A method for medical insurance cost control and review and a medical insurance cost control and review system

Abstract

The present application provides a method for medical insurance cost control review, including: obtaining a medical order item dictionary of a target hospital, a medical settlement form item dictionary of the target hospital, and a mapping relationship between items in the medical order item dictionary and items in the medical settlement form item dictionary; determining a target medical knowledge graph based on the medical order item dictionary, the medical settlement form item dictionary, the mapping relationship, and a public medical knowledge graph; determining whether a judgment object T in a target medical settlement form and a corresponding search object S in a target medical order are consistent based on the target medical knowledge graph. The solution provided by the present application can improve the efficiency and accuracy of medical insurance cost control review.

Description

Technical Field

[0001] The present application belongs to the field of information processing, and specifically relates to a method for medical insurance cost control and audit and a medical insurance cost control and audit system. Background Art

[0002] As medical insurance becomes more and more popular among the people, medical insurance audits and inspections have gradually become an important part of the insurance claims process.

[0003] During the medical insurance audit, if an item (such as CT) has been charged, that is, it is recorded in the settlement form, but the corresponding medical order is not found, this is a problem of inconsistency between the settlement form item and the medical order.

[0004] Currently, there are two ways of medical insurance auditing. One is manual auditing. However, due to limited manpower, it is impossible to check all settlement statements. Only limited spot checks can be carried out, and the inspection coverage is limited and the inspection efficiency is very low. The other is a computer medical insurance cost control review system. In the computer automation process, only string fuzzy matching technology is used. It cannot determine whether two items are consistent from the medical connotation, and a lot of misjudgments will occur. It can solve the problem of inconsistency between settlement statements and medical orders.

[0005] Therefore, there is an urgent need for a method for medical insurance cost control audit and a medical insurance cost control audit system to improve the efficiency and accuracy of medical insurance cost control audit. Summary of the invention

[0006] The embodiments of the present application provide a method for medical insurance cost control auditing and a medical insurance cost control auditing system thereof, which can improve the efficiency and accuracy of medical insurance cost control auditing.

[0007] In a first aspect, a method for medical insurance cost control review is provided, comprising: obtaining a medical order item dictionary of a target hospital, a medical settlement form item dictionary of the target hospital, and a mapping relationship between items in the medical order item dictionary and items in the medical settlement form item dictionary; determining a target medical knowledge graph based on the medical order item dictionary, the medical settlement form item dictionary, the mapping relationship, and a public medical knowledge graph; determining whether a judgment object T in a target medical settlement form and a corresponding search object S in a target medical order are consistent based on the target medical knowledge graph.

[0008] In combination with the first aspect, in a first possible implementation method of the first aspect, determining whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent based on the target medical knowledge graph includes: determining the similarity simF(S, T) between the judgment object T and the search object S based on the target medical knowledge graph, and when simF(S, T) is greater than a first preset threshold, determining that the judgment object T is consistent with the search object S.

[0009] In combination with the first aspect and its above-mentioned implementation method, in a second possible implementation method of the first aspect, determining whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent based on the target medical knowledge graph, including: when simF(S, T) is less than or equal to a first preset threshold, performing upper and lower matching on the judgment object T and the search object S according to the target medical knowledge graph, and determining the similarity simG(S, T) between the judgment object T and the search object S based on the result of the upper and lower matching; when simG(S, T) is greater than a second preset threshold, determining that the judgment object T is consistent with the search object S.

[0010] In combination with the first aspect and its above-mentioned implementation method, in a third possible implementation method of the first aspect, determining whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent based on the target medical knowledge graph, also includes: when simG(S, T) is less than or equal to the second preset threshold, using the target BERT training model to determine whether the judgment object T is consistent with the search object S.

[0011] In combination with the first aspect and its above-mentioned implementation method, in a fourth possible implementation method of the first aspect, the method also includes: selecting N corresponding relationships in the mapping relationship for consistency annotation to obtain a training set; performing BERT model training on the training set to obtain the target BERT training model.

