Claim settlement method and device, computer equipment and storage medium

Abstract

The invention belongs to the field of artificial intelligence and financial science and technology, and relates to a claim settlement method which comprises the following steps: analyzing an electronic medical record by adopting a trained semantic analysis model to obtain disease information; standardizing the disease information, and constructing initial multi-dimensional disease features; generating medication suggestions and side effect monitoring schemes according to the online inquiry records and the electronic prescription information, and integrating the medication suggestions and the side effect monitoring schemes into the initial multi-dimensional disease features to obtain multi-dimensional disease features; and analyzing the obtained insurance product to obtain a corresponding claim settlement rule, generating a personalized claim settlement scheme according to the claim settlement rule and the multi-dimensional disease features, and automatically executing a claim settlement process according to the personalized claim settlement scheme. The invention further provides a claim settlement device, computer equipment and a storage medium. In addition, the invention also relates to a block chain technology, and the electronic medical record can be stored in a block chain. The claim settlement result is accurately judged, a more personalized claim settlement scheme is provided, and the claim settlement accuracy and efficiency are improved.

Description

Technical Field

[0001] The present application relates to the fields of artificial intelligence and financial technology, and in particular to a claims settlement method, apparatus, computer equipment and storage medium. Background Art

[0002] In the current insurance industry, with the rapid development of information technology and the popularization of the Internet, online claims services have become an important trend to improve insurance claims efficiency and optimize user experience. However, in the field of online claims involving medical services, although some progress has been made, it still faces many technical bottlenecks and challenges, which limit its further development and application.

[0003] First of all, the cumbersome offline medical treatment process is one of the key factors that restrict the improvement of online claims efficiency. Under the traditional model, users need to go to the medical institutions designated by the insurance company for diagnosis and treatment in person, including registration, queuing, consultation and drug purchase, which is not only time-consuming and labor-intensive, but also particularly inconvenient for users in remote locations or with limited mobility. In addition, after the diagnosis and treatment, users are required to keep and organize relevant invoices, medical records and other claims materials, and then initiate reimbursement applications through online channels. The entire process is complicated and inefficient.

[0004] Secondly, the high threshold of user cognition is another major obstacle to the popularization of online claims services. Insurance claims terms are usually highly professional and difficult for non-professional users to understand, which can easily lead to misunderstandings or disputes during the claims process. At the same time, some insurance companies lack transparency in the claims process and fail to clearly and comprehensively inform users of the claims process, required materials, and compensation standards, further increasing users' uncertainty and doubts.

[0005] Furthermore, the complex claims process and strict material requirements are also important issues facing online claims services. Although online claims have been greatly simplified compared to the traditional offline model, users still need to submit materials and fill in information according to the established process, which is cumbersome and error-prone. At the same time, online claims often have more stringent requirements for materials. Users need to ensure that the materials are complete and accurate in order to pass the review, otherwise they will face problems such as repeated submissions and delayed claims. In addition, most online claims services require users to advance medical expenses and receive the claim payment only after submitting all claim procedures, which undoubtedly increases the user's financial burden and time cost.

[0006] Finally, the general and lack of personalization of claims standards is an important factor affecting user satisfaction. Some insurance companies fail to fully consider the actual needs and medical conditions of different users when formulating claims standards, resulting in standards that are too general or vague and difficult to accurately grasp in actual operations. This generalized claims standard not only increases the complexity and uncertainty of claims, but may also lead to the inability to meet the reasonable demands of some users, thereby affecting their overall satisfaction and loyalty.

[0007] In summary, the online claims service for medical services in the existing technology still has many shortcomings in terms of improving user experience, simplifying the claims process, lowering the user's cognitive threshold, and realizing the personalization of claims standards. Summary of the invention

[0008] The purpose of the embodiments of the present application is to propose a claims settlement method, apparatus, computer equipment and storage medium to solve the technical problems in the prior art that the claims settlement process of online claims settlement services is complex, difficult to control, the claims settlement efficiency is low, and the user experience is poor.

[0009] In order to solve the above technical problems, the present application embodiment provides a claims settlement method, which adopts the following technical solutions:

[0010] Obtain the target patient's electronic medical record, parse the electronic medical record using the trained semantic analysis model, and extract disease information;

[0011] The disease information is converted into a standardized disease description through a preset medical term mapping rule library, and the initial multi-dimensional disease characteristics of the patient are constructed according to the disease description;

[0012] Obtaining the online consultation record and electronic prescription information of the target patient, and generating medication recommendations and side effect monitoring plans based on the online consultation record and the electronic prescription information;

[0013] Integrating the medication recommendation and the side effect monitoring program into the initial multidimensional disease signature to obtain a multidimensional disease signature;

[0014] Obtaining the insurance product corresponding to the target patient, parsing the insurance product to obtain the corresponding claim rules, and generating a personalized claim plan based on the multi-dimensional disease characteristics and the claim rules;

[0015] Through the preset smart contract, the claims process is automatically executed according to the personalized claims plan.

[0016] In order to solve the above technical problems, the embodiment of the present application further provides a claims settlement device, which adopts the following technical solution:

[0017] An acquisition module is used to acquire the electronic medical records of the target patient, parse the electronic medical records using the trained semantic analysis model, and extract disease information;

[0018] A standardization module, used for converting the disease information into a standardized disease description through a preset medical term mapping rule library, and constructing an initial multi-dimensional disease feature of the patient according to the disease description;

[0019] A medication plan generation module, used to obtain the online consultation record and electronic prescription information of the target patient, and generate medication recommendations and side effect monitoring plans based on the online consultation record and the electronic prescription information;

[0020] An integration module, used for integrating the medication recommendation and the side effect monitoring scheme into the initial multi-dimensional disease feature to obtain a multi-dimensional disease feature;

[0021] A claim settlement plan generation module, used to obtain the insurance product corresponding to the target patient, parse the insurance product to obtain the corresponding claim settlement rules, and generate a personalized claim settlement plan according to the multi-dimensional disease characteristics and the claim settlement rules;

[0022] The claims settlement module is used to automatically execute the claims settlement process according to the personalized claims settlement plan through a preset smart contract.

[0023] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0024] The computer device includes a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the claim settlement method described above when executing the computer-readable instructions.

[0025] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0026] The computer-readable storage medium stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the steps of the claim settlement method described above are implemented.

