Family doctor intelligent service system and service method based on big data

Abstract

The present invention discloses a family doctor intelligent service system and service method based on big data, which relates to the field of intelligent medical service technology, performs health assessment on high-risk patients, and provides corresponding guidance to high-risk patients in combination with their health status; when abnormal conditions are identified in high-risk patients, the diagnosis and treatment-related data of high-risk patients are sent to a telemedicine platform, and the telemedicine platform outputs the diagnosis results; the condition recognition model is used to assess the condition of high-risk patients and determine the corresponding condition level, and when high-risk patients enter the registration appointment state, the corresponding appointment queuing strategy is selected according to the condition level; the appointment order of high-risk patients is optimized, and after the diagnosis state is ended, a personalized health management plan is given to high-risk patients by a health management knowledge map. The appointments of different mild patients are optimized and adjusted, so that the appointments for consultations are more reasonable and the waste of medical resources is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent medical services, and in particular to a family doctor intelligent service system and service method based on big data. Background Art

[0002] Community smart medical services are gradually becoming an important way to improve the health and well-being of residents. Through intelligent diagnosis and decision-making support systems, residents can enjoy more accurate disease diagnosis and treatment plans; health management and monitoring systems use Internet of Things technology to monitor residents' vital signs in real time and upload them to the cloud for in-depth analysis, providing residents with customized health management plans.

[0003] Telemedicine services break geographical restrictions, allowing residents to enjoy consultation and guidance from professional doctors at home, while health education and publicity popularize health knowledge and enhance residents' health awareness through smart terminals and social media platforms.

[0004] A one-stop family doctor service system and method based on blockchain is disclosed in application publication number CN117393125A, including: the present invention communicates with the user module and the family doctor module respectively through the smart contract module to complete the information interaction between the family doctor module and the user module; at the same time, the user module and the family doctor module are used to achieve information transparency, trust establishment, efficiency improvement and other goals. Users use digital identities and wallets to achieve identity authentication and payment settlement, and communicate and collaborate through the user end and the family doctor end; users and family doctors can evaluate the satisfaction of information feedback services, and motivate both parties to improve the level and reputation of services through rewards or penalties. The present invention can effectively solve the problems of information opacity, lack of trust and low efficiency between family doctors and residents.

[0005] Combined with the above application and the contents of the prior art:

[0006] Considering that high-risk patients have poor health conditions and their health fluctuates greatly, they may need frequent or regular medical treatment. However, due to the limited medical load of high-level hospitals, it may be difficult to fully accept all the patients. In this case, it is necessary to set up some family doctors in the radiation area of ​​the high-level hospitals. These family doctors will provide initial consultation for patients with mild symptoms or non-emergency issues. If the situation is difficult to handle, these high-risk patients will be guided to make appointments at high-level hospitals. Furthermore, in order to reduce the load on family doctors, a telemedicine platform can be introduced. Based on the collection of various health data of high-risk patients, the telemedicine platform can intervene in the diagnosis to improve the efficiency of diagnosis and treatment.

[0007] However, in the existing intelligent service methods of family doctors, if there are many high-risk patients in the target area at the current stage and the degree of concentration is high, resulting in a large hospital reception load, when personalized health management is needed for high-risk patients and emergency medical response is provided, before the telemedicine platform or high-level hospital intervenes, due to the lack of classification of high-risk patients and the lack of understanding of their specific health status and disease level, the allocation of medical resources lacks specificity. All high-risk patients are received and queued in a unified manner, resulting in the inability to quickly deal with some high-risk patients. Therefore, there are certain medical safety risks.

[0008] To this end, the present invention provides a family doctor intelligent service system and service method based on big data. Summary of the invention

[0009] 1. Technical issues to be resolved

[0010] In view of the deficiencies of the prior art, the present invention provides a family doctor intelligent service system and service method based on big data, which conducts health assessments on high-risk patients and provides corresponding guidance to high-risk patients in combination with their health status; when abnormal conditions are identified in high-risk patients, the diagnosis and treatment-related data of high-risk patients are sent to the telemedicine platform, and the telemedicine platform outputs the diagnosis results; the condition recognition model is used to evaluate the condition of high-risk patients and determine the corresponding condition level. When high-risk patients enter the registration appointment state, the corresponding appointment queuing strategy is selected according to the condition level; the appointment order of high-risk patients is optimized, and after the diagnosis state is ended, the health management knowledge map provides personalized health management plans for high-risk patients. The appointments of different mild patients are optimized and adjusted, so that the appointments for consultations are more reasonable and the waste of medical resources is avoided, thereby solving the technical problems raised in the background technology.

