. For example, the symptom description information is: fever appeared 3 days ago, the highest body temperature is 39℃, accompanied by cough, sore throat, and chest tightness appeared today. Then the symptom entities extracted by the BERT model are: fever, cough, sore throat and chest tightness.

[0072] S103: Determine the medication recommendation information of the target disease corresponding to the symptom entity based on a pre-set medication knowledge graph.

[0073] Specifically, in the embodiments of the present application, the medication knowledge graph is a Neo4j knowledge graph, which includes different symptoms, diseases, medication information entities and the mutual relationship between the entities, i.e., the knowledge graph obtained by organizing the symptoms, diseases, medication information and mutual relationship in a graph structure. The medication information includes at least one drug and the corresponding drug dosage. The Neo4j is an open-source graph database (Graph Database) that is specially used for efficiently storing, querying and managing graph structure data. The Neo4j knowledge graph refers to a knowledge graph constructed based on the Neo4j database, which structures and represents the entities and their associations in the real world in a graphical manner through a triple model of "node (Node)-relationship (Relationship)-property (Property)", and is particularly suitable for processing complex semantic relationship networks. The symptoms, diseases and medication information are nodes in the medication knowledge graph. Further, through multi-hop reasoning technology, the target disease corresponding to the symptom entity and the medication recommendation information for the target disease are determined from the medication knowledge graph, and the medication recommendation information includes at least one initial recommended drug and the corresponding initial recommended dosage for the target disease. Multi-hop reasoning is a complex relationship retrieval and logical deduction technology based on graph structure, which realizes deep semantic association analysis across entities and relationships by multi-step jumping in the network composed of nodes and edges.

[0074] S104: According to the preset health data platform, determine the commonly used drugs of the target user, and optimize the medication recommendation information according to the commonly used drugs to obtain target medication recommendation information.

[0075] Specifically, the health data platform refers to a unified data management and service hub constructed around the patient's whole life cycle health data. It integrates patient health data scattered in different systems and institutions, standardizes governance, structures storage, and provides intelligent services, and finally provides efficient, safe, and unified data support for clinical diagnosis and treatment, health management, medical research, and other scenarios. Under the premise of obtaining the authorization or consent of the target user, based on the health data platform, the commonly used drugs of the target user are obtained. For example, the target user himself has a basic disease of hypertension, and the commonly used drug is a depressor. Finally, the target drug recommendation information is optimized according to the commonly used drug, and the target drug recommendation information is obtained. One implementable implementation is: based on the preset contraindication reference table, it is judged whether the commonly used drug and each initial recommended drug in the drug recommendation information exist a contraindication relationship. If the commonly used drug and each initial recommended drug do not exist a contraindication relationship, then the drug recommendation information is directly determined as the target drug recommendation information. If the commonly used drug and the initial recommended drug exist a contraindication relationship, then based on the drug knowledge graph, the target drug recommendation information is determined through multi-hop reasoning, so that each drug in the target drug recommendation information and the commonly used drug do not exist a contraindication relationship. The contraindication reference table includes different drugs that exist a contraindication relationship. The contraindication relationship between drugs means that when two or more drugs are used at the same time, they may produce harmful interactions, leading to reduced drug efficacy, increased toxicity, or the risk of serious adverse reactions.

[0076] In other embodiments, step S104 further includes: obtaining integrated data by integrating multi-source health data of the target user through a preset Apache NiFi, wherein the multi-source health data includes but is not limited to electronic medical records, test reports, and historical medication information of the target user. Then, the integrated data is cleaned through a preset Flink to obtain cleaned data, thereby improving the accuracy and reliability of the health data of the target user. Finally, the cleaned data is stored in a lake warehouse integrated architecture to obtain the health data platform. The lake warehouse integrated architecture is a new data management architecture that combines the advantages of data lake and data warehouse, aiming to provide a unified, flexible, and high-performance data storage and processing platform.

[0077] S105: determining the target image of the target user according to the health data platform, and optimizing the target drug recommendation information according to the target disease and the target image to obtain the final drug recommendation information corresponding to the target user.

[0078] Specifically, under the premise of obtaining the authorization or consent of the target user, the basic information of the target user can be obtained according to the health data center, including but not limited to age, gender, medical history and other information. According to the basic information, the target image of the target user is determined. A feasible determination method is to input the basic information into the trained image prediction model to obtain the target image of the target user. The image prediction model can be a recurrent neural network model or a random forest model. The training process is as follows: the basic information and the correct image of the historical patients are used as training pairs to train the model, and the model is adjusted through the backpropagation algorithm to obtain the image prediction model.