[0027] Compared with the prior art, this application has the following beneficial effects:

[0028] The present application provides a claims method, which can improve the efficiency and accuracy of information extraction by parsing electronic medical records with a trained semantic analysis model to obtain disease information; normalize disease information and construct initial multi-dimensional disease features to ensure the accuracy and consistency of information and facilitate information management and application; generate medication recommendations and side effect monitoring plans based on online consultation records and electronic prescription information, and integrate them into the initial multi-dimensional disease features to obtain multi-dimensional disease features, which can improve the accuracy of disease diagnosis, improve the accuracy of claims, speed up claims processing, and improve claims efficiency; obtain corresponding claims rules by parsing insurance products, generate personalized claims plans based on claims rules and multi-dimensional disease features, and automatically execute claims processes based on personalized claims plans, which can accurately judge claims results, provide more personalized claims plans, reduce manual review and processing links, enhance the fairness of claims, increase claims speed, and improve user experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0030] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0031] Figure 2 is a flow chart of an embodiment of a claims settlement method according to the present application;

[0032] Figure 3 is a schematic structural diagram of an embodiment of a claims settlement device according to the present application;

[0033] Figure 4 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0034] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0035] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0036] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0037] like Figure 1 As shown, the system architecture 100 may include a terminal device 101, a network 102 and a server 103. The terminal device 101 may be a laptop 1011, a tablet computer 1012 or a mobile phone 1013. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links or optical fiber cables.

[0038] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0039] The terminal device 101 can be any electronic device with a display screen and supporting web browsing. In addition to a laptop computer 1011, a tablet computer 1012 or a mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV), a laptop computer, a desktop computer, etc.

[0040] The server 103 may be a server that provides various services, such as a background server that provides support for web pages displayed on the terminal device 101 .

[0041] It should be noted that the claims settlement method provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the claims settlement device is generally set in the server / terminal device.

[0042] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0043] Continue to refer Figure 2 , shows a flow chart of an embodiment of a claims settlement method according to the present application, comprising the following steps:

[0044] Step S201, obtain the electronic medical record of the target patient, use the trained semantic analysis model to parse the electronic medical record, and extract disease information.

[0045] The target patients are users who consult online or purchase medicines online. The target patients are identified and their complete electronic medical records are retrieved from the electronic medical record system according to their identity, which may include outpatient medical records, inpatient medical records, examination reports, doctor's orders and other types of medical record data.

[0046] Obtain the target patient's electronic medical record and pre-process it, including noise removal, format standardization, etc. Input the pre-processed electronic medical record into the trained semantic analysis model, which will deeply analyze the electronic medical record and identify key information related to the disease as disease information, including medical terms and diagnostic information. Medical terms include disease names, symptoms and signs, drug names, etc.

[0047] It should be emphasized that in order to further ensure the privacy and security of electronic medical records, the above-mentioned electronic medical records can also be stored in a blockchain node.

[0048] The blockchain referred to in this application is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0049] It should be understood that this application uses blockchain technology to build a secure, transparent, and tamper-proof data sharing platform to achieve secure data exchange between patients, medical institutions, and insurance companies.

[0050] In this embodiment, the semantic analysis model can be obtained by training existing models such as BERT, Transformer, GPT (Generative Pre-trained Transformer), etc., or it can be pre-built according to the task.

[0051] In some optional implementations, an initial semantic analysis model is constructed, including an embedding layer, an entity extraction network, an encoder layer, and an output layer, and a historical medical record dataset is obtained to train the initial semantic analysis model, and a final semantic analysis model that meets expectations is output.

[0052] In this embodiment, the steps of using the trained semantic analysis model to parse the electronic medical record and extract disease information include:

[0053] Input the electronic medical records into the semantic analysis model, which includes an embedding layer, an entity extraction network, an encoder layer, and an output layer;

[0054] The text data in the electronic medical record is vectorized through the embedding layer to obtain the text embedding vector, and the text embedding vector is input into the entity extraction network;

[0055] Entities are extracted from the text embedding vector through the entity extraction network to obtain entity relationship triples, and the entity relationship triples and text data are input into the encoder layer;

[0056] The semantic and contextual features of entity relationship triples and text data are extracted through the attention mechanism of the encoder layer to obtain a semantic feature vector.

[0057] The semantic feature vector is calculated through the output layer to output disease information.

[0058] In this embodiment, the embedding layer is used to convert the text data in the electronic medical record into vectors, including position embedding, word embedding and character embedding of the text data, and correspondingly obtaining position embedding vectors, word embedding vectors and character embedding vectors, and adding the position embedding vectors, word embedding vectors and character embedding vectors to obtain text embedding vectors. Among them, position embedding provides information about the position of words in a sentence, which helps to understand the grammatical structure and contextual relationship of a sentence; word embedding captures the semantic information of vocabulary, so that the model can understand the meaning of vocabulary in context; character embedding captures character-level information, which helps the model process unknown vocabulary or spelling variants, and improves the model's ability to process complex data by incorporating character-level information into the model.

[0059] In the medical field, entities refer to specific pieces of information in text that have clear meanings, such as disease names, symptoms, drug names, dosages, medical procedures, body parts, etc. These entities are crucial for understanding and processing medical record text data, and also play an important role in identifying the intent of medical record text data.

[0060] In this embodiment, the entity extraction network uses the CasRel algorithm to identify and extract causal relationships, parallel relationships, etc. in medical record texts, including causal relationships between entities such as symptoms, diseases, treatments, and drugs. For example, the relationship between the use of a certain drug and specific side effects, or the relationship between a certain drug and the efficacy of a disease.

[0061] The entity extraction network includes a head entity recognition layer and a relationship and tail entity joint recognition layer. In some optional implementations of this embodiment, the above-mentioned step of extracting entities from text embedding vectors through the entity extraction network to obtain entity relationship triples includes:

[0062] The text embedding vector is input into the head entity recognition layer for recognition, and all head entities are output; the head entity and the text embedding vector are input into the relation and tail entity joint recognition layer for entity relation recognition, and all specific relations and tail entities corresponding to each head entity are obtained; the head entity, specific relation and tail entity are combined to obtain entity relation triples.

[0063] In this embodiment, the text embedding vector is decoded by the head entity recognition layer, and then the start position (starting position) and end position (ending position) of the entity are obtained to identify all possible head entities. When there are multiple head entity recognitions in the text, the matching principle of the closest start and end position marks is adopted to determine the scope of the entity, and this matching method ensures the integrity of the entity.