[0012] In combination with the first aspect and the above implementation manner, in a fifth possible implementation manner of the first aspect, determining the similarity simF(S, T) between the judgment object T and the search object S includes: determining simF(S, T) using the following formula, simF(S, T)=f(1, S, T)*f(2, S, T)*h(S, T), wherein,

[0013] f(n, S, T) = count(Set(n, S) ∩Set(n, T)) / count(Set(n, S) ∪Set(n, T)), where Set(n, S) represents the set of all nodes in the target medical knowledge graph whose shortest distance to the search object S is n, and Set(n, T) represents the set of all nodes in the target medical knowledge graph whose shortest distance to the judgment object T is n. Count() represents the statistical number, and n is a positive integer. h(S, T) = ∑e -dis(propriety[i],S,T) , propriety[i] represents the i-th item in the medical attribute set of the judgment object T, and dis(propriety[i], S, T) represents the shortest distance between the judgment object T and the search object S along the i-th medical attribute in the target medical knowledge graph.

[0014] In combination with the first aspect and the above-mentioned implementation manner, in a sixth possible implementation manner of the first aspect, the judgment object T and the search object S are matched at the upper and lower levels according to the target medical knowledge graph, and the similarity simG(S, T) between the judgment object T and the search object S is determined according to the result of the upper and lower level matching, including: using the following formula to determine simG(S, T), simG(S, T) = α*g(S, T)+1-α)*h(S, T), wherein 0<α<1, g(S, T) is used to characterize whether the judgment object T is a hyponym of the search object S in the target knowledge graph, and h(S, T) = ∑e -dis(propriety[i],S,T) , propriety[i] represents the i-th item in the medical attribute set of the judgment object T, and dis(propriety[i], S, T) represents the shortest distance between the judgment object T and the search object S along the i-th medical attribute in the target medical knowledge graph.

[0015] In a second aspect, a system for medical insurance cost control and audit is provided, including: an acquisition unit, the acquisition unit is used to acquire a doctor's order item dictionary of a target hospital, a medical settlement form item dictionary of the target hospital, and a mapping relationship between each item in the doctor's order item dictionary and each item in the medical settlement form item dictionary; a processing unit, the processing unit is used to determine a target medical knowledge graph based on the doctor's order item dictionary, the medical settlement form item dictionary, the mapping relationship and the public medical knowledge graph; the processing unit is also used to determine whether a judgment object T in a target medical settlement form and a corresponding search object S in a target doctor's order are consistent based on the target medical knowledge graph, including In summary: According to the target medical knowledge graph, the similarity simF(S, T) between the judgment object T and the search object S is determined. When simF(S, T) is greater than a first preset threshold, it is determined that the judgment object T is consistent with the search object S; when simF(S, T) is less than or equal to the first preset threshold, according to the target medical knowledge graph, the search object S is matched in the upper and lower positions, and according to the result of the upper and lower matching, the similarity simG(S, T) between the judgment object T and the search object S is determined. When simG(S, T) is greater than a second preset threshold, it is determined that the judgment object T is consistent with the search object S.

[0016] Optionally, the medical insurance cost control review system can implement the first aspect or any possible implementation method of the first aspect.

[0017] In a third aspect, a device for medical insurance cost control and auditing is provided, comprising: a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein when the program or instruction is executed by the processor, the steps of the medical insurance cost control and auditing method as described in the first aspect or any one of the first aspects are implemented.

[0018] In a fourth aspect, a readable storage medium is provided, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the medical insurance cost control review method as described in the first aspect or any one of the first aspects are implemented.

[0019] Therefore, the medical insurance cost control review method provided in this application constructs a medical knowledge graph of the target hospital, and makes consistency judgments based on the matching relationship between the medical settlement form and the medical order in the medical knowledge graph. While ensuring accuracy, it greatly improves efficiency compared to manual inspection.

[0020] Furthermore, in the process of matching the relationship between the medical settlement form and the medical order in the medical knowledge graph, the present application further judges the consistency between the medical settlement form and the medical order through a deep learning model to improve the accuracy of the review. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic flowchart of a method for medical insurance cost control review according to an embodiment of the present application is shown.

[0022] Figure 2 A schematic diagram of a medical knowledge graph according to an embodiment of the present application is shown.

[0023] Figure 3 A schematic flowchart of a method for medical insurance cost control and review according to another embodiment of the present application is shown.

[0024] Figure 4 A schematic structural diagram of a system for medical insurance cost control and auditing according to an embodiment of the present application is provided.

[0025] Figure 5 A schematic structural diagram of an electronic device provided by an embodiment of the present application is provided DETAILED DESCRIPTION

[0026] The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments in the present application belong to the scope of protection of this application.