[0079] Further, the target drug recommendation information is optimized according to the target disease and the target image. A realizable implementation is as follows: based on the historical statistical records of the side effects of drugs, a plurality of historical drugs that cause side effect reactions in historical users are obtained, wherein the historical users are users with the target image of the target disease. Side effect reactions include but are not limited to skin rash, nausea, liver damage and other reactions. The historical statistical records include information such as historical drugs taken by users of different images when they have side effect reactions and the dose range at the time of taking. The first frequency of repeated occurrence of a single historical drug in all historical drugs is counted. If the first frequency exceeds the corresponding frequency threshold, it means that the corresponding historical drug is easy to cause side effect reactions when taken by historical users, and the corresponding historical drug is determined as the target drug, i.e. the historical drug that is easy to cause side effect reactions.

[0080] Further, based on the above historical statistical records, a plurality of drug dose ranges of a single target drug taken by historical users when they have side effect reactions are obtained. The second frequency of repeated occurrence of a single drug dose range in all different drug dose ranges is counted. If the second frequency exceeds the corresponding frequency threshold, the corresponding drug dose range is determined as the target dose range of the target drug, i.e. taking the target drug in this target dose range is easy to cause side effect reactions. Then, the first weight coefficient of each target drug is determined, and the second weight coefficient of the target dose range corresponding to each target drug is determined, wherein the first weight coefficient is the ratio of the first frequency of each target drug to the sum of the first frequencies of all target drugs. The second weight coefficient is the ratio of the second frequency of the single target dose range corresponding to the target drug to the sum of the second frequencies of all target dose ranges corresponding to the target drug. The drug dose range is the dose range that takes effect on the target disease.

[0081] Finally, the target drug recommendation information is optimized according to the first weight coefficient and the second weight coefficient. A feasible optimization method is that the target drug recommendation information includes at least one actual recommended drug and a corresponding actual recommended dose. If each actual recommended drug is the target drug, when the corresponding actual recommended dose exists in the target dose range of a single actual recommended drug, the corresponding target dose range is determined as a key dose range. Then, the first weight coefficient of each actual recommended drug and the first coefficient product of the second weight coefficient of the corresponding key dose range are calculated. The larger the first coefficient product is, the greater the possibility that the target user takes the actual recommended drug at the corresponding actual recommended dose and the body has a side effect reaction. The first coefficient product is compared with a preset first threshold value. If the first coefficient product exceeds the first threshold value, it means that the possibility of the body having a side effect reaction is large, and the corresponding actual recommended dose needs to be adjusted and optimized. Then, the corresponding actual recommended drug is determined as a to-be-adjusted drug.

[0082] Further, the second coefficient product of the first weight coefficient of the to-be-adjusted drug and the second weight coefficient of each target dose range is calculated. The minimum second coefficient product is selected from each second coefficient product. The possibility of causing a side effect reaction is small when the to-be-adjusted drug is taken at the target dose range corresponding to the minimum second coefficient product. Then, the maximum value in the target dose range corresponding to the minimum second coefficient product is determined as the appropriate recommended dose corresponding to the to-be-adjusted drug. Then, the minimum second coefficient product corresponding to the to-be-adjusted drug and the first coefficient product corresponding to the remaining recommended drug are summed to obtain a first summation result. The larger the first summation result is, the greater the overall possibility of causing a side effect reaction when the target user takes each actual recommended drug combination. The remaining recommended drug is an actual recommended drug other than the to-be-adjusted drug.

[0083] The first summation result is compared with a preset second threshold value. If the first summation result does not exceed the second threshold value, it means that the overall possibility of causing a side effect reaction when each actual recommended drug combination is taken is small. Then, only the actual recommended dose of the corresponding to-be-adjusted drug needs to be adjusted and optimized according to the appropriate recommended dose, and the drug combination does not need to be adjusted. Finally, the target drug recommendation information is optimized to obtain a final drug recommendation information. It should be noted that the second threshold value is greater than the first threshold value.

[0084] Further, if the first sum result exceeds the second threshold value, indicating that the overall possibility of side effect reactions when taking the actual recommended drug combination is larger, the actual recommended drugs in the target medication recommendation information need to be adjusted, the drug combination is adjusted, and the final medication recommendation information is obtained. An implementable embodiment is: based on the historical follow-up records, a plurality of groups of historical drug combinations effective for the target disease of the historical user are obtained, and the historical follow-up records include information such as historical drug combinations that are effective for the target disease of the patient and the time range of the effect. The number of repeated occurrences of a single historical drug combination in different historical drug combinations is counted. If the number of repeated occurrences exceeds a preset number threshold, the corresponding historical drug combination is determined as a target drug combination, that is, a historical drug combination that is easy to take effect or effective for the target disease. It should be noted that the historical drug combination is a drug combination taken by the historical user for the target disease.

