Intelligent management method and system for medical insurance settlement list based on knowledge graph

Abstract

The application provides a medical insurance settlement list intelligent management method and system based on a knowledge graph, which comprises the following steps: analyzing obtained medical insurance settlement list data to obtain business keywords, and associating and matching the business keywords in a pre-constructed medical insurance knowledge graph to obtain medical insurance settlement entities; verifying the medical insurance settlement entities based on the medical insurance knowledge graph to obtain entity verification results; if the entity verification results are reasonable, performing priority logic matching based on first entity attributes of the medical insurance settlement entities and second entity attributes of medical insurance reimbursement rules in the medical insurance knowledge graph to determine settlement reimbursement rules of the medical insurance settlement entities; determining a medical insurance settlement reimbursement ratio based on the settlement reimbursement rules, and determining a medical insurance settlement reimbursement amount based on the medical insurance settlement reimbursement ratio and an actual settlement amount of the medical insurance settlement entities. The application realizes quick response to changes in medical insurance settlement rules, and improves the efficiency and accuracy of medical insurance settlement.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer, and in particular to a medical insurance settlement list intelligent management method and system based on a knowledge graph. BACKGROUND

[0002] In the process of continuous promotion of medical informatization, medical insurance settlement list intelligent management has become the key to improving the efficiency of medical insurance settlement. However, the current medical insurance settlement list management method has exposed significant deficiencies in dealing with complex and variable medical insurance rules and settlement rules. Medical insurance rules are constantly updated with social development and medical needs, and settlement rules are increasingly complex and detailed. Traditional management methods rely on fixed rule bases and manual adjustment modes. Whenever the rules change, staff need to manually modify the rule base one by one, which not only consumes time and effort, but also easily causes rule conflicts due to human operation errors, affecting settlement accuracy. At the same time, the integration of newly introduced rule clauses with existing rules also faces many difficulties, and cannot quickly respond to rule changes, resulting in a significant reduction in medical insurance settlement efficiency, seriously restricting the high-quality development of medical insurance settlement work. SUMMARY

[0003] The present application provides a medical insurance settlement list intelligent management method and system based on a knowledge graph, aiming to quickly respond to changes in medical insurance settlement rules and improve the efficiency and accuracy of medical insurance settlement.

[0004] In a first aspect, the present application provides a medical insurance settlement list intelligent management method based on a knowledge graph, comprising:

[0005] The medical insurance settlement list data obtained is parsed to obtain business keywords, and the business keywords are associated and matched in a pre-constructed medical insurance knowledge graph to obtain medical insurance settlement entities;

[0006] The medical insurance settlement entities are verified based on the medical insurance knowledge graph to obtain entity verification results; verification includes treatment method, treatment dosage, reimbursement range, and treatment cost rationality verification;

[0007] If the entity verification result is reasonable, the first entity attribute of the medical insurance settlement entity and the second entity attribute of the medical insurance reimbursement rule in the medical insurance knowledge graph are matched in priority logic to determine the settlement reimbursement rule of the medical insurance settlement entity;

[0008] The medical insurance settlement reimbursement proportion is determined based on the settlement reimbursement rule, and the medical insurance settlement reimbursement amount is determined based on the medical insurance settlement reimbursement proportion and the actual settlement amount of the medical insurance settlement entity.

[0009] In a second aspect, the present application further provides a medical insurance settlement list intelligent management system based on a knowledge graph, which is applied to the medical insurance settlement list intelligent management method based on a knowledge graph as described in the first aspect; the medical insurance settlement list intelligent management system based on a knowledge graph comprises:

[0010] a settlement entity matching module, configured to parse the obtained medical insurance settlement list data to obtain business keywords, and to associate and match the business keywords in a pre-constructed medical insurance knowledge graph to obtain medical insurance settlement entities;

[0011] an entity verification module, configured to verify the medical insurance settlement entities based on the medical insurance knowledge graph to obtain entity verification results; the verification includes treatment method rationality verification, treatment dosage rationality verification and reimbursement range rationality verification;

[0012] a reimbursement rule matching module, configured to, if the entity verification results are reasonable, perform priority logic matching based on first entity attributes of the medical insurance settlement entities and second entity attributes of medical insurance reimbursement rules in the medical insurance knowledge graph to determine settlement reimbursement rules of the medical insurance settlement entities;

[0013] a medical insurance list settlement module, configured to determine a medical insurance settlement reimbursement ratio based on the settlement reimbursement rules, and to determine a medical insurance settlement reimbursement amount based on the medical insurance settlement reimbursement ratio and an actual settlement amount of the medical insurance settlement entities.

[0014] In a third aspect, the present application further provides an electronic device, comprising: a memory configured to store a computer software program; and a processor configured to read and execute the computer software program to realize the medical insurance settlement list intelligent management method based on a knowledge graph as described in any one of the above aspects.

[0015] In a fourth aspect, the present application further provides a non-transitory computer readable storage medium, wherein the storage medium stores a computer software program, and the computer software program is executed by a processor to realize the medical insurance settlement list intelligent management method based on a knowledge graph as described in any one of the above aspects.

[0016] In a fifth aspect, the present application further provides a computer program product, comprising a computer program, and the computer program is executed by a processor to realize the medical insurance settlement list intelligent management method based on a knowledge graph as described in any one of the above aspects.

[0017] The medical insurance settlement list intelligent management method based on the knowledge graph provided by the embodiment of the present application can accurately verify the rationality of the business in the medical insurance settlement list data by performing entity matching between the medical insurance knowledge graph and the medical insurance settlement list data, and performing rationality verification on treatment methods, treatment dosage, reimbursement range and treatment cost, and then performing priority logic determination through entity attributes and the medical insurance knowledge graph to ensure that complex and variable rules can be executed in order in actual settlement, avoid rule conflicts, and perform medical insurance reimbursement according to the medical insurance settlement reimbursement proportion determined according to the settlement reimbursement rule, so that no matter how the medical insurance settlement rule changes, the medical insurance settlement list can be reasonably verified, the reimbursement rule matched, the reimbursement proportion determined and the reimbursement amount settled through the medical insurance knowledge graph, the entire process does not require a large amount of manual intervention, can quickly respond to changes in the medical insurance settlement rule, and improves the efficiency and accuracy of the medical insurance settlement. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 is a flowchart of the medical insurance settlement list intelligent management method based on the knowledge graph provided by the embodiment of the present application;

[0019] Figure 2 is a structural diagram of the medical insurance settlement list intelligent management system based on the knowledge graph provided by the embodiment of the present application;

[0020] Figure 3 is an embodiment diagram of an electronic device provided by the embodiment of the present application;

[0021] Figure 4 is an embodiment diagram of a computer readable storage medium provided by the embodiment of the present application. DETAILED DESCRIPTION

[0022] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0023] In the description of the present application, the terms "first", "second" are used only for description purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "multiple" is two or more, unless otherwise specifically limited.

[0024] In the description of the present application, the term "for example" is used to mean "serving as an example, instance, or illustration." Any embodiment described as "for example" in the present application is not necessarily to be construed as preferred or advantageous over other embodiments. The following description is presented to enable any person skilled in the art to make and use the application. In the following description, for the purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the present application. It will be apparent, however, to one skilled in the art that the present application can be practiced without using these specific details. In other instances, well-known structures and processes are not elaborated upon in order to avoid unnecessary detail, which can obscure the description of the present application. Thus, the present application is not intended to be limited by the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed herein.

[0025] Optionally, referring to Figure 1 , Figure 1 is a flowchart of the medical insurance settlement list intelligent management method based on a knowledge graph provided by the present application. The execution subject of the medical insurance settlement list intelligent management method based on a knowledge graph in the present application is a medical insurance intelligent management system. Therefore, the medical insurance settlement list intelligent management method based on a knowledge graph comprises the following steps.

[0026] Step 10: The medical insurance settlement list data obtained is parsed to obtain business keywords, and the business keywords are associated and matched in the pre-constructed medical insurance knowledge graph to obtain medical insurance settlement entities.

[0027] Optionally, when medical insurance settlement reimbursement is needed, the medical insurance settlement list data is input into the medical insurance intelligent management system, or the medical insurance settlement list data is sent to the medical insurance intelligent management system through a terminal device. Therefore, the medical insurance intelligent management system parses the data after obtaining the medical insurance settlement list data. The present application extracts business keywords in the data through natural language processing (NLP) technology, wherein the business keywords usually include disease names, drug names, examination items, treatment methods, etc.

[0028] Further, the medical insurance intelligent management system associates and matches the extracted business keywords with the pre-constructed medical insurance knowledge graph, wherein the medical insurance knowledge graph is a structured knowledge base containing a large number of medical insurance related concepts, entities and their relationships. Therefore, the nodes matched with the business keywords are searched in the knowledge graph to determine the medical insurance settlement entities, i.e. specific medical insurance business objects related to the settlement list.