[0027] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0028] The existing computer medical insurance cost control and audit system cannot solve the problem of inconsistency between the settlement form and the doctor's order. There are two main reasons. First, they only use the technology of string fuzzy matching, which cannot judge whether the two items are consistent from the medical connotation, and there will be a lot of misjudgments. For example, "venipuncture catheterization" and "extraction of venipuncture catheterization" are inconsistent in connotation, but the string fuzzy matching will judge them as consistent. "Mast" and "aspartate aminotransferase" are consistent in connotation, but the string fuzzy matching will judge them as inconsistent; second, the inconsistency between the settlement form and the doctor's order can also be expressed in a hierarchical relationship. For example, the doctor's order is "four infection items (hospitalization)", and the corresponding settlement form items are "hepatitis C antibody (quantitative)", "human immunodeficiency virus antibody", "syphilis spirochete antibody (specific)", "hepatitis B surface antigen", "hepatitis B surface antibody", "hepatitis B e antigen", "hepatitis B e antibody", "hepatitis B core antibody". This hierarchical relationship cannot be solved by the existing computer medical insurance cost control and audit system.

[0029] The following, in conjunction with the accompanying drawings, describes in detail the method for medical insurance cost control review provided by the embodiment of the present application through specific embodiments and their application scenarios.

[0030] Figure 1 A schematic flow chart of a method for medical insurance cost control review according to an embodiment of the present application is shown as follows: Figure 1 As shown, the method 100 includes:

[0031] Step 110, obtaining a doctor's order item dictionary of a target hospital, a medical settlement form item dictionary of the target hospital, and a mapping relationship between items in the doctor's order item dictionary and items in the medical settlement form item dictionary;

[0032] Step 120, determining a target medical knowledge graph according to the medical order item dictionary, the medical settlement form item dictionary, the mapping relationship and the public medical knowledge graph;

[0033] Step 130, based on the target medical knowledge graph, determine whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent.

[0034] Specifically, in step 110, the target hospital generally has its own medical order item dictionary, or in other words, the target hospital has its own personalized medical order item set. These medical order dictionaries or medical order items include multiple medical order items, such as medical order: four items of infection (West Hospital), and a medical settlement form will also be generated in the target hospital. The medical settlement form items include charge items, such as drugs, examination items, specifically, such as X-ray computed tomography (CT) imaging, sodium determination; in addition, there is a mapping relationship between the medical order items actually generated in the hospital and the corresponding medical settlement form items. Therefore, each item in the medical order item dictionary in the hospital is also mapped to the medical settlement form item dictionary. There is a mapping relationship between the items, which can exist in a table form. For example, the doctor's order: four items of infection (West Hospital), corresponding to the charging items: hepatitis C antibody, human immunodeficiency virus antibody, syphilis spirochete antibody, hepatitis B antibody), and another example is the five items of hepatitis B corresponding to the charging items: hepatitis B surface antibody (anti-HBs); hepatitis B virus e antigen; hepatitis B virus surface antibody; hepatitis B e antibody (anti-HBe); hepatitis B virus core antibody; hepatitis B e antigen (HBeAg); hepatitis B core antibody (anti-HBc); hepatitis B virus e antibody; two items of hand, foot and mouth disease corresponding to the charging items: viral antibody quantitative determination, EV71 virus antibody detection. This application does not make specific restrictions.

[0035] In step 120, the sources of the public medical knowledge graph include publicly available authoritative medical knowledge graphs, medical books, journals, real medical record data, manual additions by experts, etc., which are not limited in this application.

[0036] The target medical knowledge graph includes multiple nodes, which are connected by medical attributes such as project attributes, hierarchical relationships, or diagnostic methods. Figure 2 As shown, a schematic diagram of a medical knowledge graph of an embodiment of the present application is shown, which includes multiple nodes and the relationships between these multiple nodes. If there are new settlement items and medical order items, they will be used as new nodes and associated with the nodes in the public medical knowledge graph through medical attributes. It should be understood that the specific content of the above medical attributes varies depending on the different medical knowledge graphs, and may include etiology, pathology, location, clinical manifestations, independent symptoms, physical activities, metabolites, abnormal conditions, instruments or techniques, surgical procedures, surgical approaches, etc., and this application does not limit them.

[0037] Optionally, as an embodiment of the present application, determining whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent according to the target medical knowledge graph includes: determining the similarity simF(S, T) between the judgment object T and the search object S according to the target medical knowledge graph, and when simF(S, T) is greater than a first preset threshold, determining that the judgment object T is consistent with the search object S.