[0085] Based on the above historical follow-up records, a plurality of time ranges in which the target disease takes effect under a single target drug combination are obtained. The number of repeated occurrences of a single time range in different time ranges is counted. If the number of repeated occurrences exceeds a preset number threshold, the corresponding time range is determined as a target time range corresponding to the target drug combination, that is, a time range that is easy to take effect or effective. Then, the first weight of each target drug combination is determined, and the second weight of each target time range corresponding to the target drug combination is determined. The first weight is the ratio of the number of repeated occurrences of each target drug combination to the sum of the number of repeated occurrences of all target drug combinations. The second weight is the ratio of the number of repeated occurrences of a single target time range corresponding to the target drug combination to the sum of the number of repeated occurrences of all target time ranges corresponding to the target drug combination. The first weight represents the possibility of the target drug combination taking effect on the target disease of the user of the target image, and the second weight represents the possibility of the interval time being in the target time range. It should be noted that the time range refers to the range of the interval time from the start of taking the drug to the taking effect.

[0086] Further, a target duration range in a duration interval of 0 to a preset duration threshold is determined as a first duration range. A first product of a first weight value of each target drug combination and a second weight value of a corresponding respective first duration range is calculated, and the greater the first product, the easier the target drug combination is to take effect in the corresponding first duration range. Wherein, the preset duration threshold is a critical value for measuring the length of the effective time. Then, the respective first products are summed to obtain a second summation result corresponding to the target drug combination, and the greater the second summation result, the greater the possibility that the target user will take the corresponding target drug combination and take effect quickly. Then, the maximum second summation result is selected from the respective second summation results, and the target drug combination corresponding to the maximum second summation result is determined as the final drug combination, and the target user takes the final drug combination, and the possibility of faster effect is greater. Finally, the actual recommended drug in the target drug recommendation information is adjusted to the drug in the final drug combination, if the drug in the final drug combination is the target drug, the first weight coefficient of the drug is calculated and the second weight coefficient of the corresponding respective target dose range is multiplied to obtain a coefficient product, the minimum coefficient product is selected from the respective coefficient products, and the target dose range corresponding to the minimum coefficient product is determined as the final recommended dose of the drug; if the drug in the final drug combination is not the target drug, then based on the drug knowledge graph, the final recommended dose of the drug is determined, and the final drug recommendation information is finally obtained, so that the target user can have a good curative effect when taking the medicine according to the final drug recommendation information, and the risk of side effects is reduced.

[0087] S106: After receiving the confirmation instruction of the doctor terminal to the target disease and the final drug recommendation information, the target disease and the final drug recommendation information are sent to the terminal of the target user, and a suitable follow-up plan corresponding to the target user is generated.

[0088] Specifically, after the final drug recommendation information is determined, the final drug recommendation information includes different final recommended drugs and corresponding final recommended dosages, the target user's corresponding target disease and the final drug recommendation information are sent to a doctor terminal, and the doctor terminal is a smart phone or a personal computer of a doctor corresponding to the target user. If a confirmation instruction of the doctor terminal is received for the target disease and the final drug recommendation information, it means that the doctor combines the test report and the symptom description information of the target user and other information to evaluate that the target disease and the final drug recommendation information are relatively reasonable, and then the target disease and the final drug recommendation information are sent to the terminal of the target user. Further, a suitable follow-up plan corresponding to the target user is generated, and a feasible generation method is as follows: the combination of each final recommended drug in the final drug recommendation information is determined as a final drug combination, when the final drug combination is a target drug combination, the product of the first weight value of the final drug combination and the second weight value of the corresponding target time range is calculated, the maximum product is selected from each product, the target time range corresponding to the maximum product is the time range in which the final drug combination is most likely to take effect, and finally the maximum time length in the target time range corresponding to the maximum product is selected, and whether the target user sees effect after the interval of the maximum time length is inquired to determine the suitable follow-up plan, so as to more targetedly understand the curative effect of the target user and make the follow-up more suitable for the target user. In other embodiments, if the final recommended drug is a target drug, the final recommended drug is determined as a drug to be concerned, when there is a corresponding final recommended dosage in the target dosage range corresponding to the drug to be concerned, the corresponding target dosage range is determined as a range to be concerned, the third coefficient product of the first weight coefficient of the drug to be concerned and the second weight coefficient of the corresponding range to be concerned is calculated, and according to the third coefficient product, the follow-up frequency of the side effect reaction of the corresponding drug to be concerned is determined. The larger the third coefficient product is, the more likely the side effect reaction is, and then the follow-up frequency of the side effect reaction of the corresponding drug to be concerned is higher, and finally, the follow-up frequencies of the corresponding drugs to be concerned are determined as the suitable follow-up plan, so as to more targetedly pay attention to the side effect of the target user during taking the drug. It should be noted that the side effect reaction of the drug to be concerned can be matched according to a preset side effect matching table, and the side effect matching table includes different drugs and corresponding drug side effect reactions, which are all set based on human experience.