[0029] In an embodiment, the medical insurance settlement list is "patient hospitalized due to pneumonia, using ceftriaxone sodium for anti-infection treatment, chest CT examination, 5 days of hospitalization", and the business keywords such as "pneumonia", "ceftriaxone sodium", "chest CT examination", and "hospitalization treatment" are extracted from the list by using the NLP technology. Then, it is found in the medical insurance knowledge graph that "pneumonia" corresponds to a disease entity in the knowledge graph, "ceftriaxone sodium" corresponds to a drug entity, "chest CT examination" corresponds to an examination item entity, and "hospitalization treatment" corresponds to a treatment mode entity, and the medical insurance settlement entity of the settlement list is obtained.

[0030] In step 20, the medical insurance settlement entity is verified based on the medical insurance knowledge graph, and an entity verification result is obtained. The verification includes treatment mode verification, treatment amount verification, reimbursement range verification, and treatment cost reasonableness verification.

[0031] Further, the medical insurance settlement entity is verified by the medical insurance knowledge graph, wherein the verification includes treatment mode verification, checking whether the current treatment mode conforms to the conventional treatment means of the disease; treatment amount verification, judging whether the drug use dosage, examination item frequency, etc. is within a reasonable range; reimbursement range verification, confirming whether the related drugs, examination items, etc. are within the medical insurance reimbursement directory; and treatment cost reasonableness verification, evaluating whether the treatment cost is reasonable in combination with the local medical insurance policy and market price. Therefore, the medical insurance intelligent management system verifies each medical insurance settlement entity according to the medical insurance policy, medical knowledge, etc. stored in the knowledge graph, and obtains an entity verification result, which is specifically described in steps 201 to 205.

[0032] Continuing the above embodiment, for the settlement list of the above "pneumonia" treatment, in the treatment mode verification, the knowledge graph shows that the conventional treatment mode of pneumonia includes the use of antibiotics and necessary imaging examination, and the current "ceftriaxone sodium anti-infection treatment" and "chest CT examination" conform to the convention, and the verification is passed. In the treatment amount verification, the conventional dosage range of ceftriaxone sodium for pneumonia is queried in the knowledge graph, and it is found that the dosage in the list is within the reasonable range, and the verification is passed. In the reimbursement range verification, it is confirmed that ceftriaxone sodium and chest CT examination are within the medical insurance reimbursement directory, and the verification is passed. In the treatment cost reasonableness verification, the total cost of the settlement list is found to be reasonable by referring to the cost standard of pneumonia hospitalization treatment in the local medical insurance policy and the market price of ceftriaxone sodium and chest CT examination, and the verification is passed, and therefore, the entity verification result of the medical insurance settlement entity is finally obtained as meeting the requirements.

[0033] In step 30, if the entity verification result meets the reasonableness, the first entity attribute of the medical insurance settlement entity and the second entity attribute of the medical insurance reimbursement rule in the medical insurance knowledge graph are matched based on the priority logic to determine the settlement reimbursement rule of the medical insurance settlement entity.

[0034] Further, if the entity verification result is reasonable, the medical insurance intelligent management system performs priority logic matching based on the first entity attribute (such as disease type, treatment method, drug category, etc.) of the medical insurance settlement entity and the second entity attribute of the medical insurance reimbursement rule in the medical insurance knowledge graph.

[0035] In the embodiment of the present application, the medical insurance knowledge graph also stores a plurality of medical insurance reimbursement rules, each of which corresponds to different entity attribute conditions and priorities. Therefore, the medical insurance intelligent management system matches the attributes of the medical insurance settlement entity with the attributes of the reimbursement rules according to the preset priority order, finds the most suitable reimbursement rule, and determines the settlement reimbursement rule of the medical insurance settlement entity, specifically as steps 301 to 304.

[0036] Continuing with the medical insurance settlement entity of the above-mentioned pneumonia treatment, the first entity attribute includes disease as pneumonia, treatment method as hospitalization treatment, and drug as ceftriaxone sodium, etc. The medical insurance reimbursement rule in the medical insurance knowledge graph is: rule A is for pneumonia hospitalization treatment and use of antibiotics in the directory, with a reimbursement ratio of 80%; rule B is for general hospitalization treatment, with a reimbursement ratio of 70%. Since rule A is more specific and completely matches the attributes of the medical insurance settlement entity, and the priority of rule A is higher than that of rule B, the medical insurance intelligent management system determines the settlement reimbursement rule of the medical insurance settlement entity as rule A, i.e. the reimbursement ratio is 80%.

[0037] Step 40, determine the medical insurance settlement reimbursement ratio based on the settlement reimbursement rule, and determine the medical insurance settlement reimbursement amount based on the medical insurance settlement reimbursement ratio and the actual settlement amount of the medical insurance settlement entity.

[0038] Further, the medical insurance intelligent management system determines the medical insurance settlement reimbursement ratio according to the rule, and obtains the actual settlement amount of the medical insurance settlement entity. Further, the medical insurance intelligent management system calculates the medical insurance settlement reimbursement amount by multiplying the reimbursement ratio and the actual settlement amount. Continuing with the above embodiment, the medical insurance settlement entity of the above-mentioned pneumonia treatment determines the reimbursement ratio as 80%, and the actual settlement amount of the settlement list is 5000 yuan. The medical insurance intelligent management system calculates 5000 x 80% = 4000 yuan, thereby determining the medical insurance settlement reimbursement amount of the medical insurance settlement list as 4000 yuan.

[0039] The embodiment of the present application can accurately verify the rationality of the business in the medical insurance settlement list data by performing entity matching between the medical insurance knowledge graph and the medical insurance settlement list data, and performing rationality verification of treatment methods, rationality verification of treatment dosage, rationality verification of reimbursement range, and rationality verification of treatment cost. Then, the priority logic is determined through entity attributes and the medical insurance knowledge graph, so that the complex and variable rules can be executed in order in the actual settlement, and rule conflicts are avoided. Then, the medical insurance reimbursement is performed according to the medical insurance settlement reimbursement ratio determined according to the settlement reimbursement rule. Therefore, no matter how the medical insurance settlement rule changes, the medical insurance settlement list can be reasonably verified, the reimbursement rule is matched, the reimbursement ratio is determined, and the reimbursement amount is settled through the medical insurance knowledge graph. The entire process does not require a large amount of manual intervention, can quickly respond to changes in the medical insurance settlement rule, and improves the efficiency and accuracy of the medical insurance settlement.

[0040] In an embodiment, steps 201 to 205 are described as follows:

[0041] In step 201, if any first target entity and second target entity in the medical insurance settlement entity do not have a direct relationship in the medical insurance knowledge graph, the first target entity and the second target entity are taken as end points, and path traversal is performed in the medical insurance knowledge graph with a preset step length to obtain initial indirect paths. The medical insurance business logic of each initial indirect path is determined to determine the target indirect path.

[0042] Optionally, for any first target entity and second target entity in the medical insurance settlement entity, if it is determined that the first target entity and the second target entity have a direct relationship in the medical insurance knowledge graph, the medical insurance intelligent management system determines that the medical insurance settlement entity meets the rationality of the treatment method.

[0043] Further, if it is determined that the first target entity and the second target entity do not have a direct relationship in the medical insurance knowledge graph, the medical insurance intelligent management system takes the two entities as end points and performs path traversal in the medical insurance knowledge graph according to a preset step length. The preset step length can be set according to the complexity of the knowledge graph, for example, set to 2-5 steps, that is, at most 2-5 intermediate nodes are used to connect the two end points. During the traversal process, the medical insurance intelligent management system obtains all possible initial indirect paths.

[0044] Further, the medical insurance intelligent management system analyzes the medical insurance business logic of each initial indirect path to determine whether it meets the common association mode in the medical insurance business, such as disease-symptom-treatment method, drug-indication-disease, etc. The target indirect path meeting the requirement is selected.

[0045] In the medical insurance settlement entity for continuing hospitalization of pneumonia, for example, the first target entity is "chest CT examination", and the second target entity is "ceftriaxone sodium", which has no direct relationship in the medical insurance knowledge graph. The preset step length is set to 3, and "chest CT examination" and "ceftriaxone sodium" are taken as endpoints to traverse in the knowledge graph, and three initial indirect paths are obtained: path 1 is "chest CT examination-pneumonia diagnosis-antibiotic treatment-ceftriaxone sodium"; path 2 is "chest CT examination-lung image features-bacterial infection-ceftriaxone sodium"; and path 3 is "chest CT examination-doctor diagnosis suggestion-drug regimen-ceftriaxone sodium". Then, the medical insurance intelligent management system analyzes the medical insurance business logic of each path, finds that path 1 and path 2 conform to the conventional association mode of disease diagnosis and treatment and drug use, and the "doctor diagnosis suggestion-drug regimen" association of path 3 is not clear and specific enough, and finally determines that path 1 and path 2 are target indirect paths

[0046] Step 202, the public association entity in which the frequency of each entity in the target indirect path is greater than the preset frequency threshold.