[0038] Specifically, the judgment object T refers to a specific item in the medical insurance settlement form, and the search object S refers to the item in the medical order. It is necessary to determine whether there is a corresponding relationship between the judgment object and the search object S in the medical order. Specifically, it is necessary to consider the distance between the judgment object T and the search object S in the target medical map, as well as the proportion of the nearest identical neighbors of the search object S and the judgment object T in the target medical map.

[0039] Optionally, determining the similarity simF(S, T) between the judgment object T and the search object S includes: determining simF(S, T) using the following formula: simF(S, T)=f(1, S, T)*f(2, S, T)*h(S, T), wherein,

[0040] f(n, S, T) = count(Set(n, S)∩Set(n, T)) / count(Set(n, S)∪Set(n, T)), where Set(n, S) represents the set of all nodes in the target medical knowledge graph whose shortest distance to the search object S is n, and Set(n, T) represents the set of all nodes in the target medical knowledge graph whose shortest distance to the judgment object T is n. Count() represents the statistical number, and n is a positive integer.

[0041] h(S,T)=∑e -dis(propriety[i],S,T) , propriety[i] represents the i-th item in the medical attribute set of the judgment object T, and dis(propriety[i], S, T) represents the shortest distance between the judgment object T and the search object S along the i-th medical attribute in the target medical knowledge graph.

[0042] For example, the medical attribute propriety here = {cause, pathology, location, clinical manifestations, independent symptoms, physical activities, metabolites, abnormal conditions, instruments or techniques, surgical procedures, surgical approaches, nature of the disease, examination methods}. It should be understood that the specific content of the above medical attributes varies depending on the target medical knowledge graph, and this application does not limit it.

[0043] Optionally, as an embodiment of the present application, determining whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent according to the target medical knowledge graph, including: when simF(S, T) is less than or equal to a first preset threshold, performing upper and lower matching on the judgment object T and the search object S according to the target medical knowledge graph, and determining the similarity simG(S, T) between the judgment object T and the search object S according to the result of the upper and lower matching, and when simG(S, T) is greater than a second preset threshold, determining that the judgment object T is consistent with the search object S.

[0044] Specifically, the upper and lower matching of the judgment object T and the search object S refers to: determining whether there is a subordinate relationship between the judgment object T and the search object by searching the target medical atlas.

[0045] It should be understood that the above-mentioned first threshold value may be 0.56, and the first threshold value may be adjusted according to actual needs, and this application does not limit it.

[0046] Optionally, as an embodiment of the present application, performing upper and lower matching on the judgment object T and the search object S according to the target medical knowledge graph, and determining the similarity simG(S, T) between the judgment object T and the search object S according to the result of the upper and lower matching, includes:

[0047] Use the following formula to determine simG(S, T): simG(S, T) = α*g(S, T) + (1-α)*h(S, T), where 0<α<1, g(S, T) is used to characterize whether the judgment object T is a hyponym of the search object S in the target knowledge graph, and h(S, T) = ∑e -dis(propriety[i],S,T) , propriety[i] represents the i-th item in the medical attribute set of the judgment object T, and dis(propriety[i], S, T) represents the shortest distance between the judgment object T and the search object S along the i-th medical attribute in the target medical knowledge graph.

[0048] Specifically, g(S,T) indicates whether S is on the hyponym node of T in the knowledge graph. If yes, it returns 1, otherwise it returns 0.

[0049] Optionally, as an embodiment of the present application, the target medical knowledge graph is used to determine whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent, including: when simG(S, T) is less than or equal to a second preset threshold, using the target BERT training model to determine whether the judgment object T is consistent with the search object S.

[0050] It should be understood that the second threshold value mentioned above may be 0.63, and the first threshold value may be adjusted according to actual needs, which is not limited in this application.

[0051] Specifically, after inputting the judgment object T and the search object S into the above-mentioned target BERT training model, a consistent result will be obtained.

[0052] Optionally, as an embodiment of the present application, the method further includes: selecting N corresponding relationships in the mapping relationship for consistency annotation to obtain a training set; and performing BERT model training on the training set to obtain the target BERT training model.

[0053] Specifically, some items were randomly selected from the mapping relationship of the target hospital, and the training set was obtained by annotating the consistency through manual experts. The Bidirectional Encoder Representations from Transformers (BERT) model was used for learning and training until the model converged, and the target BERT model that can judge the consistency of items was obtained.