[0064] After identifying the head entity, the joint identification of the relation and tail entity is performed. The head entity and text embedding vector are input into the joint identification layer of the relation and tail entity for identification, and all specific relations and tail entities of each head entity are obtained. After the head entity and its corresponding specific relation and tail entity are identified, the entity relation triple generated is combined and expressed as (head entity, specific relation, tail entity).

[0065] In this embodiment, the encoder layer uses multiple stacked encoders to encode entity relationship triples and text data in turn through multiple layers of encoders to extract the semantic and contextual features of the medical record text data. The input of the latter encoder is the output of the previous encoder and the text data. After the encoder is processed, a deep semantic feature representation of the text data can be obtained, and the output of the last layer of encoders is usually used as a semantic feature vector.

[0066] Specifically, each encoder includes a multi-head self-attention layer and an encoding feedforward network layer. The entity relationship triples and text data are input into the multi-head self-attention layer for weight calculation to obtain the entity relationship semantic features; the entity relationship semantic features are optimized through the encoding feedforward network layer to obtain the semantic feature vector.

[0067] Finally, the probability of the semantic feature vector is calculated through the output layer to obtain the entity probability distribution representation, and the disease information is obtained based on the entity probability distribution representation.

[0068] In this embodiment, the output layer can adopt a CRF layer, or a design combining a fully connected layer and a softmax layer. When the output layer adopts a CRF layer, the CRF layer can effectively mine the relationship between texts, set constraints to ensure that the final prediction is valid, and the constraints can be automatically learned by the CRF layer when training data. The CRF layer uses Viterbi decoding to predict the label corresponding to the sequence, and uses joint probability to further optimize the overall effect of the prediction. When the output layer adopts a fully connected layer + softmax layer, the output semantic feature vector is mapped to a longer sequence through the fully connected layer, and the text at each position represents the score of the corresponding text. The score is converted into a probability distribution through the softmax layer, and a beam search algorithm (BeamSearch) is used to search and expand according to the probability distribution to obtain disease information.

[0069] Disease information extraction through semantic analysis models can identify key entities in medical record texts, enhance semantic understanding, capture complex relationships between entities, better extract deep semantic feature representations of medical record texts, and thus better capture long-distance dependencies in the texts, generate coherent and contextually relevant texts, and improve the accuracy and efficiency of disease information extraction.

[0070] Step S202, converting the disease information into a standardized disease description through a preset medical term mapping rule library, and constructing the patient's initial multi-dimensional disease characteristics based on the disease description.

[0071] In this embodiment, the identified disease information is standardized according to a preset medical term mapping rule base, mapped into normalized medical concepts, and the differences between synonyms and abbreviations are eliminated.

[0072] The steps for building the medical term mapping rule base are as follows:

[0073] 1) Collect medical terms from medical dictionaries, textbooks, professional guidelines, industry standards and authoritative databases. Medical terms cover diseases, symptoms, signs, diagnosis, treatment and other aspects; in particular, collect different expressions of the same medical concept in different medical terminology databases and medical texts to solve the problem of diversified expression;

[0074] 2) Establish a mapping relationship between non-standard medical terms and standard medical terms, where the mapping relationship should be based on medical norms and standards to ensure accuracy and authority;

[0075] 3) The established mapping relationships are organized into a rule base, which should include term pairs (non-standard terms-standard terms), synonym lists, contextual information, etc. The construction of the medical term mapping rule base has been completed.

[0076] The disease information is matched with the standardized terms in the medical term mapping rule library and mapped into standardized disease descriptions. For example, the disease information is structured according to dimensions such as disease name, symptoms and signs, severity, and medication information.

[0077] In this embodiment, the target patient's disease category is obtained by classification based on the disease description, such as "diabetes", "hypertension", etc. For each disease category, a multi-dimensional disease feature is constructed, including dimensions such as symptoms, signs, and test results, to form a disease feature set under this category. For example, for the disease category "diabetes", symptom features such as "polydipsia, polyuria, and fatigue" and test result features such as "fasting blood sugar ≥ 0mmol / L" can be extracted to form a disease feature set for diabetes. If the target patient suffers from multiple diseases at the same time, a feature fusion algorithm is used to fuse the features of different diseases to obtain a comprehensive multi-dimensional disease feature of the target patient.

[0078] Step S203, obtaining the target patient's online consultation record and electronic prescription information, and generating medication recommendations and side effect monitoring plans based on the online consultation record and electronic prescription information.

[0079] In this embodiment, multi-dimensional disease characteristics such as medication data, diagnostic information, and treatment plans are extracted from the obtained online consultation records and electronic prescription information, and the rationality analysis of medication data is performed. By comparing the standard medication plan and clinical guidelines, it is determined whether the actual medication of the target patient meets the specifications; an intelligent drug interaction detection algorithm is used to analyze the electronic prescription information, identify the potential drug interaction risks therein, and generate early warning information according to the risk level; based on the diagnostic information and treatment plan, combined with the medication pattern and treatment response data of the target patient, personalized medication recommendations are generated in real time. Based on medication data and diagnostic information, possible adverse drug reactions and side effects are identified. If potential adverse drug reactions and side effects are identified, a targeted side effect monitoring plan is generated based on the adverse drug reaction knowledge base.

[0080] Specifically, medication data, diagnostic information and treatment plans are extracted from online consultation records and electronic prescription information; through knowledge graph technology, medication data, diagnostic information and the treatment plan are semantically associated to build a personalized knowledge graph for the patient; node association relationships are identified in the personalized knowledge graph through a graph algorithm to obtain patient health characteristics; patient health characteristics are input into a trained drug recommendation model for drug screening, and medication recommendations are output; based on the drug information in the medication recommendations, an association rule mining algorithm is used to associate and analyze the drug information with the diagnostic information in the personalized knowledge graph to obtain association rules between drugs and diagnoses; based on the association rules between drugs and diagnoses, a preset adverse drug reaction knowledge base is queried to obtain adverse drug reactions, medication recommendations are updated based on the adverse drug reactions, and corresponding side effect monitoring plans are generated.