[0047] Further, a preset frequency threshold is set, for example, 30%, and the medical insurance intelligent management system counts the frequency of each entity appearing in the target indirect path determined in step 201. If the frequency of an entity appearing in all target indirect paths is greater than the threshold, it is identified as a public association entity.

[0048] Continuing with the above determined target indirect path 1 "chest CT examination-pneumonia diagnosis-antibiotic treatment-ceftriaxone sodium" and path 2 "chest CT examination-lung image features-bacterial infection-ceftriaxone sodium", the system counts the frequency of each entity. The frequency of "chest CT examination" and "ceftriaxone sodium" as endpoint entities is 100%, the frequency of "pneumonia diagnosis" is 50%, the frequency of "antibiotic treatment" is 50%, the frequency of "lung image features" is 50%, and the frequency of "bacterial infection" is 50%. The preset frequency threshold is set to 30%, and "chest CT examination", "ceftriaxone sodium", "pneumonia diagnosis", "antibiotic treatment", "lung image features", and "bacterial infection" are all determined as public association entities.

[0049] Step 203, based on the matching relationship between the entity type of the public association entity and the entity scene of the first target entity and the second target entity, respectively, determine the first scene matching strength and the second scene matching strength, and based on the entity distance between the public association entity and the first target entity and the second target entity, respectively, determine the first entity association strength and the second entity association strength.

[0050] Further, the medical insurance intelligent management system matches and analyzes the entity type (such as disease, medicine, examination item, etc.) of the public associated entity with the entity scene (such as diagnosis scene, treatment scene, etc.) of the first target entity and the second target entity respectively. Optionally, the present embodiment adopts the Jaccard similarity coefficient formula to calculate the scene matching strength, and the formula is: J (A, B) = |A∩B| / |A∪B|, wherein A and B respectively represent the entity type set and the entity scene set. Further, the medical insurance intelligent management system determines the entity association strength according to the shortest path length of the public associated entity in the knowledge graph and the first target entity and the second target entity, and adopts the formula S = 1 / d, wherein S is the entity association strength, and d is the shortest path length. Therefore, through the two formulas, the medical insurance intelligent management system respectively calculates the first scene matching strength, the second scene matching strength, the first entity association strength and the second entity association strength.

[0051] Continue to the public associated entity "pneumonia diagnosis", which belongs to the diagnosis type entity, the entity scene corresponding to the first target entity "chest CT examination" is the diagnosis auxiliary examination scene, and the entity scene corresponding to the second target entity "ceftriaxone sodium" is the treatment drug scene. For example, the entity type set of "pneumonia diagnosis" is A1, the entity scene set of "chest CT examination" is B1, and the entity scene set of "ceftriaxone sodium" is B2, and the first scene matching strength and the second scene matching strength are obtained by calculating J (A1, B1) and J (A1, B2). In the knowledge graph, the shortest path length from "pneumonia diagnosis" to "chest CT examination" is 1, and the shortest path length to "ceftriaxone sodium" is 2, so the first entity association strength is calculated as 1, and the second entity association strength is calculated as 0.5. The strength values of each entity are obtained by calculating all the public associated entities.

[0052] In step 204, if the first scene matching strength and the second scene matching strength are both greater than the preset matching strength threshold, and the first entity association strength and the second entity association strength are both less than or equal to the preset association strength threshold, it is determined that the medical insurance settlement entity meets the treatment mode rationality. Otherwise, it does not meet.

[0053] Further, the medical insurance intelligent management system compares the first scene matching strength and the second scene matching strength with the preset matching strength threshold (such as 0.6), and simultaneously compares the first entity association strength and the second entity association strength with the preset association strength threshold (such as 0.8). Only when the first scene matching strength and the second scene matching strength are both greater than the preset matching strength threshold, and the first entity association strength and the second entity association strength are both less than or equal to the preset association strength threshold, it is determined that the medical insurance settlement entity meets the treatment mode rationality; otherwise, it is determined that it does not meet the treatment mode rationality.

[0054] In an embodiment, it is calculated that the minimum values of the first scene matching strength and the second scene matching strength corresponding to all public associated entities are 0.7 and 0.65 respectively, both of which are greater than the preset matching strength threshold 0.6; the maximum values of the first entity association strength and the second entity association strength are 0.7 and 0.75 respectively, both of which are less than or equal to the preset association strength threshold 0.8. At this time, it is determined that the medical insurance settlement entity meets the treatment mode rationality. If any one of the strength values does not meet the condition, it is determined that the treatment mode rationality is not met.

[0055] In step 205, if the medical insurance settlement entity meets the treatment mode rationality, the medical insurance settlement entity is verified for treatment dosage rationality based on the medical insurance knowledge graph, and an entity verification result is obtained.

[0056] Further, if the medical insurance settlement entity meets the treatment mode rationality, the medical insurance intelligent management system verifies the medical insurance settlement entity for treatment dosage rationality according to the medical insurance knowledge graph, and obtains an entity verification result, which is specifically described in steps 2051 to 2055.

[0057] The embodiment of the present application can analyze the relationship between each entity in the medical insurance settlement entity, not only determine whether the treatment mode is reasonable, but also verify the rationality of the treatment dosage on the basis of the reasonable treatment mode, so as to accurately verify the rationality of the business in the medical insurance settlement list data, thereby quickly responding to the change of medical insurance settlement rules, improving the efficiency and accuracy of medical insurance settlement, and ensuring the healthy and stable operation of the medical insurance system.

[0058] In an embodiment, steps 2051 to 2055 are described as follows:

[0059] In step 2051, a local association network is constructed based on the nodes and node edges related to the first target entity and the second target entity in the medical insurance knowledge graph, and actual treatment dosage data is obtained by screening the nodes and node edges related to the treatment dosage in the local association network.

[0060] Optionally, the medical insurance intelligent management system extracts the nodes and node edges related to the first target entity and the second target entity according to the medical insurance knowledge graph, constructs a local association network with the first target entity and the second target entity as the core, wherein the local association network covers various types of information directly or indirectly associated with the two entities in the medical insurance business logic. In the constructed local association network, the medical insurance intelligent management system screens the nodes and node edges related to the treatment dosage to obtain the actual treatment dosage data. For example, for a drug entity, its usage dosage, usage frequency and other data are obtained; for an examination item entity, the number of examinations and other data are obtained.

[0061] Continuing in the medical insurance settlement entity of pneumonia inpatient treatment, the first target entity is "ceftriaxone sodium", and the second target entity is "chest CT examination". The medical insurance intelligent management system extracts related nodes and node edges in the medical insurance knowledge graph with the two entities as the core, and constructs a local correlation network containing nodes such as "pneumonia", "antibiotic treatment", "pulmonary imaging diagnosis" and their mutual relationships. In the local correlation network, the nodes and edges related to the dosage and frequency of use of "ceftriaxone sodium", and the nodes and edges related to the frequency of "chest CT examination" are screened out, and the actual treatment dosage data is obtained as 1.5 grams of ceftriaxone sodium per day, 1 time per day, and 1 time of chest CT examination.

[0062] Step 2052, determining the standard treatment dosage range in the treatment scene where the first target entity and the second target entity are located based on the medical insurance knowledge graph, and determining the initial treatment dosage difference data based on the actual treatment dosage data and the standard treatment dosage range.

[0063] Further, the medical insurance intelligent management system determines the standard treatment dosage range in the treatment scene where the first target entity and the second target entity are located according to the medical insurance knowledge graph, wherein the standard treatment dosage range is pre-set in the knowledge graph based on medical norms, medical insurance policies and other factors.

[0064] Further, the medical insurance intelligent management system compares the actual treatment dosage data with the standard treatment dosage range, and determines the initial treatment dosage difference data by calculating the difference between the two.

[0065] Continuing for the above-mentioned pneumonia treatment scene, the medical insurance intelligent management system obtains from the medical insurance knowledge graph that the standard daily dosage range of ceftriaxone sodium in the pneumonia treatment scene is 1-2 grams, the use frequency is 1-2 times per day, and the reasonable examination frequency of chest CT examination is 1-2 times. Comparing the actual treatment dosage data (ceftriaxone sodium 1.5 grams per day, 1 time per day, and 1 time of chest CT examination) with the standard treatment dosage range, it is calculated that the dosage of ceftriaxone sodium is within the standard range, the difference is 0; the frequency of chest CT examination is also within the standard range, the difference is also 0, and the initial treatment dosage difference data is obtained.

[0066] Step 2053, based on the initial treatment dosage difference data, traversing the influence factors in the medical insurance knowledge graph to obtain dosage difference associated factors, and modifying the initial treatment dosage difference data based on the dosage difference associated factors to obtain target treatment dosage difference data. The dosage difference associated factors include patient characteristic factors and treatment environment factors.