[0054] It should be understood that other machine learning algorithm models can be used in this application to determine the consistency of items in medical settlement forms and medical orders.

[0055] Figure 3 FIG. 1 is a schematic flow chart of a method according to an embodiment of the present application. Figure 3 As shown, the method includes: steps 301-302, connecting with the hospital system;

[0056] Step 303, obtaining the hospital's project dictionary;

[0057] Step 305, obtaining the hospital's medical settlement form, medical order items, and the mapping relationship between the two;

[0058] Step 304, obtaining an existing medical knowledge graph, for example, the aforementioned public medical knowledge graph;

[0059] Step 306, searching the knowledge graph for medical settlement items and medical order items;

[0060] Step 307, through the physical therapy knowledge graph, find out whether the shortest path between the medical settlement form item and the doctor's order item meets the preset conditions. If the preset conditions are met, the two are matched; execute step 308, mark as passed the medical insurance cost control review; execute step 309, add the result to the medical insurance review list, and step 314, output the list result;

[0061] If the two do not match in step 307, then execute step 310 to perform knowledge graph upper and lower matching on the medical settlement form item and the doctor's order item. Further, in step 311, if the medical settlement form item and the doctor's order item are in a matching upper and lower relationship, then execute step 308 to mark as having passed the medical insurance cost control review; execute step 309 to add the result to the medical insurance review list, and step 314 to output the list result;

[0062] If in step 311, if the medical settlement form item and the doctor's order item are matched and are not in a hierarchical relationship, then execute step 312, input the two into the consistency judgment model, that is, the aforementioned target BERT training model, and judge whether the match is determined by the output result of the target BERT training model. In step 313, if it matches, execute step 308 and mark it as passed the medical insurance cost control review; execute step 309 and add the result to the medical insurance review list. In step 314, output the list result; if the matching result of step 313 is inconsistent, then the review result between the medical settlement form item and the doctor's order item is failed.

[0063] Therefore, the medical insurance cost control review method provided in this application constructs a medical knowledge graph of the target hospital, and makes consistency judgments based on the matching relationship between the medical settlement form and the medical order in the medical knowledge graph. While ensuring accuracy, it greatly improves efficiency compared to manual inspection.

[0064] Furthermore, in the process of matching the relationship between the medical settlement form and the medical order in the medical knowledge graph, the present application further judges the consistency between the medical settlement form and the medical order through a deep learning model to improve the accuracy of the review.

[0065] It should be noted that the method provided in the embodiment of the present application can be performed by a medical insurance cost control and audit system. In the embodiment of the present application, the medical insurance cost control and audit system performing the above method is taken as an example to illustrate the medical insurance cost control and audit system provided in the embodiment of the present application.

[0066] Figure 4 A schematic structural block diagram of a medical insurance cost control and audit system provided for one embodiment of the present application.

[0067] like Figure 4As shown, the system 400 includes: an acquisition unit 410, the acquisition unit 410 is used to acquire the doctor's order item dictionary of the target hospital, the medical settlement form item dictionary of the target hospital, and the mapping relationship between each item in the doctor's order item dictionary and each item in the medical settlement form item dictionary; a processing unit 420, the processing unit 420 is used to determine the target medical knowledge graph according to the doctor's order item dictionary, the medical settlement form item dictionary, the mapping relationship and the public medical knowledge graph; the processing unit 420 is also used to determine whether the judgment object T in the target medical settlement form and the corresponding search object S in the target doctor's order are the same according to the target medical knowledge graph. The method comprises: determining the similarity simF(S, T) between the judgment object T and the search object S according to the target medical knowledge graph, and when simF(S, T) is greater than a first preset threshold, determining that the judgment object T is consistent with the search object S; when simF(S, T) is less than or equal to the first preset threshold, performing upper and lower matching on the search object S according to the target medical knowledge graph, and determining the similarity simG(S, T) between the judgment object T and the search object S according to the result of the upper and lower matching, and when simG(S, T) is greater than a second preset threshold, determining that the judgment object T is consistent with the search object S.

[0068] The medical insurance cost control and audit system provided in the embodiment of the present invention can implement each process in the embodiment corresponding to the above message processing method. To avoid repetition, it will not be described here.