[0081] Keywords and entities in medication data, diagnostic information, and treatment plans are extracted through named entity recognition algorithms such as BiLSTM-CRF, including drug names, disease names, symptoms, etc.; based on the medical knowledge base and drug instructions, the semantic relationships between entities and keywords are identified and extracted through the TransR-based relationship extraction algorithm, such as "drug-treatment-disease", "drug-indication", "drug-ingredient", "disease-symptom", "disease-cause" and other relationships; the extracted entities and relationships are semantically associated through knowledge graph technology to build a personalized knowledge graph for patients. Among them, the nodes in the knowledge graph represent entities such as drugs, diseases, and symptoms, and the edges represent the relationships between entities, such as drug-indication, disease-symptom, etc. In addition, basic information of the target patient, such as age, gender, allergy history, etc., is obtained and associated with the personalized knowledge graph to enrich the content of the knowledge graph so that it can more comprehensively represent the patient's health status. If the target patient has multiple consultation records, the graph fusion algorithm is used to identify the correlation information between the personalized knowledge graphs of multiple consultations, such as changes in condition, medication adjustments, etc., and the correlation information is merged to construct the patient's complete medical history knowledge graph.

[0082] Use graph algorithms such as PersonalRank to analyze and mine personalized knowledge graphs, identify important nodes and association patterns in the knowledge graph, such as key drugs, diseases, common complications, etc., and combine them to obtain patient health characteristics.

[0083] In this embodiment, the drug recommendation model is based on a large language model, and uses natural language processing technology and deep learning algorithms to deeply analyze the patient's health characteristics to obtain medication recommendations. Specifically, the drug recommendation model includes a drug screening layer, a recommendation result generation layer, and an explanation layer. The drug screening layer extracts disease-related drug features from the patient's health features, and selects matching candidate drugs from existing drug databases (such as DrugBank, PharmGKB, and RxNorm) based on the drug features to generate a candidate drug recommendation list, where the drug database is an online resource for storing and retrieving drug-related information, which usually contains detailed information such as drug name, drug chemical structure, pharmacological action, indications, side effects, dosage, drug interaction, and drug category; the recommendation result generation layer uses collaborative filtering to recommend The algorithm sorts and screens the candidate drug list according to the medication preferences of other patients similar to the target patient to obtain personalized medication recommendation results, and then matches the personalized medication recommendation results with the electronic medical record data of the target patient to determine whether the recommended drug conflicts with or is contraindicated with the target patient's existing treatment plan. If there is a conflict or contraindication, the corresponding drug is removed from the candidate drug list and the personalized medication recommendation results are regenerated; the explanation layer can use a large language model to annotate the personalized medication recommendation results with drug instructions, and generate detailed drug instructions and potential risk warnings for drug use as medication recommendation outputs, which are easy for target patients to understand.

[0084] The Apriori association rule mining algorithm is used to associate the drug information in the medication recommendation with the diagnosis information in the personalized knowledge graph to obtain the association rules between the drug and the diagnosis. For example, the diagnosis information is "hypertension" and the corresponding drug is "nifedipine". According to the association rules between drugs and diagnoses, combined with the knowledge base of adverse drug reactions, it is judged whether the current medication recommendation for the target patient is likely to produce adverse reactions and side effects; if adverse reactions and side effects are likely to occur, the information of alternative drugs is obtained from the knowledge base of drug substitutions; by comparing the personalized knowledge graph of the target patient with the indication information of the alternative drugs, the alternative medication regimen that best suits the target patient is determined, and the drug information in the medication recommendation is updated according to the alternative medication regimen to obtain a new medication recommendation, and the corresponding side effect monitoring scheme is generated according to the knowledge base of adverse drug reactions.

[0085] In this embodiment, after obtaining the alternative medication regimen, the medication data in the personalized knowledge graph is updated.

[0086] In some optional implementations, a risk prediction model is used to assess the risk level of adverse drug reactions and side effects in target patients when adopting new medication recommendations; if the risk level exceeds a preset threshold, an early warning mechanism is triggered and the early warning information is sent to the attending physician and pharmacist; based on the early warning information and the specific situation of the target patient, the doctor and pharmacist determine whether the medication plan needs to be further adjusted, communicate with the target patient if necessary, determine the final adjustment plan, and update the target patient's personalized knowledge graph.

[0087] By generating medication recommendations and side effect monitoring plans, personalized treatment can be provided based on the patient's specific condition, medication data, past treatment responses and other personalized factors, making drug treatment more targeted and effective; by accurately matching drugs and patients, treatment effects can be improved, ineffective treatments and drug waste can be reduced, thereby reducing medical costs; at the same time, the occurrence of adverse reactions can be reduced, safe medication can be ensured, and the quality and efficiency of overall medical services can be improved.

[0088] Step S204, integrating medication recommendations and side effect monitoring plans into the initial multi-dimensional disease features to obtain multi-dimensional disease features.

[0089] In this embodiment, the medication recommendations and side effect monitoring schemes are converted into feature vectors and integrated into the initial multidimensional disease features to obtain a complete multidimensional disease feature.

[0090] The health records of target patients are constructed based on multi-dimensional disease characteristics, and the health records are encrypted and stored through blockchain technology to ensure data security and privacy protection, and to achieve secure sharing with medical institutions and insurance companies.

[0091] Step S205, obtain the insurance product corresponding to the target patient, parse the insurance product to obtain the corresponding claim rules, and generate a personalized claim plan based on the multi-dimensional disease characteristics and claim rules.

[0092] In this embodiment, all insurance products insured by the target patient are obtained based on the identification information of the target patient, the policy text corresponding to each insurance product is obtained, and features are extracted from the policy text to obtain key policy information; candidate claim rules in the key policy information are identified, and conflicts are detected for the candidate claim rules to obtain a consistent claim rule set; the consistent claim rule set is combined and optimized according to multi-dimensional disease characteristics to generate an optimal claim rule combination; and a personalized claim plan is generated based on the optimal claim rule combination.

[0093] Specifically, keyword extraction algorithms (such as TextRank, LDA, etc.) are used to extract key policy information from the policy text. The key policy information includes basic information (such as policy number, policyholder information, insured information, etc.), insurance contract terms, premiums, payment methods, etc. The candidate claims rules in the insurance contract terms are identified through a semantic understanding algorithm, and conflicts are detected for the candidate claims rules through a rule conflict detection algorithm to eliminate logical conflicts between rules and obtain a consistent set of claims rules. Among them, the semantic understanding algorithm can be extracted using a trained claims rule recognition model, and the claims rule recognition model can be trained based on a Transformer or BERT model.