[0089] In an embodiment, before step S106, when the final drug combination is the target drug combination, a second product of the first weight value of the final drug combination and the second weight value corresponding to each second time length range is calculated, each second product is summed to obtain a corresponding third summation result, and the larger the third summation result is, the faster the final drug combination is likely to take effect. The second time length range is a target time length range in a time length interval from 0 to a preset time length threshold. If the third summation result exceeds a preset third threshold, it indicates that the target user is more likely to take the final drug combination quickly, and each final recommended drug in the final drug recommendation information is verified to be reasonable, thereby improving the accuracy of the recommended drug for the target user.

[0090] In another embodiment, an intersection operation is performed on each target drug combination to obtain at least one common drug, and each target drug combination containing a single common drug is determined as an associated drug combination. A third product of the first weight value of each associated drug combination and the second weight value corresponding to each third time length range is calculated, each third product is summed to obtain a fourth summation result of the corresponding associated drug combination, and the larger the fourth summation result is, the faster the corresponding associated drug combination is likely to take effect. The third time length range is a target time length range in a time length interval from 0 to a preset time length threshold.

[0091] Further, if the fourth summation result exceeds a preset third threshold, it indicates that the corresponding associated drug combination has a relatively obvious effect, and is determined as a key drug combination. The first number of associated drug combinations corresponding to a single common drug and the second number of key drug combinations are counted. The ratio of the second number to the first number corresponding to the same common drug is calculated, the larger the ratio is, the more likely the associated drug combination containing the common drug is to have a significant effect, and thus the higher the contribution of the common drug to the effect is. Further, when each final recommended drug in the final drug recommendation information is a common drug, the medication reminder frequency of the corresponding final recommended drug is determined according to the ratio, the larger the ratio is, the higher the contribution of the corresponding final recommended drug to the effect is, and the more important the timely medication of the final recommended drug is, and thus the more medication reminder frequencies of the final recommended drug are. It should be noted that in the embodiments of the present application, the medication reminder frequency of the final recommended drug can be matched from a preset frequency matching table, and the frequency matching table includes different ratio ranges and corresponding medication reminder frequencies. For example, the ratio range 0-0.2 has a medication reminder frequency of 2 times, the ratio range 0.2-0.4 has a medication reminder frequency of 4 times, and so on. If the ratio of the final recommended drug a is 0.3, the medication reminder frequency of the final recommended drug a is 4 times.

[0092] The implementation principle of the intelligent health management method of the embodiment of the present application is that: after the symptom description information is obtained, the symptom description information is analyzed through the BERT model to accurately identify the symptom entity, and then the symptom entity related target disease and corresponding medication recommendation information are quickly and accurately matched through the medication knowledge graph, so as to realize efficient inquiry, diagnosis and medication intervention for the target user. Further, the target medication recommendation information is obtained by optimizing the medication recommendation information in combination with the commonly used drugs of the target user, so as to avoid the conflict between the recommended medication and the commonly used drugs, and then the target medication recommendation information is optimized again in combination with the target portrait and the target disease to obtain the final medication recommendation information, so that the medication recommendation is more suitable for the needs of the target user. Finally, after the doctor confirms, the target disease and the final medication recommendation information are sent to the terminal of the target user, and the appropriate follow-up plan corresponding to the target user is generated, so as to effectively avoid the artificial health management of the target user and improve the efficiency of health management.

[0093] The following is an embodiment of the system of the present application, which can be used to execute the method embodiments of the present application. For details not disclosed in the system embodiments of the present application, please refer to the method embodiments of the present application.

[0094] Please refer to Figure 3 The structure diagram of the intelligent health management system provided by the embodiment of the present application. The intelligent health management system can be realized by software, hardware or a combination of the two to become all or part of the system. The system includes an information acquisition module 11, a symptom identification module 12, a medication recommendation module 13, a first optimization module 14, a second optimization module 15 and a follow-up determination module 16.