[0011] (II) Technical solution

[0012] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0013] The big data-based intelligent family doctor service method includes, after identifying high-risk patients based on medical records, constructing the concentration of high-risk patients in the hospital radiation area, and if the concentration exceeds expectations, collecting the diagnosis and treatment and related data of high-risk patients;

[0014] Real-time monitoring of the health data of high-risk patients, the health assessment model conducts health assessment on high-risk patients, and provides corresponding guidance to high-risk patients based on their health status;

[0015] When abnormal conditions are identified in high-risk patients, the diagnosis and treatment data of high-risk patients are sent to the telemedicine platform, which then outputs the diagnosis results;

[0016] Use the condition recognition model to evaluate the condition of high-risk patients and determine the corresponding condition level. When high-risk patients enter the registration appointment state, select the corresponding appointment queuing strategy based on the condition level;

[0017] The appointment order of high-risk patients is optimized, and after they finish the diagnosis state, the health management knowledge graph provides personalized health management plans for high-risk patients.

[0018] Furthermore, patient information is collected based on the patient medical records in the hospital, high-risk patients are identified and corresponding location information is obtained, and the location information is marked on an electronic map;

[0019] Based on the location information of different high-risk patients, the aggregation of high-risk patients in the radiation area is constructed. If the aggregation degree exceeds the aggregation threshold, a health record construction instruction is issued to the high-risk patients in the radiation area.

[0020] Furthermore, after receiving the instruction to build a health record, the diagnosis and treatment and related data of high-risk patients in the radiation area are collected, including at least the identity information, basic information and medical records of high-risk patients, and the patient's electronic health record is built by collecting the diagnosis and treatment and related data.

[0021] Furthermore, wearable devices are used to monitor the real-time health data of high-risk patients. The collected health data is used as input, and the trained health assessment model is used to conduct health assessments on high-risk patients to obtain health scores for high-risk patients; and the patient's disease type is identified based on the patient's health data.

[0022] Furthermore, after obtaining the health scores of high-risk patients at various time points, a health value is constructed based on the health scores of high-risk patients. If the health value of the high-risk patients is lower than the health threshold, an emergency treatment instruction is issued to the outside; if the health value of the high-risk patients is not lower than the health threshold, a guidance instruction for medical treatment is issued to the outside.

[0023] Furthermore, after receiving the emergency treatment order, the health data of high-risk patients, including physiological indicators and behavioral data, are monitored and collected. The health data is used as input and the trained abnormal indicator recognition model is used to identify abnormal situations. If an abnormal situation is identified in a high-risk patient, an abnormal reminder instruction is issued.

[0024] Furthermore, after receiving the abnormal reminder instruction, the health data of high-risk patients, the patient's electronic health records and the current abnormal situation are sent to the telemedicine platform, which intervenes in the diagnosis and outputs the diagnosis results:

[0025] If the high-risk patient is a mild patient, the telemedicine platform will provide the high-risk patient with a treatment plan and give guidance on medical treatment; if the high-risk patient is a severe patient, the telemedicine platform will share the high-risk patient's real-time location information, health data, condition description and electronic health records with emergency personnel.

[0026] Furthermore, if a patient with a mild illness receives a medical consultation instruction, the patient with a mild illness is guided to make a registration request to the hospital;

[0027] After receiving the registration request, the pre-trained collaborative filtering algorithm matches the attending physician for high-risk patients based on the correspondence between the health data, electronic health records and symptom description information of mild patients and the doctor's areas of expertise and diagnostic experience.

[0028] Furthermore, the health data, electronic health records and symptom description information of mild patients are used as input, and the trained condition recognition model is used to evaluate the condition of mild patients to obtain the corresponding condition level, including emergency patients, moderate patients and mild patients, and the mild patients are marked according to the condition level.

[0029] Furthermore, after the mildly ill patient enters the registration appointment state, the disease level of the mildly ill patient is identified and a corresponding appointment queuing strategy is selected for the mildly ill patient, including:

[0030] If a patient with mild symptoms becomes an emergency patient, the emergency patient will be automatically queued; if a patient with mild symptoms becomes a moderate patient, the number of people in the queue in front will be limited when the moderate patient is in the queue; if a patient with mild symptoms becomes a mild patient, they will queue according to the appointment time.