[0094] According to business needs and rule characteristics, a rule conflict detection algorithm is pre-built according to the rule type (such as absolute rules, relative rules), rule hierarchy (such as priority relationship), rule interdependence, etc. The rule conflict detection algorithm is used to detect conflicts in candidate claim rules. The consistent claim rule set is optimized through multi-dimensional disease characteristics, and the claim rules that best match the multi-dimensional disease characteristics are selected as the target claim rules corresponding to the target patient to form the optimal claim rule combination. According to the optimal claim rule combination, a personalized claim plan is dynamically generated. The personalized claim plan includes the scope of claim, compensation ratio, deductible, etc. The personalized claim plan is converted into structured data, stored in the claim plan database, and associated with the target patient. When the patient consults or buys medicine in the online hospital, the patient's personalized claim plan is read from the claim plan database and displayed to the patient. After the patient confirms that he accepts the personalized claim plan, the corresponding claim service process is started, such as advance payment, direct compensation, etc., and the entire claim process is finally completed.

[0095] By generating personalized claims plans based on multi-dimensional disease characteristics and claims rules, we can more accurately meet the patient's protection needs, improve the accuracy of claims processing, and enhance user satisfaction. At the same time, we can automatically generate personalized claims plans, reduce the materials and procedures required, and make the claims process more convenient and efficient.

[0096] Step S206: automatically executing the claims process according to the personalized claims settlement plan through a preset smart contract.

[0097] In this embodiment, blockchain technology is used to build a decentralized medical data sharing network to achieve secure data exchange between patients, medical institutions (such as online hospitals) and insurance companies.

[0098] When the target patient conducts an online consultation or purchases medicine online, a personalized claims plan is generated and the preset smart contract is called for claim review. After the review is passed, the claim function is called through the smart contract to initiate a claim request to the claim terminal (such as an insurance company).

[0099] Specifically, a preset smart contract is used to determine whether the personalized claim plan meets the claim conditions; if the claim conditions are met, the target patient's medical data is encrypted to obtain an encrypted certificate, which is sent to the claim terminal for claim review, and the claim review result returned by the claim terminal is received; when the claim review result is approved, the claim amount is automatically calculated through the smart contract, and the claim process is executed according to the claim amount.

[0100] Among them, smart contracts are computer protocols designed to disseminate, verify or execute contracts in an information-based manner. The functions of smart contracts are as follows: through preset rules and algorithms, a preliminary review of claim applications can be conducted; when the claim conditions are met, the smart contract can automatically trigger the claim process without human intervention; according to the terms and conditions in the insurance contract, the smart contract can automatically calculate the amount of compensation to ensure the accuracy and fairness of the compensation; once the claim application is reviewed, the smart contract can immediately execute the compensation operation, greatly shortening the compensation time.

[0101] In order to protect patient privacy, zero-knowledge proof algorithms such as zk-SNARKs are used to encrypt the target patient's medical data, generate an encrypted proof proving the patient's claim eligibility, and submit the encrypted proof to the insurance company. The insurance company verifies the encrypted proof and confirms the patient's claim eligibility without obtaining the patient's specific medical information. If the claim is approved, the claim amount is automatically calculated through the smart contract, and the insurance company transfers the claim amount to the online hospital through the blockchain network to complete the payment of the claim. After receiving the claim fee, the online hospital automatically updates the target patient's claim record and synchronizes the claim result to the blockchain network for the target patient and the insurance company to query and confirm.

[0102] In this embodiment, medical data is extracted from a personalized claims plan.

[0103] Through on-chain management of patient medical data and insurance claims, mutual trust and collaboration among patients, hospitals, and claims are achieved, which improves claims efficiency and protects patient privacy and security.

[0104] This application uses a trained semantic analysis model to parse electronic medical records to obtain disease information, which can improve the efficiency and accuracy of information extraction; normalizes disease information and constructs initial multi-dimensional disease features to ensure the accuracy and consistency of information and facilitate information management and application; generates medication recommendations and side effect monitoring plans based on online consultation records and electronic prescription information, and integrates them into the initial multi-dimensional disease features to obtain multi-dimensional disease features, which can improve the accuracy of disease diagnosis, improve the accuracy of claims, and improve claims efficiency; obtains corresponding claims rules by parsing insurance products, generates personalized claims plans based on claims rules and multi-dimensional disease features, and automatically executes the claims process based on personalized claims plans, which can accurately judge the claims results, provide more personalized claims plans, reduce the links of manual review and processing, enhance the fairness of claims, increase the claims rate, and improve user experience.

[0105] In some optional implementations of this embodiment, after the step of generating a personalized claim settlement plan according to the multi-dimensional disease characteristics and claim settlement rules, the following steps are further included:

[0106] Using a preset similarity calculation algorithm, calculate the similarity scores between the personalized claims solution and each historical claims case in the historical claims case database;

[0107] Based on the similarity scores, a preset number of historical claims cases with the highest similarity are selected as reference cases;

[0108] Obtain the claims result data of the reference case, optimize and adjust the personalized claims plan based on the claims result data, and obtain the optimized personalized claims plan.

[0109] Among them, the preset similarity calculation algorithm can adopt commonly used vector similarity measurement methods such as Euclidean distance and cosine similarity, and sort the historical claims cases according to the calculated similarity scores to obtain the Top-N historical claims cases most similar to the personalized claims plan, where N is a positive integer.

[0110] Obtain the claim result data of the reference case, including the claim amount, the applicability of the claim terms, etc., and conduct statistical analysis on these claim result data; obtain the key indicator data such as the claim amount distribution of the reference case and the applicable frequency of each term through statistical analysis; compare the estimated claim amount of the personalized claim plan with the claim amount distribution of the reference case to determine whether the estimated amount is reasonable; if it deviates from the claim amount distribution range of the reference case, adjust the estimated claim amount of the personalized claim plan. Use the association rule mining algorithm to analyze the correlation between the claim terms in the reference case, and obtain the rules that affect the claim results of different clause combinations, which are association rules; match the clause combination of the personalized claim plan with the mined association rules to determine whether the personalized claim plan meets the association rules; if not, optimize the clause combination of the personalized claim plan according to the association rules; after adjusting and optimizing the estimated claim amount and clause combination, generate an optimized personalized claim plan.

[0111] Optionally, the optimized personalized claims settlement plan is compared with the original personalized claims settlement plan to evaluate the optimization effect. If the optimization effect is significant, the optimized personalized claims settlement plan is determined to be used; otherwise, the original personalized claims settlement plan is retained or further optimized.

[0112] This application can help better meet customer needs, further improve claims efficiency and accuracy, enhance market competitiveness, and achieve sustainable development by optimizing personalized claims solutions.