[0095] The information acquisition module 11 is used to acquire the symptom description information of the target user;

[0096] The symptom identification module 12 is used to determine the symptom entity corresponding to the symptom description information through the preset BERT model;

[0097] The medication recommendation module 13 is used to determine the medication recommendation information of the target disease corresponding to the symptom entity based on the preset medication knowledge graph, and the medication knowledge graph is a knowledge graph obtained by organizing the knowledge graph in a graph structure of symptoms, diseases, medication information and mutual relationship;

[0098] The first optimization module 14 is used to determine the commonly used drugs of the target user according to the preset health data platform, and optimize the medication recommendation information according to the commonly used drugs to obtain the target medication recommendation information;

[0099] The second optimization module 15 is configured to determine a target portrait of the target user according to the health data middle platform, and optimize the target drug recommendation information according to the target disease and the target portrait, to obtain the final drug recommendation information corresponding to the target user.

[0100] The follow-up determination module 16 is configured to send the target disease and the final drug recommendation information to the terminal of the target user and generate a suitable follow-up plan corresponding to the target user after receiving a confirmation instruction of the target disease and the final drug recommendation information from the doctor terminal.

[0101] Optionally, as shown in Figure 4 The system further includes a middle platform construction module 17, which is specifically configured to:

[0102] The multiple-source health data of the target user is integrated through a preset Apache NiFi to obtain integrated data.

[0103] The integrated data is cleaned through a preset Flink to obtain cleaned data.

[0104] The cleaned data is stored in a lake-warehouse integrated architecture to obtain the health data middle platform.

[0105] Optionally, the second optimization module 15 is specifically configured to:

[0106] At least one target drug is determined according to a plurality of historical drugs that have caused side effect reactions of a historical user, the target drug being a historical drug that is prone to cause side effect reactions, and the historical user being a user with the target portrait of the target disease;

[0107] At least one target dose range is determined according to a single target drug dose range of the historical user when the side effect reactions occur, the target dose range being a drug dose range that is prone to cause side effect reactions, and the drug dose range being a dose range that takes effect on the target disease;

[0108] A first weight coefficient of each target drug is determined, and a second weight coefficient of each target dose range corresponding to the target drug is determined, the first weight coefficient representing the possibility of causing side effect reactions of the target drug, and the second weight coefficient representing the possibility of causing side effect reactions when the drug dose is in the target dose range;

[0109] The target drug recommendation information is optimized according to the first weight coefficient and the second weight coefficient, to obtain the final drug recommendation information corresponding to the target user.

[0110] Optionally, the second optimization module 15 is specifically configured to:

[0111] If each actual recommended drug is the target drug, when there is a corresponding actual recommended dose in the target dose range corresponding to a single actual recommended drug, the corresponding target dose range is determined as the key dose range;

[0112] The first weight coefficient of each actual recommended drug and the second weight coefficient of the corresponding key dose range are calculated to obtain a first coefficient product, and if the first coefficient product exceeds a preset first threshold value, the corresponding actual recommended drug is determined as the drug to be adjusted;

[0113] The first weight coefficient of the drug to be adjusted and the second weight coefficient of each target dose range corresponding thereto are calculated to obtain a second coefficient product, and the minimum second coefficient product is selected from the second coefficient products;

[0114] According to the target dose range corresponding to the minimum second coefficient product, the appropriate recommended dose corresponding to the drug to be adjusted is determined, and the minimum second coefficient product corresponding to the drug to be adjusted and the first coefficient products corresponding to the remaining recommended drugs are summed to obtain a first summation result, the remaining recommended drugs being the actual recommended drugs other than the drug to be adjusted;

[0115] If the first summation result does not exceed a preset second threshold value, the target medication recommendation information is optimized according to the appropriate recommended dose to obtain the final medication recommendation information corresponding to the target user;

[0116] If the first summation result exceeds the preset second threshold value, the actual recommended drugs in the target medication recommendation information are adjusted to obtain the final medication recommendation information corresponding to the target user.