[0067] Further, the medical insurance intelligent management system traverses the influencing factors in the medical insurance knowledge graph based on the initial treatment dosage difference data, wherein, the nodes and edges related to patient characteristic factors (such as age, weight, underlying diseases, etc.) and treatment environment factors (such as hospital level, treatment region, etc.) are focused on, and the dosage difference correlation factors are obtained. Further, the medical insurance intelligent management system corrects the initial treatment dosage difference data according to the correlation factors to obtain target treatment dosage difference data. Optionally, the present embodiment adopts a correction method based on conditional probability, for patient characteristic factors, for example, the patient characteristic set is P, and the dosage correction coefficient under different characteristic combinations is C P , for treatment environment factors, for example, the treatment environment set is E, and the dosage correction coefficient under different environment combinations is C E , the correction formula is: D 修正 =D 初始 *C P *C E , wherein D 修正 is the target treatment dosage difference data, and D 初始 is the initial treatment dosage difference data.

[0068] Continuing in the above-mentioned pneumonia treatment embodiment, the medical insurance intelligent management system finds through traversal in the medical insurance knowledge graph that the patient is 65 years old (belongs to an elderly patient, corresponding to a specific correction coefficient), and the treatment hospital is a third-grade A-level hospital (corresponding to a different treatment environment correction coefficient). According to the settings in the knowledge graph, the correction coefficient C P of the elderly patient for the dosage of ceftriaxone sodium is 1.1, and the correction coefficient C E of the third-grade A-level hospital for the number of chest CT examinations is 0.9. For ceftriaxone sodium, the initial difference is 0, so the target treatment dosage difference data after correction is still 0; for chest CT examination, the initial difference is 0, so the target treatment dosage difference data after correction is also 0.

[0069] Step 2054, if the target treatment dosage difference data is within the dosage difference threshold range, it is determined that the medical insurance settlement entity meets the treatment dosage rationality. Otherwise, it does not meet.

[0070] Further, a dosage difference threshold range is set, and the medical insurance intelligent management system compares the target treatment dosage difference data with the threshold range. If the target treatment dosage difference data is within the dosage difference threshold range, the medical insurance intelligent management system determines that the medical insurance settlement entity meets the treatment dosage rationality; if it exceeds the dosage difference threshold range, the medical insurance intelligent management system determines that it does not meet the treatment dosage rationality.

[0071] Continuing the above pneumonia treatment embodiment, the dosage difference threshold range is [-0.2, 0.2], and since the target treatment dosage difference data of ceftriaxone sodium and chest CT examination are both 0, which are within the set dosage difference threshold range, the medical insurance intelligent management system determines that the medical insurance settlement entity conforms to the treatment dosage rationality.

[0072] In step 2055, if the medical insurance settlement entity conforms to the treatment dosage rationality, the medical insurance settlement entity is verified for the reimbursement range rationality based on the medical insurance knowledge graph, and an entity verification result is obtained.

[0073] Further, if the medical insurance settlement entity conforms to the treatment dosage rationality, the medical insurance intelligent management system verifies the medical insurance settlement entity for the reimbursement range rationality according to the medical insurance knowledge graph, and obtains an entity verification result, which is specifically described in steps 20551 to 20554.

[0074] The embodiment of the present application can comprehensively and scientifically verify the treatment dosage rationality of the medical insurance settlement entity, and further verify the reimbursement range rationality, so as to accurately verify the rationality of the business in the medical insurance settlement list data, thereby quickly responding to the changes of medical insurance settlement rules, improving the efficiency and accuracy of medical insurance settlement, and ensuring the healthy and stable operation of the medical insurance system.

[0075] In an embodiment, steps 20551 to 20554 are described as follows:

[0076] In step 20551, the reimbursement range rules of the first target entity and the second target entity are extracted according to the local correlation network. The reimbursement range rules include reimbursable items, reimbursement processes and reimbursement additional conditions.

[0077] Optionally, the medical insurance intelligent management system retrieves the reimbursement range rules related to the first target entity and the second target entity in the medical insurance knowledge graph according to the local correlation network, wherein the reimbursement range rules include three types of reimbursable items, reimbursement processes and reimbursement additional conditions. The reimbursable items specify which drugs, inspection items, etc. can be included in the medical insurance reimbursement; the reimbursement process specifies the links and involved entities that the reimbursement needs to go through; and the reimbursement additional condition sets additional limitation conditions for the reimbursement, such as specific diseases, treatment duration, etc.

[0078] Continuing in the medical insurance settlement entity of pneumonia inpatient treatment, the first target entity is "ceftriaxone sodium", and the second target entity is "chest CT examination", and the reimbursement range rules related to the two entities are extracted in the medical insurance knowledge graph according to the local correlation network. The reimbursable items show that ceftriaxone sodium is a reimbursable drug for treating pneumonia, and chest CT examination is a reimbursable examination item under the demand of pneumonia diagnosis; the reimbursement process includes nodes such as patient application submission, hospital review, and medical insurance department review, and the corresponding process; the reimbursement additional conditions are that pneumonia needs to meet inpatient treatment and meet certain diagnosis standards.

[0079] Step 20552, determine whether the first target entity and the second target entity belong to the entity nodes in the reimbursable items based on the medical insurance knowledge graph, and determine whether the first target entity and the second target entity belong to the entity nodes in the reimbursement process based on the medical insurance knowledge graph, and determine whether the first target entity and the second target entity belong to the entity nodes in the reimbursement additional conditions based on the medical insurance knowledge graph.

[0080] Further, the medical insurance intelligent management system determines whether the first target entity and the second target entity belong to the entity nodes in the reimbursable items, the reimbursement process and the reimbursement additional conditions according to the medical insurance knowledge graph. The embodiments of the present application determine whether the entity meets the requirements of the corresponding rules by searching for the association relationship between the entity and each rule node in the knowledge graph. For the reimbursable items, check whether the entity is in the prescribed reimbursable list; for the reimbursement process, confirm whether the entity participates in the prescribed process link; for the reimbursement additional conditions, verify whether the entity meets the set conditions.

[0081] Continuing the above embodiment, for "ceftriaxone sodium" and "chest CT examination", in the medical insurance knowledge graph, it is confirmed that "ceftriaxone sodium" is in the reimbursable drug list for treating pneumonia, and "chest CT examination" is in the reimbursable examination item list for pneumonia diagnosis; in the reimbursement process, it is found that the treatment and examination behaviors corresponding to the two entities involve process nodes such as patient application submission and hospital review; in the reimbursement additional conditions, since the patient is hospitalized due to pneumonia and the diagnosis meets the standard, "ceftriaxone sodium" and "chest CT examination" meet the requirements of the reimbursement additional conditions.

[0082] Step 20553, if the first target entity and the second target entity belong to the nodes in the reimbursable items, the reimbursement process or / and the reimbursement additional conditions, it is determined that the medical insurance settlement entity meets the reimbursement range rationality. Or, if the first target entity or the second target entity belongs to the nodes in the reimbursable items, the reimbursement process or / and the reimbursement additional conditions, it is determined that the first target entity or the second target entity meets the reimbursement range rationality.

[0083] Further, the medical insurance intelligent management system comprehensively determines according to the judgment result of the reimbursement range rule. If the first target entity and the second target entity both belong to the nodes in the reimbursable item, the reimbursement process or / and the reimbursement additional condition, or at least one of the first target entity or the second target entity belongs to the nodes, the medical insurance intelligent management system determines that the medical insurance settlement entity or the corresponding entity meets the reimbursement range rationality.

[0084] Further, if it is determined that neither the first target entity nor the second target entity meets the condition, the medical insurance intelligent management system determines that the medical insurance settlement entity does not meet the reimbursement range rationality.

[0085] Continuing the above embodiment, since both "ceftriaxone sodium" and "chest CT examination" meet the requirements of the reimbursable item, the reimbursement process and the reimbursement additional condition, the medical insurance intelligent management system determines that the medical insurance settlement entity meets the reimbursement range rationality.

[0086] In step 20554, if the medical insurance settlement entity meets the treatment method rationality, or the first target entity or the second target entity meets the reimbursement range rationality, the medical insurance intelligent management system verifies the treatment cost rationality of the medical insurance settlement entity based on the medical insurance knowledge graph to obtain an entity verification result.

[0087] Further, if the medical insurance settlement entity meets the treatment method rationality, or the first target entity or the second target entity meets the reimbursement range rationality, the medical insurance intelligent management system verifies the treatment cost rationality of the medical insurance settlement entity based on the medical insurance knowledge graph to obtain an entity verification result, which is specifically described in steps 205541 to 205545.

[0088] The embodiment of the present application can accurately determine whether the medical insurance settlement is compliant and reasonable from the complete verification chain of reimbursement rule review to cost verification, so as to accurately verify the rationality of the business in the medical insurance settlement list data, thereby quickly responding to changes in medical insurance settlement rules, improving the efficiency and accuracy of medical insurance settlement, and ensuring the healthy and stable operation of the medical insurance system.

[0089] In an embodiment, steps 205541 to 205545 are described as follows:

[0090] In step 205541, the medical insurance intelligent management system determines the entity path between the first target entity and the second target entity based on the medical insurance knowledge graph, and determines the path node related to the treatment cost in the entity path.