[0031] Furthermore, the location information of patients with mild symptoms is collected. If the patients with mild symptoms have not arrived at the hospital and are in a queue, the time when the patients arrived at the hospital is obtained;

[0032] According to the actual workload of doctors, the arrival time and condition level of patients with mild symptoms, the optimization goal is to improve the efficiency of patient reception. The pre-trained genetic algorithm is used to optimize the appointment order of patients with mild symptoms, obtain the optimized appointment order and assign corresponding appointment numbers.

[0033] Furthermore, after the diagnosis status of mild and severe patients ends, the medical information, including symptoms, diagnosis and treatment plans and prescriptions, is recorded, and the electronic health records of mild patients are updated;

[0034] After keyword extraction from the updated electronic health records, the corresponding health characteristics are obtained;

[0035] Taking patient health management as the target word, a health management knowledge graph for high-risk patients is pre-built; based on the correspondence between health characteristics and health management plans, the health management knowledge graph provides personalized health management plans for mild patients, including: dietary advice, exercise plans, drug management, and lifestyle adjustments.

[0036] The big data-based family doctor intelligent service system includes a cluster analysis unit that identifies high-risk patients based on medical records and constructs the clustering of high-risk patients in the hospital radiation area. If the clustering exceeds expectations, the diagnosis and treatment and related data of high-risk patients are collected;

[0037] The data evaluation unit monitors the health data of high-risk patients in real time, uses the health evaluation model to evaluate the health of high-risk patients, and provides corresponding guidance to high-risk patients based on their health status;

[0038] The remote processing unit, when identifying abnormal conditions in high-risk patients, sends the diagnosis and treatment-related data of the high-risk patients to the telemedicine platform, which then outputs the diagnosis results;

[0039] The appointment unit uses the condition recognition model to evaluate the condition of high-risk patients and determine the corresponding condition level. When high-risk patients enter the registration appointment state, the corresponding appointment queuing strategy is selected according to the condition level;

[0040] The personalized management unit optimizes the appointment order of high-risk patients, and after they finish the diagnosis state, the health management knowledge graph provides personalized health management plans for high-risk patients.

[0041] (III) Beneficial effects

[0042] The present invention provides a family doctor intelligent service system and service method based on big data, which has the following beneficial effects:

[0043] 1. Determine the degree of demand for smart medical services in the region, and provide smart medical services when there is a corresponding demand, so as to achieve health protection for many high-risk patients; collect various information of high-risk patients and build electronic health records of patients, which can be used as a reference when high-risk patients are in the treatment status, and pay attention to high-risk patients by updating the electronic health records.

[0044] 2. Based on the real-time monitoring of various physical health indicators of high-risk patients through wearable devices, the health status of high-risk patients can be evaluated in real time, so that high-risk patients can understand their physical health status in real time and deal with abnormal situations in a timely manner.

[0045] 3. Comprehensively evaluate the health status of high-risk patients based on their health values, adopt different treatment methods for high-risk patients based on their health values, and allocate different medical resources to make the actual response more targeted, avoid lack of attention to high-risk patients, and reduce the waste of medical resources.

[0046] 4. When external reminders are issued when abnormal conditions occur in high-risk patients, high-risk patients can be treated in a timely manner when abnormalities occur, thereby ensuring the health status of high-risk patients; remote diagnosis and treatment of high-risk patients can be achieved through the intervention of the telemedicine platform, and high-risk patients can be divided into different types accordingly. While ensuring the health of patients, it can also reduce the pressure of subsequent medical intervention and the burden on hospitals.

[0047] 5. Assess and classify the conditions of mild patients through the trained disease recognition model, and adopt adaptive treatment methods for patients of different severity when medical treatment is needed; quickly handle mild patients by matching corresponding appointment queuing strategies for different mild patients, and ensure that patients with the highest demand are given priority; optimize and adjust the appointments for different mild patients based on the conditions of each patient and the corresponding appointment queuing strategies, so as to make the appointments more reasonable and avoid wasting medical resources.

[0048] 6. Provide corresponding health management plans through the health management knowledge map, provide personalized health management plans for high-risk patients after receiving the diagnosis, and refer to the personalized health management plans to accelerate the recovery of high-risk patients after they enter the rehabilitation stage. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 This is a flow chart of the family doctor intelligent service method of the present invention;

[0050] Figure 2 It is a schematic diagram of the structure of the family doctor intelligent service system of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] See also Figure 1 The present invention provides a family doctor intelligent service method based on big data, including:

[0053] Step 1: After identifying high-risk patients based on medical records, build the concentration of high-risk patients in the hospital radiation area , if the aggregation Exceed expectations and collect diagnosis and treatment data of high-risk patients;