[0117] Optionally, the second optimization module 15 is specifically configured to:

[0118] According to a plurality of groups of historical drug combinations that are effective for the target disease of the historical user, at least one group of target drug combinations is determined, the target drug combination being a historical drug combination that is easy to be effective for the target disease;

[0119] A plurality of time length ranges in which the target disease of the historical user is effective under a single target drug combination are obtained, and at least one target time length range is determined according to the time length ranges, the target time length range being an easy-to-effect time length range;

[0120] The first weight value of each target drug combination is determined, and the second weight value of each target time length range corresponding to the target drug combination is determined;

[0121] The first weight value of each target drug combination and the second weight value of each first time length range corresponding thereto are calculated to obtain a first product, and the first products are summed to obtain a second summation result corresponding to the target drug combination, the first time length range being a target time length range within a time length interval from 0 to a preset time length threshold value;

[0122] selecting a maximum second summation result from the second summation results, determining a final drug combination corresponding to the maximum second summation result as a target drug combination, and adjusting actual recommended drugs in the target drug recommendation information according to the target drug combination to obtain final drug recommendation information corresponding to the target user.

[0123] Optionally, the system further comprises a drug use verification module 18, specifically configured to:

[0124] determining at least one target drug combination from the multiple groups of historical drug combinations for the target disease of the historical user, the target drug combination being a historical drug combination that is easy to take effect on the target disease;

[0125] obtaining multiple time length ranges in which the target drug combination takes effect on the target disease of the historical user, and determining at least one target time length range from the time length ranges, the target time length range being a time length range in which the target drug combination is easy to take effect;

[0126] determining a first weight value of each target drug combination and a second weight value of each target time length range corresponding to the target drug combination;

[0127] summarizing each final recommended drug in the final drug recommendation information to obtain a final drug combination, and when the final drug combination is a target drug combination, calculating a second product of the first weight value of the final drug combination and the second weight value of each second time length range corresponding to the final drug combination, and summing the second products to obtain a corresponding third summation result, the second time length range being a target time length range within a time length interval from 0 to a preset time length threshold;

[0128] If the third summation result exceeds a preset third threshold, it is verified that each final recommended drug in the final drug recommendation information is reasonable.

[0129] Optionally, the system further comprises a drug use reminding module 19, specifically configured to:

[0130] performing an intersection operation on each target drug combination to obtain at least one common drug, and determining each target drug combination containing a single common drug as an associated drug combination;

[0131] calculating a third product of a first weight value of each associated drug combination and a second weight value of each third time length range corresponding to the associated drug combination, and summing the third products to obtain a fourth summation result of the corresponding associated drug combination, the third time length range being a target time length range within a time length interval from 0 to a preset time length threshold;

[0132] If the fourth summation result exceeds the third threshold, the corresponding associated drug combination is determined as a key drug combination, and a first number of the associated drug combinations corresponding to a single common drug and a second number of the corresponding key drug combinations are counted.

[0133] The ratio of the second number corresponding to the common drug to the first number is calculated, and when each final recommended drug in the final drug recommendation information is the common drug, the use reminding frequency of the corresponding final recommended drug is determined according to the ratio, and the larger the ratio is, the more the use reminding frequency of the corresponding final recommended drug is.

[0134] It should be noted that the intelligent health management system provided in the above embodiment is only used as an example for the division of the above functional modules when the intelligent health management method is executed. In actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above. In addition, the intelligent health management system and the intelligent health management method provided in the above embodiment belong to the same concept, and the implementation process is detailed in the method embodiment, which will not be repeated here.

[0135] The embodiment of the present application also discloses a computer readable storage medium, and the computer readable storage medium stores a computer program, wherein the computer program is executed by a processor to realize the intelligent health management method of the above embodiment.

[0136] The computer program can be stored in the computer readable medium, and the computer program includes computer program code, which can be in the form of source code, object code, executable file or some middleware form, etc. The computer readable medium includes any entity or device, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc. that can carry computer program code. It should be noted that the computer readable medium includes but is not limited to the above components.

[0137] The computer readable storage medium stores the intelligent health management method of the above embodiment in the computer readable storage medium, and is loaded and executed on the processor to facilitate the storage and application of the above method.

[0138] The embodiment of the present application also discloses an electronic device, and the computer readable storage medium stores a computer program, and the computer program is loaded and executed by the processor to realize the above intelligent health management method.

[0139] The electronic device can be a desktop computer, a notebook computer or a cloud server, etc. and the electronic device includes but is not limited to a processor and a memory. For example, the electronic device can also include an input / output device, a network access device and a bus, etc.

[0140] The processor can be a central processing unit (CPU), and can also be other general purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), programmable logic devices (PLD), field programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, etc. The general purpose processor can be a microprocessor or any conventional processor, etc. The present application is not limited in this regard.