[0091] Optionally, the medical insurance intelligent management system finds all entity paths between the first target entity and the second target entity according to the medical insurance knowledge graph, wherein the entity path reflects the association relationship between the two entities in the medical insurance business logic, for example, connected through intermediate nodes such as diseases, treatment methods, drug use, etc.

[0092] Further, for each entity path, the medical insurance intelligent management system screens out the path nodes related to treatment cost, wherein the path nodes usually include drug price, examination item charging standard, treatment service cost and other information directly related to cost calculation.

[0093] Continuing in the medical insurance settlement entity of pneumonia hospitalization treatment, the first target entity is "ceftriaxone sodium" and the second target entity is "chest CT examination". The medical insurance intelligent management system searches in the medical insurance knowledge graph and obtains two entity paths: path 1 is "ceftriaxone sodium-pneumonia treatment-chest CT examination" and path 2 is "ceftriaxone sodium-antibiotic drug category-medical examination item category-chest CT examination". In the two paths, the path nodes related to treatment cost are screened out, such as the drug unit price node corresponding to "ceftriaxone sodium", the examination item charging standard node corresponding to "chest CT examination", and the related cost nodes such as bed fee and nursing fee involved in "pneumonia treatment".

[0094] Step 205542, based on the cost standard interval of the path nodes in each entity path in the medical insurance knowledge graph and the influence coefficient of the association relationship between the path nodes on the cost, the expected cost value of each entity path is determined, and the expected cost range of the medical insurance settlement entity is determined based on the expected cost value of each entity path.

[0095] Further, for each entity path, the medical insurance intelligent management system obtains the cost standard interval corresponding to the path nodes from the medical insurance knowledge graph, and determines the influence coefficient of the association relationship between the path nodes on the cost, wherein the influence coefficient can be set according to different business logic relationships recorded in the knowledge graph, for example, the influence degree of different treatment methods on drug use cost. Optionally, in the embodiment of the present application The expected cost value of each entity path is calculated, wherein E p is the expected cost value, C i is the cost standard interval value of the i-th path node, I i is the influence coefficient of the association relationship between the i-th path node and the next path node on the cost, and n is the number of path nodes.

[0096] Further, after obtaining the expected cost value of each entity path, the medical insurance intelligent management system takes the minimum value and the maximum value in all path expected cost values to determine the expected cost range of the medical insurance settlement entity.

[0097] Continuing the above example, for path 1 "ceftriaxone sodium-pneumonia treatment-chest CT examination", for example, the drug single price standard interval of "ceftriaxone sodium" is [100, 150] yuan, the median C1=125 yuan, and the influence coefficient I1=1.2 of the associated relationship with "pneumonia treatment" on cost; the cost standard interval of "pneumonia treatment" is [2000, 3000] yuan, the median C2=2500 yuan, and the influence coefficient I2=1.1 of the associated relationship with "chest CT examination" on cost; the charge standard interval of "chest CT examination" is [500, 800] yuan, and the median C3=360 yuan. According to the formula, the expected cost E of the path is calculated as follows: p1 =125×1.2×2500×1.1×650=25593750 yuan. For path 2 "ceftriaxone sodium-antibiotic drug category-medical examination item category-chest CT examination", the expected cost E is calculated as follows: p2 =23000000 yuan. Taking the minimum and maximum values in E p1 and E p2 , the expected cost range is determined as [23000000, 25593750] yuan.

[0098] Step 205543, if the actual treatment cost in the medical insurance settlement entity is within the expected cost range, it is determined that the entity verification result is in line with the rationality.

[0099] Further, the medical insurance intelligent management system compares the actual treatment cost in the medical insurance settlement entity with the expected cost range. If the actual treatment cost is within the expected cost range, i.e., the medical insurance settlement entity is in line with the treatment cost rationality, it can be understood that the medical insurance settlement entity is in line with the treatment method rationality, the treatment dosage rationality, the reimbursement range rationality, and the treatment cost rationality at this time. Therefore, the medical insurance intelligent management system directly determines that the entity verification result is in line with the rationality. Continuing the above example, the actual treatment cost of the medical insurance settlement entity for the inpatient treatment of pneumonia is 24000000 yuan, which is within the expected cost range of [23000000, 25593750] yuan, and the medical insurance intelligent management system determines that the entity verification result is in line with the rationality.

[0100] Step 205544, if the actual treatment cost is not within the expected cost range, the cost deviation path set is obtained. The expected cost value of the cost deviation path in the cost deviation path set has an intersection with an interval centered on the actual treatment cost and having a preset error cost as a radius.

[0101] Further, if the actual treatment cost is not within the cost expectation range, the medical insurance intelligent management system obtains a cost deviation path set, wherein a cost deviation path in the cost deviation path set is a path whose cost expectation value intersects with an interval centered at the actual treatment cost and having a preset error cost as a radius. Therefore, the medical insurance intelligent management system filters out paths meeting the condition by traversing the cost expectation values of all the entity paths, and forms the cost deviation path set. Continuing with the above embodiment, if the actual treatment cost is 26,000,000 yuan, which exceeds the cost expectation range [23,000,000, 25,593,750] yuan, and the preset error cost is 1,000,000 yuan, the interval centered at 26,000,000 yuan and having a radius of 1,000,000 yuan is [25,000,000, 27,000,000] yuan. By traversing all the entity paths, it is found that there is a path whose cost expectation value is 25,800,000 yuan, which is within the interval, and this path is included in the cost deviation path set.

[0102] In step 205545, if the number of cost deviation paths in the cost deviation path set is less than a preset number threshold, it is determined that the entity verification result meets the rationality.

[0103] Further, the medical insurance intelligent management system compares the number of cost deviation paths in the cost deviation path set with the preset number threshold. If the number of cost deviation paths is less than the preset number threshold, the medical insurance intelligent management system determines that the entity verification result meets the rationality; otherwise, it is determined that the entity verification result does not meet the rationality, wherein the preset number threshold in the embodiment of the present application is 1.

[0104] Therefore, it can be understood that as long as the medical insurance settlement entity does not meet at least one of the treatment mode rationality, the treatment dosage rationality, the reimbursement range rationality, and the treatment cost rationality, the medical insurance intelligent management system determines that the entity verification result of the medical insurance settlement entity does not meet the rationality.

[0105] The embodiment of the present application analyzes the treatment cost rationality by means of the entity path, the cost-related node, and the various possible cost-related paths, and therefore can scientifically and comprehensively verify the treatment cost rationality of the medical insurance settlement entity, reduce the misjudgment caused by special circumstances or complex business logic, accurately verify the rationality of the business in the medical insurance settlement list data, quickly respond to changes in medical insurance settlement rules, improve the efficiency and accuracy of medical insurance settlement, and ensure the healthy and stable operation of the medical insurance system.

[0106] In an embodiment, steps 301 to 304 are described as follows:

[0107] Step 301, the first entity attribute is associated with the second entity attribute, and the candidate reimbursement rule in the medical insurance reimbursement rule is determined. The second entity attribute of the candidate reimbursement rule at least includes one entity attribute in the first entity attribute.

[0108] Optionally, the medical insurance intelligent management system performs entity association operation on the first entity attribute of the medical insurance settlement entity and the second entity attribute of the medical insurance reimbursement rule in the medical insurance knowledge graph, traverses all medical insurance reimbursement rules, checks whether the second entity attribute of the rule at least contains one entity attribute in the first entity attribute. If the condition is met, this medical insurance reimbursement rule is included in the candidate reimbursement rule set.

[0109] Continue in the medical insurance settlement entity of pneumonia hospitalization treatment, the first entity attribute includes disease "pneumonia", treatment mode "hospitalization treatment", and drug "ceftriaxone sodium" is used. The medical insurance reimbursement rule in the medical insurance knowledge graph has: rule A "for pneumonia hospitalization treatment and using antibiotics in the catalog, the reimbursement ratio is 80%", the second entity attribute contains "pneumonia", "hospitalization treatment" and "antibiotics in the catalog"; rule B "general hospitalization treatment, the reimbursement ratio is 70%", the second entity attribute contains "hospitalization treatment"; rule C "for cold outpatient treatment using specific drugs, the reimbursement ratio is 60%", the second entity attribute contains "cold", "outpatient treatment" and "specific drugs". The medical insurance intelligent management system compares and finds that the second entity attribute of rule A and rule B at least contains one (such as "hospitalization treatment") in the first entity attribute, and rule C does not contain, so rule A and rule B are determined as the candidate reimbursement rule.

[0110] Step 302, based on the second entity attribute of each rule in the candidate reimbursement rule and the first entity attribute, the target entity attribute matched with the first entity attribute in each rule is obtained.

[0111] Further, for each candidate reimbursement rule, the medical insurance intelligent management system performs entity mapping on the second entity attribute and the first entity attribute of the medical insurance settlement entity, checks each attribute in the second entity attribute of the candidate reimbursement rule one by one, judges whether it matches the attribute in the first entity attribute, and determines the matched attribute as the target entity attribute, wherein the target entity attribute reflects the association degree of the candidate reimbursement rule and the actual situation of the current medical insurance settlement entity.