[0054] The step 1 includes the following contents:

[0055] Step 101: After determining the radiation area of ​​the hospital, construct an electronic map covering the radiation area; collect patient information based on the patient medical records in the hospital, identify high-risk patients and obtain corresponding location information, and mark the location information on the electronic map; high-risk patients include patients with chronic diseases, elderly patients, high-risk groups, pregnant women and postpartum women, and patients recovering from surgery;

[0056] Under dimensionless conditions, the concentration of high-risk patients in the radiation area is constructed based on the location information of different high-risk patients. , as follows:

[0057] ;

[0058] Where: are the weights for geographically weighted regression, based on location and patient location Calculate and obtain, the value falls into ; is the median value of aggregation, The number of high-risk patients, The bandwidth parameter controls the degree of smoothing, and its value falls within , and Respectively represent the number of rows and columns of the grid;

[0059] Pre-set aggregation thresholds based on historical data and management expectations for high-risk patients;

[0060] If the degree of aggregation exceeds the aggregation threshold, it means that the number of high-risk patients may be large. In this scenario, if the high-risk patients seek medical examinations completely on their own, considering the negligence of the high-risk patients themselves and the load of the hospital during the consultation, the high-risk patients will face greater health risks. At this time, health record building instructions can be issued to high-risk patients in the radiation area through public channels;

[0061] When using it, after the introduction of smart medical services, the smart service area is pre-defined, and the location information and number of high-risk patients in the area are counted and recorded. By building a high-risk patient aggregation , according to the concentration It can judge the demand for smart medical services in the region and provide smart medical services when there is a corresponding demand, thus achieving health protection for many high-risk patients;

[0062] Step 102: After receiving the health record building instruction, collect the diagnosis and treatment and related data of high-risk patients in the radiation area, including the following:

[0063] Identity information, such as name, gender, age, ID number, contact information, emergency contact, etc.; basic information, such as past medical history, allergy information, family medical history and lifestyle habits (diet, smoking, drinking, etc.); medical records, such as electronic medical records, including past diagnosis and treatment records, examination results, imaging data, etc.; building patient electronic health records from the collection of diagnosis and treatment and related data;

[0064] When using, combine the contents in steps 101 and 102:

[0065] Before seeing a high-risk patient, various information about the high-risk patient is collected and an electronic health record of the patient is constructed. The electronic health record can be used as a reference when the high-risk patient is in the seeing state. At the same time, by updating the electronic health record, attention can be paid to high-risk patients.

[0066] However, in the existing intelligent service methods of family doctors, if there are many high-risk patients in the target area at the current stage and the degree of concentration is high, resulting in a large reception load for the hospital, if personalized health management is required for high-risk patients and emergency medical response is provided, there is a lack of classification of high-risk patients before the intervention of the telemedicine platform or high-level hospitals, and their specific health status and disease level are not understood. Therefore, there is a lack of specificity in the allocation of medical resources. All high-risk patients are received and queued for appointments in a unified manner, resulting in the inability to quickly deal with some high-risk patients. Therefore, there are certain medical safety risks.

[0067] Step 2: Real-time monitoring of the health data of high-risk patients, using the health assessment model to conduct health assessments on high-risk patients, and providing corresponding guidance to high-risk patients based on their health status;

[0068] The step 2 includes the following contents:

[0069] Step 201: A high-risk patient is bound to a wearable device, such as a smart watch and a heart rate monitor, and the wearable device is used to monitor the high-risk patient's real-time health data, including: physiological indicators, such as monitoring heart rate, blood pressure, blood sugar, body temperature, sleep quality, exercise volume, etc.; and behavioral data, such as eating habits, exercise frequency, sleep patterns, etc.; the collected physiological indicators and behavioral data are aggregated to generate a patient health indicator data set;

[0070] Step 202: training a convolutional neural network with the labeled sample data to obtain a trained health assessment model for high-risk patients; using the patient health data as input, using the trained health assessment model to perform a health assessment on the high-risk patients to obtain a health score for the high-risk patients; and identifying the patient's disease type, such as cardiovascular disease, hypertension, or diabetes, based on the patient's health data;

[0071] When using it, when smart medical services are needed for high-risk patients, wearable devices can be used to monitor the various health indicators of high-risk patients in real time, and the health status of high-risk patients can be evaluated in real time, so that high-risk patients can understand their health status in real time and deal with abnormal situations in a timely manner;

[0072] Step 203: After obtaining the health scores of high-risk patients at each time point, the health value constructed based on the health scores of high-risk patients , high-risk patients are marked with health values ​​as follows:

[0073] ;

[0074] Where: is the number of time nodes, For high-risk patients Health scores at each time point, is the weight coefficient, satisfying ; is the initial time node; For the weighted moving average part, in the weighted moving average calculation, represents the index of the lag period, for example, if is the current time node, Indicates hysteresis Data points at time nodes;

[0075] Based on historical data and health management expectations for high-risk patients, health thresholds are pre-set; if the health value of a high-risk patient is lower than the health threshold, it means that the current health status of the high-risk patient is poor and needs to be treated immediately. At this time, an emergency treatment instruction is issued to the outside world;

[0076] If the health value of the high-risk patient is not lower than the health threshold and the health status has not deteriorated further for the time being, then an instruction to guide the patient to seek medical treatment will be issued to the outside world;

[0077] When using, combine the contents in steps 201 to 203:

[0078] After periodically outputting the health scores of high-risk patients, corresponding health values ​​are generated based on the health scores of high-risk patients and their changing states, and a comprehensive assessment of the health status of high-risk patients is achieved based on the health values. Different treatment methods are adopted for high-risk patients based on the health values, and different medical resources are allocated to make the actual response more targeted, avoid lack of attention to high-risk patients, and reduce waste of medical resources.

[0079] Step 3: When an abnormal condition is identified in a high-risk patient, the diagnosis and treatment data of the high-risk patient is sent to the telemedicine platform, and the telemedicine platform outputs the diagnosis result;

[0080] The step three includes the following contents:

[0081] Step 301: Train a machine learning algorithm using the labeled sample data to obtain an abnormal indicator recognition model;

[0082] After receiving the emergency treatment order, the system monitors and collects the physiological indicators and behavioral data of high-risk patients in real time. The high-risk patients’ health data is used as input, and the trained abnormal indicator recognition model is used to identify abnormal conditions, such as arrhythmia, acute hypertension or hypoglycemia.

[0083] If abnormal conditions are identified in high-risk patients, abnormal reminder instructions will be issued through mobile applications, text messages or smart devices to notify the user, emergency contacts and family doctors;

[0084] When in use, when emergency treatment is needed for some high-risk patients, based on the collected physiological indicators and behavioral data, external reminders are issued when abnormal conditions occur in high-risk patients, ensuring that high-risk patients can be treated in time when abnormalities occur, thereby ensuring the health status of high-risk patients;

[0085] Step 302: After receiving the abnormal reminder instruction, the health data of the high-risk patient, the patient's electronic health record and the current abnormal situation are sent to the family doctor, and the family doctor responsible for the high-risk patient intervenes in the diagnosis through the telemedicine platform, and the telemedicine platform outputs the diagnosis result:

[0086] If the symptoms of high-risk patients are mild, they will be treated as mild patients, and the telemedicine platform will provide treatment plans and guidance for medical treatment to high-risk patients;

[0087] If the high-risk patient has a serious condition, he or she will be treated as a critically ill patient, and the telemedicine platform will share the high-risk patient's real-time location information, health data, condition description and electronic health records with emergency personnel;

[0088] When used, combine the contents in steps 301 and 302:

[0089] When the health status of high-risk patients becomes abnormal, the telemedicine platform can intervene in the diagnosis based on the collected data of high-risk patients, so as to realize remote diagnosis and treatment of high-risk patients, and classify high-risk patients into different types accordingly. While realizing the protection of patients' health, it can also reduce the pressure of subsequent medical intervention and the burden on hospitals.

[0090] Step 4: Use the condition recognition model to evaluate the condition of high-risk patients and determine the corresponding condition level. When high-risk patients enter the registration appointment state, select the corresponding appointment queuing strategy based on the condition level;

[0091] The step 4 includes the following contents:

[0092] Step 401: If the patient with a mild illness receives a medical consultation instruction, the patient with a mild illness is guided to send a registration request to the hospital through a mobile application or an online platform;

[0093] After the hospital receives the registration request, the pre-trained collaborative filtering algorithm matches the attending physician for the high-risk patient based on the correspondence between the health data, electronic health records and symptom description information of the mild patient and the doctor's expertise and diagnostic experience;

[0094] When in use, based on the acquisition of various information of high-risk patients, the treatment effect of high-risk patients can be guaranteed through data recommendation and matching of corresponding doctors;

[0095] Step 402: train a convolutional neural network with the labeled sample data to obtain a trained disease recognition model; use the health data, electronic health records and symptom description information of the mild patients as input, use the trained disease recognition model to evaluate the disease of the mild patients, obtain the corresponding disease level, including emergency patients, moderate patients and mild patients, and mark the mild patients according to the disease level;