[0141] The memory can be an internal storage unit of the electronic device, such as a hard disk or a memory of the electronic device, or an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital card (SD), or a flash memory card (FC) equipped on the electronic device. The memory can also be a combination of the internal storage unit and the external storage device of the electronic device. The memory is used to store computer programs and other programs and data required by the electronic device. The memory can also be used to temporarily store data that has been output or will be output. The present application is not limited in this regard.

[0142] The electronic device stores the intelligent health management method of the above embodiments in the memory of the electronic device, and loads and executes the method on the processor of the electronic device, which is convenient for use.

[0143] The above description is only exemplary embodiments of the present disclosure, and cannot limit the scope of the present disclosure. Any equivalent changes and modifications made in accordance with the teachings of the present disclosure are still within the scope of the present disclosure. The present application is intended to cover any variations, uses, or adaptive changes of the present disclosure that follow the general principles of the present disclosure and include common knowledge or conventional technical means in the technical field not described in the present disclosure. The specification and embodiments are only considered exemplary, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. An intelligent health management method, characterized in that, The method includes: Obtain symptom description information from the target user; The symptom entities corresponding to the symptom description information are determined using a pre-set BERT model. Based on a pre-defined medication knowledge graph, medication recommendation information for the target disease corresponding to the symptom entity is determined. The medication knowledge graph is a knowledge graph organized in a graph structure, which includes symptoms, diseases, medication information, and their interrelationships. Based on the preset health data platform, the commonly used drugs of the target user are determined, and the medication recommendation information is optimized based on the commonly used drugs to obtain the target medication recommendation information; Based on the health data platform, a target profile of the target user is determined, and the target medication recommendation information is optimized based on the target disease and the target profile to obtain the final medication recommendation information corresponding to the target user. This includes: determining at least one target drug based on multiple historical drugs for which the user has experienced side effects, wherein the target drug is a historical drug that is prone to causing side effects, and the historical user is a user with the target disease and the target profile; determining at least one target dosage range based on the dosage range of a single target drug when the user experienced side effects, wherein the target dosage range is a dosage range that is prone to causing side effects, and the dosage range is a dosage range that is effective for the target disease; determining a first weighting coefficient for each target drug and a second weighting coefficient for each target dosage range corresponding to the target drug, wherein the first weighting coefficient represents the probability of the target drug causing side effects, and the second weighting coefficient represents the probability of the drug dosage causing side effects within the target dosage range; and optimizing the target medication recommendation information based on the first weighting coefficient and the second weighting coefficient to obtain the final medication recommendation information corresponding to the target user. After receiving confirmation instructions from the doctor's terminal regarding the target disease and the final medication recommendation information, the target disease and the final medication recommendation information are sent to the target user's terminal, and a suitable follow-up plan is generated for the target user.

2. The intelligent health management method according to claim 1, characterized in that, Before determining the target user's commonly used medications based on a preset health data platform, the process further includes: The multi-source health data of the target user is integrated using the preset Apache NiFi to obtain the integrated data; The integrated data is cleaned using a pre-defined Flink algorithm to obtain cleaned data. The cleaned data is stored in the integrated lake warehouse architecture to obtain the health data platform.

3. The intelligent health management method according to claim 1, characterized in that, The target medication recommendation information includes at least one actual recommended drug and its corresponding actual recommended dosage. The optimization of the target medication recommendation information based on the first weighting coefficient and the second weighting coefficient to obtain the final medication recommendation information for the target user specifically includes: If each of the actual recommended drugs is the target drug, then when there is a corresponding actual recommended dose within the target dose range corresponding to a single actual recommended drug, the corresponding target dose range is determined as the key dose range; Calculate the product of the first weight coefficient of each actual recommended drug and the first coefficient of the second weight coefficient of the corresponding key dosage range. If the first coefficient product exceeds a preset first threshold, the corresponding actual recommended drug is determined as a drug to be adjusted. Calculate the product of the first weighting coefficient of the drug to be adjusted and the second coefficient of the corresponding second weighting coefficient for each target dose range, and select the smallest second coefficient product from all the second coefficient products; Based on the target dose range corresponding to the minimum second coefficient product, the appropriate recommended dose corresponding to the drug to be adjusted is determined, and the minimum second coefficient product corresponding to the drug to be adjusted is summed with the first coefficient product corresponding to the remaining recommended drugs to obtain a first summation result. The remaining recommended drugs are the actual recommended drugs other than the drug to be adjusted. If the first summation result does not exceed the preset second threshold, then the target medication recommendation information is optimized according to the appropriate recommended dose to obtain the final medication recommendation information corresponding to the target user; If the first summation result exceeds a preset second threshold, the actual recommended drugs in the target medication recommendation information are adjusted to obtain the final medication recommendation information corresponding to the target user.