[0112] Continuing the above example, for the candidate reimbursement rule A, its second entity attribute "pneumonia" matches the "pneumonia" in the first entity attribute of the medical insurance settlement entity, "hospitalization" matches, and "in-catalog antibiotics" matches "ceftriaxone sodium (which belongs to in-catalog antibiotics)", so the target entity attributes of rule A are "pneumonia", "hospitalization", and "in-catalog antibiotics"; for the candidate reimbursement rule B, its second entity attribute "hospitalization" matches the "hospitalization" in the first entity attribute of the medical insurance settlement entity, and the rest of the attributes do not match, so the target entity attribute of rule B is "hospitalization".

[0113] In step 303, the rule priority of each rule is determined based on the attribute inclusion relationship and the attribute priority between the target entity attributes of each rule, and the rule whose rule priority is higher than a preset priority in the candidate reimbursement rule is determined as the target reimbursement rule.

[0114] Further, the medical insurance intelligent management system analyzes the attribute inclusion relationship between the target entity attributes of each candidate reimbursement rule and the preset attribute priority, where the attribute inclusion relationship refers to whether the target entity attributes of one rule completely include the target entity attributes of another rule; the attribute priority is preset in the medical insurance knowledge graph, and different attributes have different priorities, for example, the disease type attribute priority is higher than the treatment method attribute. Optionally, the rule priority is determined according to the following logic: if the target entity attributes of one rule completely include the target entity attributes of another rule, and there is no higher priority attribute in the included rule, then the rule priority of the inclusion relationship is higher; if there is no inclusion relationship, the rule with the highest priority attribute in the target entity attribute is compared, and the rule with the higher priority attribute has a higher rule priority. A preset priority is set, and the medical insurance intelligent management system determines the rule whose rule priority is higher than the preset priority in the candidate reimbursement rule as the target reimbursement rule.

[0115] Continuing the above example, for rule A and rule B, the target entity attributes of rule A "pneumonia", "hospitalization", and "in-catalog antibiotics" completely include the target entity attributes of rule B "hospitalization", and there is no higher priority attribute in rule B (for example, the disease type attribute priority is higher than the treatment method attribute), so the priority of rule A is higher than that of rule B. The preset priority is set to medium, and the priority of rule A is higher than the preset priority, so rule A is determined as the target reimbursement rule.

[0116] In step 304, the settlement reimbursement rule is determined based on the target reimbursement rule.

[0117] Further, the medical insurance intelligent management system determines the settlement reimbursement rule according to the target reimbursement rule, which is specifically described in steps 3041 to 3044.

[0118] The embodiment of the present application can accurately match the most suitable reimbursement rule for the medical insurance settlement entity based on complex entity attribute relationship and priority logic, can effectively process the diversity and complexity of medical insurance reimbursement rules, ensure that complex and variable rules can be executed in order in actual settlement, avoid rule conflicts, ensure that the medical insurance fund is reimbursed according to reasonable and accurate rules, and improve the efficiency and accuracy of medical insurance settlement.

[0119] In an embodiment, steps 3041 to 3044 are described as follows:

[0120] Step 3041, if the number of rules in the target reimbursement rule is less than the preset number threshold, the target reimbursement rule is determined as the settlement reimbursement rule.

[0121] Optionally, the medical insurance intelligent management system counts the number of rules in the target reimbursement rule and compares it with the preset number threshold, wherein the preset number threshold in the embodiment of the present application is 2. If the number of rules in the target reimbursement rule is less than the preset number threshold, it means that there is only one rule in the target reimbursement rule currently screened out, therefore, the medical insurance intelligent management system directly determines the target reimbursement rule as the settlement reimbursement rule.

[0122] Continuing in the case of pneumonia hospitalization treatment, for example, the target reimbursement rule has only one rule A “for pneumonia hospitalization treatment and using antibiotics in the catalog, the reimbursement ratio is 80%”, therefore, the medical insurance intelligent management system directly determines rule A as the settlement reimbursement rule of the medical insurance settlement entity.

[0123] Step 3042, if the number of rules in the target reimbursement rule is greater than or equal to the preset number threshold, for any first reimbursement rule and second reimbursement rule in the target reimbursement rule, if the entity attributes of the first reimbursement rule and the second reimbursement rule overlap, and the first reimbursement rule and the second reimbursement rule are different rules, the first reimbursement rule and the second reimbursement rule are determined as conflict rules.

[0124] Further, if the number of rules in the target reimbursement rule is greater than or equal to the preset number threshold, any two rules (first reimbursement rule and second reimbursement rule) in the target reimbursement rule are analyzed, the entity attributes of the two rules are compared, if there is an overlapping part in the entity attributes, and the two rules belong to different rules (i.e. the rule contents are not completely same), the two rules are determined as conflict rules.

[0125] Continuing the above example, there are three target reimbursement rules: rule A "for pneumonia hospitalization and using in-catalog antibiotics, the reimbursement ratio is 80%", rule B "for pneumonia treatment and using in-catalog drugs, the reimbursement ratio is 75%", and rule C "for hospitalization and using in-catalog antibiotics, the reimbursement ratio is 70%", and the preset quantity threshold is 2. Upon analyzing rule A and rule B, it is found that they both contain overlapping entity attributes such as "pneumonia" and "in-catalog drugs (antibiotics belong to drugs)", and the rule contents are different, so rule A and rule B are determined as conflicting rules; similarly, rule A and rule C, and rule B and rule C also have overlapping entity attributes and are different rules, and are all determined as conflicting rules.

[0126] Step 3043, in the target reimbursement rules, the first target rule with the highest rule priority in the conflicting rules is retained, and the second target rule other than the first target rule is eliminated, to obtain the post-resolution rules.

[0127] Further, for the set of conflicting rules, the medical insurance intelligent management system processes according to the rule priority (which has been determined in step 303), finds the rule with the highest rule priority in the conflicting rules, sets it as the first target rule, and retains it; at the same time, eliminates other rules (set as the second target rule) other than the first target rule, to obtain the post-resolution rules. Continuing in the above set of conflicting rules containing rule A, rule B, and rule C, for example, the rule priority determined in step 303 is rule A > rule B > rule C. The medical insurance intelligent management system determines rule A as the first target rule and retains it, and eliminates rule B and rule C as the second target rule, to obtain the post-resolution rule as rule A.

[0128] Step 3044, if the number of rules in the post-resolution rules is less than the preset quantity threshold, the post-resolution rules are determined as the settlement reimbursement rules. If the number of rules in the post-resolution rules is greater than or equal to the preset quantity threshold, the rule including the most first entity attributes in the post-resolution rules is determined as the settlement reimbursement rule.

[0129] Further, the medical insurance intelligent management system counts the number of rules in the post-resolution rules and compares it with the preset quantity threshold. If the number of rules in the post-resolution rules is less than the preset quantity threshold, it means that the rules have been relatively clear after conflict resolution, and the post-resolution rules are directly determined as the settlement reimbursement rules.

[0130] Further, if the number of rules in the post-resolution rules is greater than or equal to the preset quantity threshold, the medical insurance intelligent management system further analyzes the post-resolution rules and finds the rule including the most first entity attributes of the medical insurance settlement entity, and determines it as the settlement reimbursement rule.

[0131] Continuing the above embodiment, if the post-resolution rules obtained through step 3043 are only rule A, at this time, the number of post-resolution rules is 1, and rule A is determined as the settlement reimbursement rule. If the post-resolution rules are rule A "for pneumonia hospitalization treatment and using antibiotics in the catalog, the reimbursement ratio is 80%" and rule D "for pneumonia hospitalization treatment, the reimbursement ratio is 70%", the preset number threshold is 2, at this time, the number of post-resolution rules is 2, which is equal to the preset number threshold. The first entity attribute of the medical insurance settlement entity includes "pneumonia", "hospitalization treatment" and "ceftriaxone sodium", rule A contains 3 first entity attributes, and rule D contains 2 first entity attributes. Rule A is determined as the settlement reimbursement rule.

[0132] The embodiment of the present application can effectively handle the complex situations such as excessive number and rule conflict that may exist in the target reimbursement rule, ensure the uniqueness and rationality of the finally determined settlement reimbursement rule, avoid the reimbursement confusion caused by unclear or conflicting rules, and thus adapt to the diversity and complexity of medical insurance reimbursement rules, quickly respond to changes in medical insurance settlement rules, and improve the efficiency and accuracy of medical insurance settlement.

[0133] Further, the knowledge graph-based medical insurance settlement list intelligent management system provided by the present application is described below. The knowledge graph-based medical insurance settlement list intelligent management system described below can be correspondingly referred to the knowledge graph-based medical insurance settlement list intelligent management method described above.