[0096] When in use, after high-risk patients enter the diagnostic state and obtain the corresponding diagnostic information and symptom description information, the trained disease recognition model is used to evaluate and classify the conditions of mild patients, so that when medical treatment is needed, adaptive treatment methods can be adopted for patients of different degrees;

[0097] Step 403: After the mildly ill patient enters the registration appointment state, the disease level of the mildly ill patient is identified and a corresponding appointment queuing strategy is selected for the mildly ill patient, including:

[0098] If a mild patient is an emergency patient, the emergency patient will be automatically queued and given priority for emergency treatment; if a mild patient is a moderate patient, the number of people in the queue in front will be limited when the moderate patient is in the queue; if a mild patient is a mild patient, the queue will be arranged according to the appointment time;

[0099] When used, combine the contents in steps 401 to 403:

[0100] After obtaining the condition level of mild patients, if mild patients need to go to the hospital for treatment, by matching different mild patients with corresponding appointment queuing strategies, rapid processing of mild patients can be achieved, ensuring that patients with the highest needs can be seen first, avoiding the occurrence of safety hazards.

[0101] Step 5: Optimize the appointment order of high-risk patients, and after they finish the diagnosis state, the health management knowledge graph will provide personalized health management plans for high-risk patients;

[0102] The step five includes the following contents:

[0103] Step 501: Collect the location information of the mildly ill patient. If the mildly ill patient has not arrived at the hospital and enters the queue state, obtain the time when the patient arrives at the hospital through data communication;

[0104] Based on the actual workload of doctors, the arrival time and condition level of patients with mild symptoms, the optimization goal is to improve the efficiency of patient reception. The pre-trained genetic algorithm is used to optimize the appointment order of patients with mild symptoms, obtain the optimized appointment order and assign corresponding appointment numbers, and send appointment confirmation, queuing progress and consultation reminder notifications to patients with mild symptoms.

[0105] When in use, when patients with mild symptoms make online appointments, if there are several patients with mild symptoms making appointments for medical treatment, the appointments for different patients with mild symptoms will be optimized and adjusted based on the conditions of each patient and the corresponding appointment queuing strategy, so as to make the medical appointments more reasonable and avoid waste of medical resources.

[0106] Step 502: After the diagnosis of the mild and severe patients ends, the medical information, including symptoms, diagnosis and treatment plans and prescriptions, is recorded, and the electronic health records of the mild patients are updated; after keyword extraction from the updated electronic health records, the corresponding health characteristics are obtained;

[0107] Taking patient health management as the target word, after deep retrieval and entity relationship construction, a health management knowledge graph for high-risk patients is pre-built; based on the correspondence between health characteristics and health management plans, the health management knowledge graph provides personalized health management plans for mild patients, including: dietary advice, exercise plans, drug management, and lifestyle adjustments.

[0108] When using, combine the contents in steps 501 and 502:

[0109] After the diagnosis of mild and severe patients is completed, the corresponding health management plan is given according to the patient's information through the health management knowledge map, so that personalized health management plans can be given to high-risk patients after receiving the diagnosis. With reference to the personalized health management plan, the recovery of high-risk patients can be accelerated after they enter the rehabilitation stage.

[0110] The specific method of building a knowledge graph for health management of high-risk patients can refer to the following:

[0111] Clarify the goals and scope of the knowledge graph

[0112] Determine the goal: clarify the specific problems that the knowledge graph aims to solve, such as identification of high-risk patients, prediction of disease progression, recommendation of personalized treatment plans, etc. Define the scope: define the medical fields, patient groups, data types, etc. that the knowledge graph needs to cover to ensure the pertinence and practicality of the knowledge graph.

[0113] Knowledge Collection

[0114] Medical data collection: Collect medical records, examination results, medication records, living habits and other information of high-risk patients. External data source integration: Integrate external data sources such as medical literature, clinical guidelines, expert opinions, etc. to ensure the authority and comprehensiveness of the knowledge graph. Data cleaning and preprocessing: Clean the collected data, remove duplicate, erroneous or irrelevant information, and perform necessary preprocessing, such as data formatting and standardization.

[0115] Knowledge Representation

[0116] Ontology construction: Construct an ontology in the medical field to define medical concepts such as diseases, symptoms, and treatment methods, and the relationships between them. Entity and relationship extraction: Use natural language processing (NLP) technology to extract entities (such as patients, diseases, and drugs) and relationships (such as patient-disease, disease-symptom, and drug-treatment relationships) from text data. Knowledge graph construction: Based on the extracted entities and relationships, construct the nodes and edges of the knowledge graph to form a directed or undirected graph structure.