[0069] It should be noted that the medical insurance cost control and audit system provided by the embodiment of the present invention and the medical insurance cost control and audit method provided by the embodiment of the present invention are based on the same inventive concept and achieve the same beneficial effects. Therefore, the specific implementation of this embodiment can refer to the implementation and beneficial effects of the aforementioned network resource configuration method, and the repeated parts will not be repeated.

[0070] Figure 5 This is a schematic structural diagram of an electronic device provided by an embodiment of the present application. Please refer to Figure 5 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and a memory. The memory may include a memory, such as a high-speed random access memory (RAM), and may also include a non-volatile memory (non-volatile memory), such as at least one disk storage. Of course, the electronic device may also include hardware required for other services.

[0071] The processor, network interface and memory can be interconnected through an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 5 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0072] The memory is used to store the program. Specifically, the program may include a program code, and the program code includes a computer operation instruction. The memory may include a memory and a non-volatile memory, and provides instructions and data to the processor.

[0073] The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it, forming a data acquisition device at the logical level. The processor executes the program stored in the memory and is specifically used to execute Figures 1 to 3 The method provided by the embodiment.

[0074] The electronic device provided by the embodiment of the present invention should be able to implement each process in the embodiment corresponding to the above message processing method, and will not be described again here to avoid repetition.

[0075] It should be noted that the electronic device provided in the embodiment of the present invention is applied to the sending end of data transmission, and is based on the same inventive concept as the data transmission method provided in the embodiment of the present invention, and achieves the same beneficial effects. Therefore, the specific implementation of this embodiment can refer to the implementation and beneficial effects of the aforementioned network resource configuration method, and the repeated parts will not be repeated.

[0076] The embodiment of the present application also provides a readable storage medium on which a program or instruction is stored. When the program or instruction is executed by a processor, the above Figures 1 to 3 The various processes of the method embodiments provided in the embodiments can achieve the same technical effect, and to avoid repetition, they will not be described again here.

[0077] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0078] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0079] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0080] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0081] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0082] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0083] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0084] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0085] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0086] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for reviewing medical insurance cost control, characterized in that: include: Obtaining a doctor's order item dictionary of a target hospital, a medical settlement form item dictionary of the target hospital, and a mapping relationship between items in the doctor's order item dictionary and items in the medical settlement form item dictionary; Determine a target medical knowledge graph according to the medical order item dictionary, the medical settlement form item dictionary, the mapping relationship and the public medical knowledge graph; According to the target medical knowledge graph, determine whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent; The step of determining, based on the target medical knowledge graph, whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent includes: According to the target medical knowledge graph, determine the similarity simF(S, T) between the judgment object T and the search object S; When simF(S, T) is greater than a first preset threshold, it is determined that the judgment object T is consistent with the search object S; When simF(S, T) is less than or equal to the first preset threshold, the judgment object T and the search object S are matched in the upper and lower positions according to the target medical knowledge graph, and the similarity simG(S, T) between the judgment object T and the search object S is determined according to the result of the upper and lower position matching. When simG(S, T) is greater than the second preset threshold, it is determined that the judgment object T is consistent with the search object S; The determining of the similarity simF(S, T) between the judgment object T and the search object S comprises: Determine simF(S, T) using the following formula, simF(S,T)=f(1,S,T)*f(2,S,T)*h(S,T), where f(n,S,T)=count(Set(n,S)∩Set(n,T)) / count(Set(n,S)∪Set(n,T)), Set(n, S) represents the set of all nodes in the target medical knowledge graph whose shortest distance to the search object S is n, Set(n, T) represents the set of all nodes in the target medical knowledge graph whose shortest distance to the judgment object T is n, count() represents the statistical number, and n is a positive integer; h(S,T)=∑e -dis(propriety[i]，S，T) , propriety[i] represents the i-th item in the medical property set of the judgment object T, and dis(propriety[i], S, T) represents the shortest distance between the judgment object T and the search object S along the i-th medical property in the target medical knowledge graph; According to the target medical knowledge graph, the judgment object T and the search object S are matched in a hierarchical manner, and according to the result of the hierarchical matching, the similarity simG(S, T) between the judgment object T and the search object S is determined, including: Determine simG(S, T) using the following formula, simG(S,T)=α*g(S,T)+(1-α)*h(S,T), where 0<α<1, g(S, T) is used to indicate whether the judgment object T is a hyponym of the search object S in the target knowledge graph. h(S,T)=∑e -dis(propriety[i]，S，T) , propriety[i] represents the i-th item in the medical attribute set of the judgment object T, and dis(propriety[i], S, T) represents the shortest distance between the judgment object T and the search object S along the i-th medical attribute in the target medical knowledge graph.