[0113] In some optional implementations, before the step of obtaining the insurance product corresponding to the target patient, the method further includes:

[0114] Obtain the insurance product characteristics of all insurance products, use a multi-dimensional matching algorithm to match the multi-dimensional disease characteristics with the insurance product characteristics, and obtain the matching degree between the target patient and each of the insurance products;

[0115] Screening out the insurance products whose matching degree is greater than or equal to a preset threshold to form a candidate insurance product set;

[0116] Through smart contract technology, a corresponding personalized insurance plan is generated for each candidate insurance product in the candidate insurance product set;

[0117] Recommend personalized insurance plans to target patients;

[0118] In response to the target patient's selection instruction for the personalized insurance plan, the candidate insurance product corresponding to the personalized insurance plan is used as the insurance product insured by the target patient.

[0119] In this embodiment, a multi-dimensional matching algorithm is used to match the multi-dimensional disease characteristics with the product characteristics of the insurance product to obtain an insurance product that matches the current disease characteristics of the target patient. The multi-dimensional matching algorithm can use a text-based matching algorithm (such as cosine similarity, Jaccard similarity) to calculate the matching score between the multi-dimensional disease characteristics and the product characteristics. If the matching score is greater than or equal to the preset threshold, the insurance product is used as a candidate insurance product. If the matching score is less than the preset threshold, the insurance product is discarded. According to the set of candidate insurance products, the target patient is judged whether it meets the underwriting conditions of the candidate insurance product through smart contract technology combined with the target patient's past medical history data; if it meets the underwriting conditions, the corresponding personalized insurance plan is generated according to the insurance terms, coverage, compensation ratio and other information of the candidate insurance product and pushed to the patient. The target patient can browse the details of the plan online. If the target patient accepts the plan, the insurance process is automatically triggered to complete the online insurance.

[0120] By matching the patient's multi-dimensional disease characteristics with the characteristics of the insurance product through a multi-dimensional matching algorithm, it is possible to achieve an accurate correspondence between the patient's characteristics and the insurance product, thereby ensuring the accuracy of the claims.

[0121] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0122] The basic technologies of artificial intelligence generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. Artificial intelligence software technology mainly includes computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning. A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it may include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium may be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0123] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0124] Further references Figure 3 , as a response to the above Figure 2 In order to realize the method shown in the figure, the present application provides an embodiment of a claims settlement device, and the device embodiment is Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0125] like Figure 3 As shown, the claim settlement device 300 described in this embodiment includes: an acquisition module 301, a standardization module 302, a medication plan generation module 303, an integration module 304, a claim settlement plan generation module 305 and a claim settlement module 306. Among them:

[0126] The acquisition module 301 is used to acquire the electronic medical record of the target patient, and parse the electronic medical record using the trained semantic analysis model to extract disease information;

[0127] The standardization module 302 is used to convert the disease information into a standardized disease description through a preset medical term mapping rule library, and construct an initial multi-dimensional disease feature of the patient according to the disease description;

[0128] The medication plan generation module 303 is used to obtain the online consultation record and electronic prescription information of the target patient, and generate medication suggestions and side effect monitoring plans according to the online consultation record and the electronic prescription information;

[0129] The integration module 304 is used to integrate the medication recommendation and the side effect monitoring scheme into the initial multi-dimensional disease feature to obtain a multi-dimensional disease feature;

[0130] The claim settlement plan generation module 305 is used to obtain the insurance product corresponding to the target patient, analyze the insurance product to obtain the corresponding claim settlement rules, and generate a personalized claim settlement plan according to the multi-dimensional disease characteristics and the claim settlement rules;

[0131] The claims module 306 is used to automatically execute the claims process according to the personalized claims solution through a preset smart contract.

[0132] It should be emphasized that in order to further ensure the privacy and security of electronic medical records, the above-mentioned electronic medical records can also be stored in a blockchain node.

[0133] Based on the above-mentioned claims settlement device 300, the efficiency and accuracy of information extraction can be improved; the accuracy and consistency of information can be ensured, and the management and application of information can be facilitated; at the same time, the accuracy of disease diagnosis can be improved, the accuracy of claims settlement can be improved, and the efficiency of claims settlement can be improved; in addition, according to the claims settlement rules and multi-dimensional disease characteristics, a personalized claims settlement plan is generated, and the claims settlement process is automatically executed according to the personalized claims settlement plan, which can accurately judge the claims settlement results, provide more personalized claims settlement plans, reduce the links of manual review and processing, enhance the fairness of claims settlement, improve the claims settlement rate, and enhance the user experience.

[0134] In some optional implementations, the acquisition module 301 includes:

[0135] An input submodule, for inputting the electronic medical record into the semantic analysis model, wherein the semantic analysis model includes an embedding layer, an entity extraction network, an encoder layer, and an output layer;

[0136] An embedding submodule, configured to perform vector encoding on the text data in the electronic medical record through the embedding layer to obtain a text embedding vector, and input the text embedding vector into the entity extraction network;

[0137] An entity extraction submodule, configured to perform entity extraction on the text embedding vector through the entity extraction network to obtain entity relationship triples, and input the entity relationship triples and the text data into the encoder layer;

[0138] An encoding submodule, configured to extract semantic and contextual features of the entity relationship triples and the text data through an attention mechanism of the encoder layer to obtain a semantic feature vector;

[0139] The output submodule is used to calculate the semantic feature vector through the output layer and output disease information.

[0140] Disease information extraction through semantic analysis models can identify key entities in medical record texts, enhance semantic understanding, capture complex relationships between entities, better extract deep semantic feature representations of medical record texts, and thus better capture long-distance dependencies in the texts, generate coherent and contextually relevant texts, and improve the accuracy and efficiency of disease information extraction.

[0141] In some optional implementations, the medication regimen generation module 303 includes:

[0142] An extraction submodule, for extracting medication data, diagnostic information and treatment plans from the online consultation record and the electronic prescription information;

[0143] A graph construction submodule, for semantically associating the medication data, the diagnostic information and the treatment plan through knowledge graph technology to construct a personalized knowledge graph for the patient;

[0144] An identification submodule, used to identify node association relationships on the personalized knowledge graph through a graph algorithm to obtain health characteristics of the patient;

[0145] A drug recommendation submodule, which is used to input the patient's health characteristics into the trained drug recommendation model to perform drug screening and output medication recommendations;

[0146] A mining submodule, for performing association analysis between the drug information in the medication recommendation and the diagnosis information in the personalized knowledge graph using an association rule mining algorithm to obtain association rules between drugs and diagnoses;

[0147] The generation submodule is used to query a preset adverse drug reaction knowledge base according to the association rules between the drug and the diagnosis, obtain the adverse drug reaction, update the medication recommendation according to the adverse drug reaction and generate a corresponding side effect monitoring plan.