[0117] Knowledge Fusion

[0118] Multi-source data fusion: Fusion of data from different sources to resolve conflicts and redundancy issues between data. Semantic alignment: Use semantic technology to achieve semantic alignment between different data sources to ensure the consistency of concepts in the knowledge graph.

[0119] Graph optimization and update

[0120] Quality assessment: Conduct quality assessment on the constructed knowledge graph, including data accuracy, completeness, consistency, etc. Continuous optimization: Optimize the knowledge graph based on the evaluation results, including correcting errors, supplementing missing information, etc. Regular updates: Extract knowledge and information from new data sources regularly, update the knowledge graph, and ensure its timeliness and accuracy.

[0121] Application and Feedback

[0122] Application scenario development: Develop health management applications for high-risk patients based on knowledge graphs, such as disease prediction, personalized treatment plan recommendations, etc. User feedback collection: Collect user feedback on the application to understand the effects and problems of the knowledge graph in actual applications. Iterative optimization: Iteratively optimize the knowledge graph based on user feedback and application effects to continuously improve its performance and practicality.

[0123] See also Figure 2 The present invention provides a family doctor intelligent service system based on big data, including:

[0124] The cluster analysis unit identifies high-risk patients based on their medical records and constructs the clustering of high-risk patients within the hospital's radiation area. If the clustering exceeds expectations, the diagnosis and treatment and related data of high-risk patients are collected;

[0125] The data evaluation unit monitors the health data of high-risk patients in real time, uses the health evaluation model to evaluate the health of high-risk patients, and provides corresponding guidance to high-risk patients based on their health status;

[0126] The remote processing unit, when identifying abnormal conditions in high-risk patients, sends the diagnosis and treatment-related data of the high-risk patients to the telemedicine platform, which then outputs the diagnosis results;

[0127] The appointment unit uses the condition recognition model to evaluate the condition of high-risk patients and determine the corresponding condition level. When high-risk patients enter the registration appointment state, the corresponding appointment queuing strategy is selected according to the condition level;

[0128] The personalized management unit optimizes the appointment order of high-risk patients, and after they finish the diagnosis state, the health management knowledge graph provides personalized health management plans for high-risk patients.

[0129] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0130] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0131] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0132] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only some logical function divisions. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0133] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0134] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0135] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program codes.