4. The intelligent health management method according to claim 3, characterized in that, The step of adjusting the actual recommended drugs in the target medication recommendation information to obtain the final medication recommendation information corresponding to the target user specifically includes: Based on multiple historical drug combinations that are effective against the target disease of the historical users, at least one target drug combination is determined, wherein the target drug combination is a historical drug combination that is likely to be effective against the target disease. Under a single target drug combination, obtain multiple duration ranges of efficacy for the target disease of historical users, and determine at least one target duration range based on each duration range, wherein the target duration range is a duration range in which efficacy is easily achieved. A first weight is determined for each of the target drug combinations, and a second weight is determined for each of the target duration ranges corresponding to the target drug combinations; Calculate the first product of the first weight of each target drug combination and the second weight of each corresponding first duration range, sum the first products to obtain the second summation result of the corresponding target drug combination, where the first duration range is the target duration range within the duration interval from 0 to a preset duration threshold; The largest second summation result is selected from each of the second summation results. The target drug combination corresponding to the largest second summation result is determined as the final drug combination. Based on the final drug combination, the actual recommended drugs in the target medication recommendation information are adjusted to obtain the final medication recommendation information corresponding to the target user.

5. The intelligent health management method according to claim 1, characterized in that, The method further includes: Based on multiple historical drug combinations that have been effective against the target disease of historical users, at least one target drug combination is determined, wherein the target drug combination is a historical drug combination that is likely to be effective against the target disease. Under a single target drug combination, obtain multiple duration ranges of efficacy for the target disease of historical users, and determine at least one target duration range based on each duration range, wherein the target duration range is a duration range in which efficacy is easily achieved. A first weight is determined for each of the target drug combinations, and a second weight is determined for each of the target duration ranges corresponding to the target drug combinations; The final recommended drugs in the final medication recommendation information are summarized to obtain the final drug combination. When the final drug combination is the target drug combination, the first weight of the final drug combination and the second product of the second weight of each corresponding second time range are calculated, and the second products are summed to obtain the corresponding third summation result. The second time range is the target time range within the time interval from 0 to the preset time threshold. If the third summation result exceeds a preset third threshold, then the rationality of each final recommended drug in the final medication recommendation information is verified.

6. The intelligent health management method according to claim 5, characterized in that, The method further includes: Perform an intersection operation on each of the target drug combinations to obtain at least one common drug, and determine each target drug combination that contains a single common drug as an associated drug combination; Calculate the third product of the first weight of each associated drug combination and the second weight of each corresponding third time range, and sum the third products to obtain the fourth summation result of the corresponding associated drug combination. The third time range is the target time range within the time interval from 0 to a preset time threshold. If the fourth summation result exceeds the third threshold, the corresponding associated drug combination is identified as a key drug combination, and the first number of associated drug combinations corresponding to a single shared drug and the second number of key drug combinations are counted. Calculate the ratio of the second number to the first number corresponding to the same shared drug. When all the final recommended drugs in the final medication recommendation information are the shared drugs, determine the medication reminder frequency of the corresponding final recommended drug based on the ratio. The larger the ratio, the more medication reminders the corresponding final recommended drug will receive.

7. An intelligent health management system for implementing the intelligent health management method according to any one of claims 1 to 6, characterized in that, include: The information acquisition module (11) is used to acquire the symptom description information of the target user; The symptom recognition module (12) is used to determine the symptom entity corresponding to the symptom description information through a preset BERT model; The medication recommendation module (13) is used to determine medication recommendation information for the target disease corresponding to the symptom entity based on a preset medication knowledge graph. The medication knowledge graph is a knowledge graph obtained by organizing symptoms, diseases, medication information and their interrelationships in a graph structure. The first optimization module (14) is used to determine the target user's commonly used drugs according to the preset health data platform, and optimize the drug recommendation information according to the commonly used drugs to obtain the target drug recommendation information; The second optimization module (15) is used to determine the target profile of the target user based on the health data platform, and optimize the target medication recommendation information based on the target disease and the target profile to obtain the final medication recommendation information corresponding to the target user. The follow-up determination module (16) is used to send the target disease and the final medication recommendation information to the target user's terminal after receiving the confirmation instruction from the doctor's terminal for the target disease and the final medication recommendation information, and to generate a suitable follow-up plan for the target user.

8. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is loaded and executed by the processor, it implements the method of any one of claims 1-6.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that, When the processor loads and executes the computer program, it implements the method of any one of claims 1-6.

Images