[0134] Optionally, referring to Figure 2 , Figure 2 is a structural schematic diagram of the knowledge graph-based medical insurance settlement list intelligent management system provided by the present application. The knowledge graph-based medical insurance settlement list intelligent management system comprises:

[0135] The settlement entity matching module 210 is configured to parse the obtained medical insurance settlement list data to obtain business keywords, and match the business keywords in the pre-constructed medical insurance knowledge graph to obtain medical insurance settlement entities.

[0136] The entity verification module 220 is configured to verify the medical insurance settlement entities based on the medical insurance knowledge graph to obtain entity verification results. The verification includes treatment method rationality verification, treatment dosage rationality verification and reimbursement range rationality verification.

[0137] The reimbursement rule matching module 230 is configured to, if the entity verification result is reasonable, perform priority logic matching based on the first entity attribute of the medical insurance settlement entity and the second entity attribute of the medical insurance reimbursement rule in the medical insurance knowledge graph to determine the settlement reimbursement rule of the medical insurance settlement entity.

[0138] The medical insurance settlement list settlement module 240 is configured to determine a medical insurance settlement reimbursement ratio based on the settlement reimbursement rule, and determine a medical insurance settlement reimbursement amount based on the medical insurance settlement reimbursement ratio and an actual settlement amount of the medical insurance settlement entity.

[0139] The medical insurance knowledge graph and the medical insurance settlement list data are matched, and the rationality of treatment methods, treatment dosage, reimbursement range and treatment cost is verified, so that the rationality of the business in the medical insurance settlement list data can be accurately verified. The priority logic is determined based on the entity attribute and the medical insurance knowledge graph, so that the complex and variable rules can be executed in order in the actual settlement, and rule conflicts are avoided. The medical insurance reimbursement is performed according to the medical insurance settlement reimbursement ratio determined based on the settlement reimbursement rule. Therefore, no matter how the medical insurance settlement rule changes, the medical insurance settlement list can be verified for rationality, the reimbursement rule is matched, the reimbursement ratio is determined, and the reimbursement amount is settled based on the medical insurance knowledge graph. The entire process does not require a large amount of manual intervention, can quickly respond to changes in the medical insurance settlement rule, and improves the efficiency and accuracy of the medical insurance settlement.

[0140] Referring to Figure 3 , Figure 3 An embodiment of an electronic device provided by the embodiment of the present application is shown in the figure. Figure 3 As shown in the figure, the electronic device 300 provided by the embodiment of the present application includes a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, the following steps are implemented:

[0141] The medical insurance settlement list data obtained is parsed to obtain business keywords, and the business keywords are associated and matched in the pre-constructed medical insurance knowledge graph to obtain medical insurance settlement entities;

[0142] The medical insurance settlement entities are verified based on the medical insurance knowledge graph to obtain entity verification results. The verification includes rationality verification of treatment methods, treatment dosage, reimbursement range and treatment cost;

[0143] If the entity verification result is reasonable, the settlement reimbursement rule of the medical insurance settlement entity is determined based on the first entity attribute of the medical insurance settlement entity and the second entity attribute of the medical insurance reimbursement rule in the medical insurance knowledge graph.

[0144] The medical insurance settlement reimbursement ratio is determined based on the settlement reimbursement rule, and the medical insurance settlement reimbursement amount is determined based on the medical insurance settlement reimbursement ratio and the actual settlement amount of the medical insurance settlement entity.

[0145] Referring to Figure 4 , Figure 4 An embodiment of a computer readable storage medium provided by the embodiment of the present application is shown in the figure.Figure 4 As shown, the embodiment provides a computer readable storage medium 400, which stores a computer program 311, and the computer program 311 is executed by a processor to implement the following steps:

[0146] The obtained medical insurance settlement list data is parsed to obtain business keywords, and the business keywords are associated and matched in the pre-constructed medical insurance knowledge graph to obtain medical insurance settlement entities;

[0147] The medical insurance settlement entities are verified based on the medical insurance knowledge graph to obtain entity verification results; the verification includes treatment method, treatment dosage, reimbursement range, and treatment cost rationality verification;

[0148] If the entity verification result is reasonable, the first entity attribute of the medical insurance settlement entity and the second entity attribute of the medical insurance reimbursement rule in the medical insurance knowledge graph are matched based on the priority logic to determine the settlement reimbursement rule of the medical insurance settlement entity;

[0149] The medical insurance settlement reimbursement proportion is determined based on the settlement reimbursement rule, and the medical insurance settlement reimbursement amount is determined based on the medical insurance settlement reimbursement proportion and the actual settlement amount of the medical insurance settlement entity.

[0150] On the other hand, the present application also provides a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program is executed by a processor, and the computer can execute the medical insurance settlement list intelligent management method based on the knowledge graph provided by the above-mentioned method, which comprises:

[0151] The obtained medical insurance settlement list data is parsed to obtain business keywords, and the business keywords are associated and matched in the pre-constructed medical insurance knowledge graph to obtain medical insurance settlement entities;

[0152] The medical insurance settlement entities are verified based on the medical insurance knowledge graph to obtain entity verification results; the verification includes treatment method, treatment dosage, reimbursement range, and treatment cost rationality verification;

[0153] If the entity verification result is reasonable, the first entity attribute of the medical insurance settlement entity and the second entity attribute of the medical insurance reimbursement rule in the medical insurance knowledge graph are matched based on the priority logic to determine the settlement reimbursement rule of the medical insurance settlement entity;

[0154] The medical insurance settlement reimbursement proportion is determined based on the settlement reimbursement rule, and the medical insurance settlement reimbursement amount is determined based on the medical insurance settlement reimbursement proportion and the actual settlement amount of the medical insurance settlement entity.