[0136] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A family doctor intelligent service method based on big data, characterized by: include, After identifying high-risk patients based on their medical records, the concentration of high-risk patients in the hospital radiation area is constructed. Exceeded expectations, collected diagnosis and treatment data of high-risk patients and related data; among them, based on the location information of different high-risk patients, constructed the concentration of high-risk patients in the radiation area , as follows: ； Where: is the weight of the geographically weighted regression, based on the location of the sample points and patient location Calculate, and the value falls into ; is the median value of aggregation, The number of high-risk patients, The bandwidth parameter controls the degree of smoothing, and its value falls within , and Respectively represent the number of rows and columns of the grid; Real-time monitoring of the health data of high-risk patients, the health assessment model conducts health assessment on high-risk patients, and provides corresponding guidance to high-risk patients based on their health status; , high-risk patients are marked with health values ​​as follows: ； Where: is the number of time nodes, For high-risk patients Health scores at each time point, is the weight coefficient, satisfying ; is the initial time node; For the weighted moving average part, in the weighted moving average calculation, An index representing the lag period; When abnormal conditions are identified in high-risk patients, the diagnosis and treatment data of high-risk patients are sent to the telemedicine platform, which then outputs the diagnosis results; Use the condition recognition model to evaluate the condition of high-risk patients and determine the corresponding condition level. When high-risk patients enter the registration appointment state, select the corresponding appointment queuing strategy based on the condition level; The appointment order of high-risk patients is optimized, and after they finish the diagnosis state, the health management knowledge graph provides personalized health management plans for high-risk patients; Collect patient information based on the patient medical records in the hospital, identify high-risk patients and obtain corresponding location information, and then mark the location information on the electronic map; build the concentration of high-risk patients in the radiation area based on the location information of different high-risk patients. If the concentration exceeds the concentration threshold, issue a health record construction instruction to the high-risk patients in the radiation area; After receiving the health record construction instruction, the diagnosis and treatment and related data of high-risk patients in the radiation area are collected, including at least the identity information, basic information and medical records of high-risk patients, and the patient electronic health record is built by collecting the diagnosis and treatment and related data; Use wearable devices to monitor the real-time health data of high-risk patients. Use the collected health data as input and use the trained health assessment model to conduct health assessments on high-risk patients to obtain the health scores of high-risk patients. Identify the type of disease of the patient based on the patient's health data. After obtaining the health scores of high-risk patients at various time points, a health value is constructed based on the health scores of high-risk patients. If the health value of the high-risk patient is lower than the health threshold, an emergency treatment instruction is issued to the outside; if the health value of the high-risk patient is not lower than the health threshold, a guidance instruction for medical treatment is issued to the outside; After receiving the emergency treatment order, the health data of high-risk patients, including physiological indicators and behavioral data, are monitored and collected. The health data is used as input and the trained abnormal indicator recognition model is used to identify abnormal conditions. If an abnormal condition is identified in a high-risk patient, an abnormal reminder order is issued. After receiving the abnormal reminder instruction, the health data of high-risk patients, the patient's electronic health records and the current abnormal situation are sent to the telemedicine platform, which then intervenes in the diagnosis and outputs the diagnosis results: If the high-risk patient is a mild case, the telemedicine platform will provide the high-risk patient with a treatment plan and give instructions for medical treatment; if the high-risk patient is a severe case, the telemedicine platform will share the high-risk patient's real-time location information, health data, condition description and electronic health records with emergency personnel; If a mild patient receives a medical consultation instruction, the patient will be guided to make a registration request to the hospital. After receiving the registration request, the pre-trained collaborative filtering algorithm will match the attending physician for the high-risk patient based on the correspondence between the mild patient's health data, electronic health records and symptom description information and the doctor's expertise and diagnostic experience. Using the health data, electronic health records and symptom description information of mild patients as input, the trained condition recognition model is used to evaluate the condition of mild patients and obtain the corresponding condition level, including emergency patients, moderate patients and mild patients, and mild patients are marked according to the condition level; After patients with mild symptoms enter the registration appointment status, the condition level of the patients is identified and the corresponding appointment queuing strategy is selected for the patients with mild symptoms, including: if the patients with mild symptoms are emergency patients, the emergency patients are automatically queued; if the patients with mild symptoms are moderate patients, when the moderate patients are in the queue, the number of people in the queue in front is limited; if the patients with mild symptoms are mild patients, they are queued according to the appointment time.

2. The big data-based family doctor intelligent service method according to claim 1 is characterized by: Collect the location information of patients with mild symptoms. If the patients with mild symptoms have not arrived at the hospital and enter the queue, obtain the time when the patients arrived at the hospital. Based on the actual workload of the doctor, the arrival time of the patients with mild symptoms and the level of their condition, take improving the efficiency of patient reception as the optimization goal, use the pre-trained genetic algorithm to optimize the appointment order of patients with mild symptoms, obtain the optimized appointment order and assign the corresponding appointment number.

3. The family doctor intelligent service method based on big data according to claim 2 is characterized by: After the diagnosis status of mild and severe patients ends, record the medical information, including symptoms, diagnosis and treatment plans and prescriptions, and update the electronic health records of mild patients; After keyword extraction from the updated electronic health records, the corresponding health characteristics are obtained; Taking patient health management as the target word, pre-build the health management knowledge map for high-risk patients; Health management plans include: dietary advice, exercise plans, medication management, and lifestyle adjustments.

4. A family doctor intelligent service system based on big data, applying the intelligent service method according to any one of claims 1 to 3, characterized in that: include, The cluster analysis unit identifies high-risk patients based on their medical records and constructs the clustering of high-risk patients within the hospital's radiation area. If the clustering exceeds expectations, the diagnosis and treatment and related data of high-risk patients are collected; The data evaluation unit monitors the health data of high-risk patients in real time, uses the health evaluation model to evaluate the health of high-risk patients, and provides corresponding guidance to high-risk patients based on their health status; The remote processing unit, when identifying abnormal conditions in high-risk patients, sends the diagnosis and treatment-related data of the high-risk patients to the telemedicine platform, which then outputs the diagnosis results; The appointment unit uses the condition recognition model to evaluate the condition of high-risk patients and determine the corresponding condition level. When high-risk patients enter the registration appointment state, the corresponding appointment queuing strategy is selected according to the condition level; The personalized management unit optimizes the appointment order of high-risk patients, and after they finish the diagnosis state, the health management knowledge graph provides personalized health management plans for high-risk patients.

Images