[0148] By generating medication recommendations and side effect monitoring plans, personalized treatment can be provided based on the patient's specific condition, medication data, past treatment responses and other personalized factors, making drug treatment more targeted and effective; by accurately matching drugs and patients, treatment effects can be improved, ineffective treatments and drug waste can be reduced, thereby reducing medical costs; at the same time, the occurrence of adverse reactions can be reduced, safe medication can be ensured, and the quality and efficiency of overall medical services can be improved.

[0149] In some optional implementations of this embodiment, the claim settlement solution generating module 305 includes:

[0150] A feature extraction submodule is used to obtain the policy text of the insurance product, perform feature extraction on the policy text, and obtain key policy information;

[0151] A conflict detection submodule, used to identify candidate claim rules in the key information of the insurance policy, perform conflict detection on the candidate claim rules, and obtain a consistent claim rule set;

[0152] An optimization submodule, for combinatorially optimizing the consistent claim settlement rule set according to the multi-dimensional disease characteristics to generate an optimal claim settlement rule combination;

[0153] The claim settlement plan generation submodule is used to generate a personalized claim settlement plan based on the optimal claim settlement rule combination.

[0154] By generating personalized claims plans based on multi-dimensional disease characteristics and claims rules, we can more accurately meet the patient's protection needs, improve the accuracy of claims processing, and enhance user satisfaction. At the same time, we can automatically generate personalized claims plans, reduce the materials and procedures required, and make the claims process more convenient and efficient.

[0155] In this embodiment, the claims settlement module 306 is further used to:

[0156] Determining whether the personalized claims settlement plan meets the claims settlement conditions through the preset smart contract;

[0157] If the claim conditions are met, the medical data of the target patient is encrypted to obtain an encrypted certificate, the encrypted certificate is sent to the claim terminal for claim review, and the claim review result returned by the claim terminal is received;

[0158] When the claim review result is approved, the claim amount is automatically calculated through the smart contract, and the claim process is executed according to the claim amount.

[0159] Through on-chain management of patient medical data and insurance claims, mutual trust and collaboration among patients, hospitals, and claims are achieved, which improves claims efficiency and protects patient privacy and security.

[0160] In some optional implementations, the claims processing device 300 further includes a solution optimization module, which is used to:

[0161] Using a preset similarity calculation algorithm, calculate the similarity score between the personalized claim settlement plan and each historical claim settlement case in the historical claim settlement case library;

[0162] According to the similarity scores, a preset number of historical claims cases with the highest similarity are selected as reference cases;

[0163] The claim result data of the reference case is obtained, and the personalized claim settlement plan is optimized and adjusted according to the claim result data to obtain an optimized personalized claim settlement plan.

[0164] By optimizing personalized claims solutions, we can help better meet customer needs, further improve claims efficiency and accuracy, enhance market competitiveness, and achieve sustainable development.

[0165] In some optional implementations, the claim settlement device 300 further includes an insurance module, including:

[0166] A matching submodule is used to obtain the insurance product characteristics of all insurance products, and use a multi-dimensional matching algorithm to match the multi-dimensional disease characteristics with the insurance product characteristics to obtain the matching degree between the target patient and each of the insurance products;

[0167] A screening submodule, used to screen out the insurance products whose matching degree is greater than or equal to a preset threshold to form a candidate insurance product set;

[0168] An insurance plan generation submodule, used to generate a corresponding personalized insurance plan for each candidate insurance product in the candidate insurance product set through smart contract technology;

[0169] A push submodule, used for recommending the personalized insurance plan to the target patient;

[0170] The insurance submodule is used to respond to the target patient's selection instruction for the personalized insurance plan and use the candidate insurance product corresponding to the personalized insurance plan as the insurance product insured by the target patient.

[0171] By matching the patient's multi-dimensional disease characteristics with the characteristics of the insurance product through a multi-dimensional matching algorithm, it is possible to achieve an accurate correspondence between the patient's characteristics and the insurance product, thereby ensuring the accuracy of the claims.

[0172] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0173] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected and communicated through a system bus. It should be noted that the figure only shows a computer device 4 with a memory 41, a processor 42, and a network interface 43, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0174] The computer device may be a computing device such as a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with a user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0175] The memory 41 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as a hard disk or memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. equipped on the computer device 4. Of course, the memory 41 can also include both the internal storage unit of the computer device 4 and its external storage device. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions of the claim settlement method, etc. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or are to be output.

[0176] The processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run computer-readable instructions stored in the memory 41 or process data, such as computer-readable instructions for running the claims settlement method.

[0177] The network interface 43 may include a wireless network interface or a wired network interface. The network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0178] By using a trained semantic analysis model to parse electronic medical records and obtain disease information, the efficiency and accuracy of information extraction can be improved; normalizing disease information and constructing initial multi-dimensional disease features can ensure the accuracy and consistency of information and facilitate information management and application; generating medication recommendations and side effect monitoring plans based on online consultation records and electronic prescription information, and integrating them into the initial multi-dimensional disease features to obtain multi-dimensional disease features, can improve the accuracy of disease diagnosis, improve the accuracy of claims, and improve claims efficiency; by parsing insurance products to obtain corresponding claims rules, generating personalized claims plans based on claims rules and multi-dimensional disease features, and automatically executing the claims process based on personalized claims plans, it can accurately judge the claims results, provide more personalized claims plans, reduce the links of manual review and processing, enhance the fairness of claims, increase the claims rate, and improve user experience.

[0179] The present application also provides another embodiment, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the claims settlement method as described above.

[0180] By using a trained semantic analysis model to parse electronic medical records and obtain disease information, the efficiency and accuracy of information extraction can be improved; normalizing disease information and constructing initial multi-dimensional disease features can ensure the accuracy and consistency of information and facilitate information management and application; generating medication recommendations and side effect monitoring plans based on online consultation records and electronic prescription information, and integrating them into the initial multi-dimensional disease features to obtain multi-dimensional disease features, can improve the accuracy of disease diagnosis, improve the accuracy of claims, and improve claims efficiency; by parsing insurance products to obtain corresponding claims rules, generating personalized claims plans based on claims rules and multi-dimensional disease features, and automatically executing the claims process based on personalized claims plans, it can accurately judge the claims results, provide more personalized claims plans, reduce the links of manual review and processing, enhance the fairness of claims, increase the claims rate, and improve user experience.