2. The method according to claim 1, characterized in that: The step of determining whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent according to the target medical knowledge graph further includes: When simG(S, T) is less than or equal to the second preset threshold, the target BERT training model is used to determine whether the determination object T is consistent with the search object S.

3. The method according to claim 2, characterized in that The method further comprises: Selecting N corresponding relationships from the mapping relationships for consistency annotation to obtain a training set; Perform BERT model training on the training set to obtain the target BERT training model.

4. A system for medical insurance cost control and review, characterized in that: include: An acquisition unit, the acquisition unit is used to acquire a doctor's order item dictionary of a target hospital, a medical settlement form item dictionary of the target hospital, and a mapping relationship between each item in the doctor's order item dictionary and each item in the medical settlement form item dictionary; A processing unit, the processing unit is used to determine a target medical knowledge graph according to the medical order item dictionary, the medical settlement form item dictionary, the mapping relationship and the public medical knowledge graph; The processing unit is also used to determine whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent according to the target medical knowledge graph, including: According to the target medical knowledge graph, determine the similarity simF(S, T) between the judgment object T and the search object S, and when simF(S, T) is greater than a first preset threshold, determine that the judgment object T is consistent with the search object S; When simF(S, T) is less than or equal to the first preset threshold, the search object S is matched with the upper and lower positions according to the target medical knowledge graph, and the similarity simG(S, T) between the judgment object T and the search object S is determined according to the result of the upper and lower position matching. When simG(S, T) is greater than the second preset threshold, it is determined that the judgment object T is consistent with the search object S; The step of determining, based on the target medical knowledge graph, whether the judgment object T in the target medical settlement form and the corresponding search object S in the target medical order are consistent includes: According to the target medical knowledge graph, determine the similarity simF(S, T) between the judgment object T and the search object S; When simF(S, T) is greater than a first preset threshold, it is determined that the judgment object T is consistent with the search object S; When simF(S, T) is less than or equal to the first preset threshold, the judgment object T and the search object S are matched in the upper and lower positions according to the target medical knowledge graph, and the similarity simG(S, T) between the judgment object T and the search object S is determined according to the result of the upper and lower position matching. When simG(S, T) is greater than the second preset threshold, it is determined that the judgment object T is consistent with the search object S; The determining of the similarity simF(S, T) between the judgment object T and the search object S comprises: Determine simF(S, T) using the following formula, simF(S,T)=f(1,S,T)*f(2,S,T)*h(S,T), where f(n,S,T)=count(Set(n,S)∩Set(n,T)) / count(Set(n,S)∪Set(n,T)), Set(n, S) represents the set of all nodes in the target medical knowledge graph whose shortest distance to the search object S is n, Set(n, T) represents the set of all nodes in the target medical knowledge graph whose shortest distance to the judgment object T is n, count() represents the statistical number, and n is a positive integer; h(S,T)=∑e -dis(propriety[i]，S，T) , propriety[i] represents the i-th item in the medical property set of the judgment object T, and dis(propriety[i], S, T) represents the shortest distance between the judgment object T and the search object S along the i-th medical property in the target medical knowledge graph; According to the target medical knowledge graph, the judgment object T and the search object S are matched in a hierarchical manner, and according to the result of the hierarchical matching, the similarity simG(S, T) between the judgment object T and the search object S is determined, including: Determine simG(S, T) using the following formula, simG(S,T)=α*g(S,T)+(1-α)*h(S,T), where 0<α<1, g(S, T) is used to indicate whether the judgment object T is a hyponym of the search object S in the target knowledge graph. h(S,T)=∑e -dis(propriety[i]，S，T) , propriety[i] represents the i-th item in the medical attribute set of the judgment object T, and dis(propriety[i], S, T) represents the shortest distance between the judgment object T and the search object S along the i-th medical attribute in the target medical knowledge graph.

5. A device for medical insurance cost control and auditing, characterized in that: include: It includes a processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of the medical insurance cost control audit method as described in any one of claims 1 to 3 are implemented.

6. A readable storage medium, characterized in that: The readable storage medium stores programs or instructions, and when the programs or instructions are executed by the processor, the steps of the medical insurance cost control review method as described in any one of claims 1-3 are implemented.

Images