[0155] The system embodiments described above are merely illustrative, wherein the units described as separate components can or can not be physically separate, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purposes of the embodiments according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0156] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and necessary universal hardware platforms, and of course can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of software products, and the computer software products can be stored in a computer readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and include a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0157] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A medical insurance settlement list intelligent management method based on a knowledge graph, characterized by, The method comprises the following steps: parsing the obtained medical insurance settlement list data to obtain business keywords, and associating and matching the business keywords in a pre-constructed medical insurance knowledge graph to obtain medical insurance settlement entities; verifying the medical insurance settlement entities based on the medical insurance knowledge graph to obtain entity verification results; verification includes treatment method, treatment dosage, reimbursement range, and treatment cost reasonableness verification; if the entity verification result is reasonable, performing priority logic matching based on the first entity attribute of the medical insurance settlement entity and the second entity attribute of the medical insurance reimbursement rule in the medical insurance knowledge graph to determine the settlement and reimbursement rule of the medical insurance settlement entity; determining the medical insurance settlement reimbursement proportion based on the settlement and reimbursement rule, and determining the medical insurance settlement reimbursement amount based on the medical insurance settlement reimbursement proportion and the actual settlement amount of the medical insurance settlement entity; wherein, the entity verification result obtained by verifying the medical insurance settlement entity based on the medical insurance knowledge graph comprises: if any first target entity and second target entity in the medical insurance settlement entity do not exist in the medical insurance knowledge graph, then taking the first target entity and the second target entity as endpoints, performing path traversal in the medical insurance knowledge graph with a preset step size to obtain initial indirect paths, and performing indirect association mode determination on the medical insurance business logic of each initial indirect path to determine a target indirect path; publicly associated entities in the target indirect path whose frequency of occurrence is greater than a preset frequency threshold; determining first scene matching strength and second scene matching strength based on the matching relationship between the entity type of the publicly associated entity and the entity scene of the first target entity and the second target entity, and determining first entity association strength and second entity association strength based on the entity distance between the publicly associated entity and the first target entity and the second target entity; if the first scene matching strength and the second scene matching strength are both greater than a preset matching strength threshold, and the first entity association strength and the second entity association strength are both less than or equal to a preset association strength threshold, it is determined that the medical insurance settlement entity is reasonable in treatment method; otherwise, it is not; if the medical insurance settlement entity is reasonable in treatment method, performing treatment dosage reasonableness verification on the medical insurance settlement entity based on the medical insurance knowledge graph to obtain the entity verification result; the entity verification result obtained by performing treatment dosage reasonableness verification on the medical insurance settlement entity based on the medical insurance knowledge graph comprises: constructing a local association network based on the nodes and node edges related to the first target entity and the second target entity in the medical insurance knowledge graph, and filtering the nodes and node edges related to the treatment dosage in the local association network to obtain actual treatment dosage data; determining the standard treatment dosage range in the treatment scene where the first target entity and the second target entity are located based on the medical insurance knowledge graph, and determining the initial treatment dosage difference data based on the actual treatment dosage data and the standard treatment dosage range; Based on the initial treatment dosage difference data, influence factor traversal is performed in the medical insurance knowledge graph to obtain dosage difference associated factors, and the initial treatment dosage difference data is corrected based on the dosage difference associated factors to obtain target treatment dosage difference data; the dosage difference associated factors include patient characteristic factors and treatment environment factors; If the target treatment dosage difference data is within a dosage difference threshold range, it is determined that the medical insurance settlement entity conforms to treatment dosage rationality; otherwise, it does not conform; If the medical insurance settlement entity conforms to treatment dosage rationality, the medical insurance settlement entity is verified for reimbursement range rationality based on the medical insurance knowledge graph to obtain the entity verification result. 2.The knowledge graph-based medical insurance settlement list intelligent management method according to claim 1, characterized in that, The medical insurance settlement entity is verified for reimbursement range rationality based on the medical insurance knowledge graph to obtain the entity verification result, including: According to the local correlation network, the reimbursement range rules of the first target entity and the second target entity are extracted; the reimbursement range rules include reimbursable items, reimbursement processes, and reimbursement additional conditions; Based on the medical insurance knowledge graph, it is determined whether the first target entity and the second target entity belong to entity nodes in the reimbursable items, and whether the first target entity and the second target entity belong to entity nodes in the reimbursement processes, and whether the first target entity and the second target entity belong to entity nodes in the reimbursement additional conditions based on the medical insurance knowledge graph; If the first target entity and the second target entity belong to nodes in the reimbursable items, the reimbursement processes, or / and the reimbursement additional conditions, it is determined that the medical insurance settlement entity conforms to reimbursement range rationality; or, if the first target entity or the second target entity belongs to nodes in the reimbursable items, the reimbursement processes, or / and the reimbursement additional conditions, it is determined that the first target entity or the second target entity conforms to reimbursement range rationality; otherwise, it does not conform; If the medical insurance settlement entity conforms to treatment method rationality, or the first target entity or the second target entity conforms to reimbursement range rationality, the medical insurance settlement entity is verified for treatment cost rationality based on the medical insurance knowledge graph to obtain the entity verification result. 3.The knowledge graph-based medical insurance settlement list intelligent management method according to claim 2, characterized in that, The medical insurance settlement entity is verified for treatment cost rationality based on the medical insurance knowledge graph to obtain the entity verification result, including: Based on the medical insurance knowledge graph, the entity path between the first target entity and the second target entity is determined, and the path nodes related to treatment cost in the entity path are determined; Based on the cost standard interval of the path nodes in each entity path in the medical insurance knowledge graph and the influence coefficient of the cost between the path nodes, the expected cost value of each entity path is determined, and the expected cost range of the medical insurance settlement entity is determined based on the expected cost value of each entity path; If the actual treatment cost in the medical insurance settlement entity is within the expected cost range, it is determined that the entity verification result conforms to rationality; or, If the actual treatment cost is not within the expected cost range, a cost deviation path set is obtained; the cost deviation path in the cost deviation path set has an intersection with an interval centered on the actual treatment cost and having a preset error cost as a radius; If the number of cost deviation paths in the cost deviation path set is less than a preset number threshold, it is determined that the entity verification result is in line with rationality. 4.The knowledge graph-based medical insurance settlement list intelligent management method according to any one of claims 1 to 3, characterized in that, The priority logic matching is performed based on the first entity attribute of the medical insurance settlement entity and the second entity attribute of the medical insurance reimbursement rule in the medical insurance knowledge graph, and a settlement and reimbursement rule of the medical insurance settlement entity is determined, including: The first entity attribute and the second entity attribute are associated to determine a candidate reimbursement rule in the medical insurance reimbursement rule; the second entity attribute of the candidate reimbursement rule at least includes one entity attribute in the first entity attribute; The second entity attribute of each rule in the candidate reimbursement rule is mapped with the first entity attribute to obtain a target entity attribute in each rule that matches the first entity attribute; Based on the attribute inclusion relationship and the attribute priority between the target entity attributes of each rule, a rule priority of each rule is determined, and a rule with a rule priority higher than a preset priority in the candidate reimbursement rule is determined as a target reimbursement rule; The settlement and reimbursement rule is determined based on the target reimbursement rule. 5.The knowledge graph-based medical insurance settlement list intelligent management method according to claim 4, characterized in that, The settlement and reimbursement rule is determined based on the target reimbursement rule, including: If the number of rules in the target reimbursement rule is less than a preset number threshold, the target reimbursement rule is determined as the settlement and reimbursement rule; or, If the number of rules in the target reimbursement rule is greater than or equal to the preset number threshold, for any first reimbursement rule and second reimbursement rule in the target reimbursement rule, if the entity attributes of the first reimbursement rule and the second reimbursement rule overlap and the first reimbursement rule and the second reimbursement rule are different rules, the first reimbursement rule and the second reimbursement rule are determined as conflict rules; In the target reimbursement rule, a first target rule with the highest rule priority in the conflict rules is retained, and a second target rule other than the first target rule is excluded to obtain a resolved rule; If the number of rules in the resolved rule is less than the preset number threshold, the resolved rule is determined as the settlement and reimbursement rule; if the number of rules in the resolved rule is greater than or equal to the preset number threshold, a rule including the first entity attribute the most in the resolved rule is determined as the settlement and reimbursement rule.

6. An intelligent management system for medical insurance settlement list based on a knowledge graph, characterized in that, The method is applied to the medical insurance settlement list intelligent management method based on a knowledge graph. The medical insurance settlement list intelligent management system based on a knowledge graph includes: A settlement entity matching module is configured to parse the obtained medical insurance settlement list data to obtain a business keyword, and associate and match the business keyword in a pre-constructed medical insurance knowledge graph to obtain a medical insurance settlement entity. The entity verification module is configured to verify the medical insurance settlement entity based on the medical insurance knowledge graph to obtain an entity verification result. The verification includes treatment method rationality verification, treatment dosage rationality verification, and reimbursement range rationality verification. The reimbursement rule matching module is configured to, if the entity verification result is reasonable, perform priority logic matching based on a first entity attribute of the medical insurance settlement entity and a second entity attribute of a medical insurance reimbursement rule in the medical insurance knowledge graph to determine a settlement reimbursement rule of the medical insurance settlement entity. The medical insurance list settlement module is configured to determine a medical insurance settlement reimbursement proportion based on the settlement reimbursement rule, and determine a medical insurance settlement reimbursement amount based on the medical insurance settlement reimbursement proportion and an actual settlement amount of the medical insurance settlement entity. The verification based on the medical insurance knowledge graph on the medical insurance settlement entity to obtain the entity verification result includes: If any first target entity and second target entity in the medical insurance settlement entity do not exist in a direct relationship in the medical insurance knowledge graph, the first target entity and the second target entity are taken as endpoints, path traversal is performed in the medical insurance knowledge graph with a preset step length to obtain an initial indirect path, and indirect association mode determination is performed on the medical insurance business logic of each initial indirect path to determine a target indirect path. A common associated entity in the target indirect path has a frequency greater than a preset frequency threshold. Based on a matching relationship between an entity type of the common associated entity and entity scenes of the first target entity and the second target entity, a first scene matching strength and a second scene matching strength are determined, and based on entity distances between the common associated entity and the first target entity and the second target entity, a first entity association strength and a second entity association strength are determined. If the first scene matching strength and the second scene matching strength are both greater than a preset matching strength threshold, and the first entity association strength and the second entity association strength are both less than or equal to a preset association strength threshold, it is determined that the medical insurance settlement entity is reasonable in treatment method; otherwise, it is not reasonable. If the medical insurance settlement entity is reasonable in treatment method, treatment dosage rationality verification is performed on the medical insurance settlement entity based on the medical insurance knowledge graph to obtain the entity verification result. The verification based on the medical insurance knowledge graph on the medical insurance settlement entity to obtain the entity verification result includes: Based on nodes and node edges related to the first target entity and the second target entity in the medical insurance knowledge graph, a local association network is constructed, and nodes and node edges related to treatment dosage in the local association network are filtered to obtain actual treatment dosage data. Based on the medical insurance knowledge graph, a standard treatment dosage range in a treatment scene in which the first target entity and the second target entity are located is determined, and based on the actual treatment dosage data and the standard treatment dosage range, initial treatment dosage difference data is determined. Based on the initial treatment dosage difference data, influence factor traversal is performed in the medical insurance knowledge graph to obtain dosage difference associated factors, and the initial treatment dosage difference data is corrected based on the dosage difference associated factors to obtain target treatment dosage difference data; the dosage difference associated factors include patient characteristic factors and treatment environment factors; If the target treatment dosage difference data is within a dosage difference threshold range, it is determined that the medical insurance settlement entity conforms to treatment dosage rationality; otherwise, it does not conform; If the medical insurance settlement entity conforms to treatment dosage rationality, the medical insurance settlement entity is verified for reimbursement range rationality based on the medical insurance knowledge graph to obtain the entity verification result.

7. An electronic device, comprising: a memory for storing a computer software program; a processor for reading and executing the computer software program, wherein when the processor executes the computer software program, the method for intelligently managing a medical insurance settlement list based on a knowledge graph according to any one of claims 1 to 5 is implemented.

8. A non-transitory computer readable storage medium having stored therein a computer software program, characterized in that, When the computer software program is executed by the processor, the method for intelligently managing a medical insurance settlement list based on a knowledge graph according to any one of claims 1 to 5 is implemented.

Images