[0181] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0182] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. A claims settlement method, characterized in that: The steps include: Obtain the target patient's electronic medical record, parse the electronic medical record using the trained semantic analysis model, and extract disease information; The disease information is converted into a standardized disease description through a preset medical term mapping rule library, and the initial multi-dimensional disease characteristics of the patient are constructed according to the disease description; Obtaining the online consultation record and electronic prescription information of the target patient, and generating medication recommendations and side effect monitoring plans based on the online consultation record and the electronic prescription information; Integrating the medication recommendation and the side effect monitoring program into the initial multidimensional disease signature to obtain a multidimensional disease signature; Obtaining the insurance product corresponding to the target patient, parsing the insurance product to obtain the corresponding claim rules, and generating a personalized claim plan based on the multi-dimensional disease characteristics and the claim rules; Through the preset smart contract, the claims process is automatically executed according to the personalized claims plan.

2. The claim settlement method according to claim 1, characterized in that: The step of using the trained semantic analysis model to parse the electronic medical record and extract disease information includes: Inputting the electronic medical record into the semantic analysis model, the semantic analysis model comprising an embedding layer, an entity extraction network, an encoder layer, and an output layer; Performing vector encoding on the text data in the electronic medical record through the embedding layer to obtain a text embedding vector, and inputting the text embedding vector into the entity extraction network; Perform entity extraction on the text embedding vector through the entity extraction network to obtain entity relationship triples, and input the entity relationship triples and the text data into the encoder layer; Extracting semantic and contextual features of the entity relationship triples and the text data through the attention mechanism of the encoder layer to obtain a semantic feature vector; The semantic feature vector is calculated through the output layer to output disease information.

3. The claim settlement method according to claim 1, characterized in that: The step of generating medication recommendations and side effect monitoring plans based on the online consultation records and the electronic prescription information includes: Extracting medication data, diagnosis information and treatment plans from the online consultation records and the electronic prescription information; By using knowledge graph technology, the medication data, the diagnosis information and the treatment plan are semantically associated to construct a personalized knowledge graph for the patient; Identify node associations on the personalized knowledge graph using a graph algorithm to obtain health characteristics of the patient; Inputting the patient's health characteristics into a trained drug recommendation model to perform drug screening and output medication recommendations; According to the drug information in the medication recommendation, an association rule mining algorithm is used to perform association analysis between the drug information and the diagnosis information in the personalized knowledge graph to obtain association rules between drugs and diagnoses; According to the association rules between the drug and the diagnosis, a preset adverse drug reaction knowledge base is queried to obtain the adverse drug reaction, and the medication recommendation is updated according to the adverse drug reaction and a corresponding side effect monitoring plan is generated.

4. The claim settlement method according to claim 1, characterized in that: The step of analyzing the insurance product to obtain corresponding claims rules and generating a personalized claims plan according to the multi-dimensional disease characteristics and the claims rules includes: Obtaining the policy text of the insurance product, performing feature extraction on the policy text, and obtaining key policy information; Identify candidate claim rules in the key information of the insurance policy, perform conflict detection on the candidate claim rules, and obtain a consistent claim rule set; Combining and optimizing the consistent claim rule set according to the multi-dimensional disease characteristics to generate an optimal claim rule combination; Generate a personalized claims settlement plan based on the optimal claim settlement rule combination.

5. The claim settlement method according to claim 1, characterized in that: The steps of automatically executing the claims process according to the personalized claims settlement plan through the preset smart contract include: Determining whether the personalized claims settlement plan meets the claims settlement conditions through the preset smart contract; If the claim conditions are met, the medical data of the target patient is encrypted to obtain an encrypted certificate, the encrypted certificate is sent to the claim terminal for claim review, and the claim review result returned by the claim terminal is received; When the claim review result is approved, the claim amount is automatically calculated through the smart contract, and the claim process is executed according to the claim amount.

6. The claim settlement method according to claim 1, characterized in that: After the step of generating a personalized claim settlement plan according to the multi-dimensional disease characteristics and the claim settlement rules, the method further includes: Using a preset similarity calculation algorithm, calculate the similarity score between the personalized claim settlement plan and each historical claim settlement case in the historical claim settlement case library; According to the similarity scores, a preset number of historical claims cases with the highest similarity are selected as reference cases; The claim result data of the reference case is obtained, and the personalized claim settlement plan is optimized and adjusted according to the claim result data to obtain an optimized personalized claim settlement plan.

7. The claim settlement method according to claim 1, characterized in that: Before the step of obtaining the insurance product corresponding to the target patient, the method further includes: Obtaining the insurance product characteristics of all insurance products, using a multi-dimensional matching algorithm to match the multi-dimensional disease characteristics with the insurance product characteristics, and obtaining the matching degree between the target patient and each of the insurance products; Screening out the insurance products whose matching degree is greater than or equal to a preset threshold to form a candidate insurance product set; Generate a corresponding personalized insurance plan for each candidate insurance product in the candidate insurance product set through smart contract technology; recommending the personalized insurance plan to target patients; In response to the target patient's selection instruction for the personalized insurance plan, the candidate insurance product corresponding to the personalized insurance plan is used as the insurance product insured by the target patient.

8. A claims settlement device, characterized in that: include: An acquisition module is used to acquire the electronic medical records of target patients, parse the electronic medical records using a trained semantic analysis model, and extract disease information; A standardization module, used for converting the disease information into a standardized disease description through a preset medical term mapping rule library, and constructing an initial multi-dimensional disease feature of the patient according to the disease description; A medication plan generation module, used to obtain the online consultation record and electronic prescription information of the target patient, and generate medication recommendations and side effect monitoring plans based on the online consultation record and the electronic prescription information; An integration module, used for integrating the medication recommendation and the side effect monitoring scheme into the initial multi-dimensional disease feature to obtain a multi-dimensional disease feature; A claim settlement plan generation module, used to obtain the insurance product corresponding to the target patient, parse the insurance product to obtain the corresponding claim settlement rules, and generate a personalized claim settlement plan according to the multi-dimensional disease characteristics and the claim settlement rules; The claims settlement module is used to automatically execute the claims settlement process according to the personalized claims settlement plan through a preset smart contract.

9. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores computer-readable instructions, and the processor implements the steps of the claims settlement method as claimed in any one of claims 1 to 7 when executing the computer-readable instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the claims settlement method according to any one of claims 1 to 7.

Images