Rule extraction method, device and equipment applicable to medical knowledge graph

Abstract

The present invention provides a rule extraction method, device and equipment suitable for medical knowledge graph, which can be applied to the field of medical knowledge graph technology. The method includes traversing the medical knowledge graph, obtaining multiple medical object nodes and multiple neighbor nodes corresponding to each medical object node, the neighbor nodes representing the patient information of the medical object; for the candidate medical object nodes among the multiple medical object nodes, based on the multiple candidate patient information related to the candidate medical object nodes, determining the candidate reasoning rules, the candidate reasoning rules including the candidate condition patient information as the candidate reasoning condition, and the candidate conclusion patient information as the candidate reasoning conclusion; according to the first co-occurrence number of the candidate condition patient information in the patient information of the multiple medical objects, determining the confidence of the candidate reasoning rule; based on the confidence of the candidate reasoning rules corresponding to each of the multiple candidate object nodes, determining the target reasoning rule for providing patient medical reference information.

Description

Technical Field

[0001] The present invention relates to the field of medical knowledge graph technology, and more specifically to a rule extraction method, device and equipment suitable for medical knowledge graphs. Background Art

[0002] As an effective knowledge representation and storage tool in the era of big data, medical knowledge graphs can effectively mine, organize and manage knowledge in large-scale medical data, improve the quality of medical information services, and thus provide more intelligent services for doctors and patients in the medical field. Rule extraction technology for medical knowledge graphs can provide decision support for clinicians by accurately extracting and analyzing the potential rules and associations contained in the data in medical knowledge graphs.

[0003] Existing knowledge graph rule extraction technologies often determine rules by identifying circular relationship chains in knowledge graphs. However, in medical knowledge graphs, each node is often expanded around the patient as the central node, and there is no circular relationship chain. As a result, it is difficult to extract accurate rules that conform to medical common sense using existing knowledge graph rule extraction technologies. Summary of the invention

[0004] In view of the above problems, the present invention provides a rule extraction method, device and equipment suitable for medical knowledge graphs.

[0005] According to a first aspect of the present invention, a rule extraction method applicable to a medical knowledge graph is provided, comprising: traversing the medical knowledge graph to obtain a plurality of medical object nodes and a plurality of neighbor nodes corresponding to each medical object node, the neighbor nodes representing the patient information of the medical object; for a candidate medical object node among the plurality of medical object nodes, based on a plurality of candidate patient information related to the candidate medical object node, determining a candidate reasoning rule, wherein the candidate reasoning rule comprises candidate conditional patient information as a candidate reasoning condition, and candidate conclusion patient information as a candidate reasoning conclusion; determining the confidence of the candidate reasoning rule according to a first co-occurrence number of the candidate conditional patient information in the patient information of the plurality of medical objects; and determining a target reasoning rule for providing patient medical reference information based on the confidence of the candidate reasoning rule corresponding to each of the plurality of candidate medical object nodes.

[0006] According to an embodiment of the present invention, the first co-occurrence number is determined based on the following operations: comparing multiple patient information corresponding to each of the multiple medical object nodes with the candidate condition patient information to obtain a first comparison result; determining the first co-occurrence number based on the number of first target comparison results that represent a match in the first comparison result.

[0007] According to an embodiment of the present invention, the confidence of a candidate reasoning rule is determined based on a first co-occurrence number of candidate conditional patient information in patient information of multiple medical objects, including: determining the confidence of the candidate reasoning rule based on a ratio between the first co-occurrence number and the second co-occurrence number, wherein the second co-occurrence number is determined based on the following operations: comparing multiple patient information related to an associated medical object node with candidate conclusion patient information to obtain a second comparison result, wherein at least one patient information of the associated medical object matches the candidate conditional patient information; determining the second co-occurrence number based on the number of second target comparison results representing a match in the second comparison result.

[0008] According to an embodiment of the present invention, the above method also includes: determining, from the medical objects in the medical knowledge graph, a medical object node to be processed that is associated with both the target condition patient information and the target conclusion patient information of the target reasoning rule; and determining the target reasoning rule as the medical reasoning rule corresponding to the medical object node to be processed.

[0009] According to an embodiment of the present invention, the patient information includes symptom information and treatment information.

[0010] According to an embodiment of the present invention, for a candidate medical object node among multiple medical object nodes, a candidate inference rule is determined based on multiple candidate patient information related to the candidate medical object node, including: sampling multiple neighbor nodes corresponding to the candidate medical object node to determine a preset number of candidate patient information; and determining the candidate inference rule based on the preset number of candidate patient information.

[0011] According to an embodiment of the present invention, candidate inference rules are determined based on a preset number of candidate patient information, including: determining one candidate patient information from the preset number of candidate patient information as candidate conclusion patient information; and determining other candidate patient information from the preset number of candidate patient information except the conclusion patient information as candidate conditional patient information.

[0012] According to an embodiment of the present invention, for a candidate medical object node among multiple medical object nodes, a candidate inference rule is determined based on multiple candidate patient information related to the candidate medical object node, and also includes: when the number of multiple candidate patient information related to the candidate medical object node does not meet a preset number, the multiple medical object nodes are resampled to determine new candidate patient information.

[0013] The second aspect of the present invention provides a rule extraction device suitable for a medical knowledge graph, comprising: an acquisition module, used to traverse the medical knowledge graph, obtain multiple medical object nodes and multiple neighbor nodes corresponding to each medical object node, the neighbor nodes representing the patient information of the medical object; a first determination module, used to determine a candidate reasoning rule for a candidate medical object node among the multiple medical object nodes based on multiple candidate patient information related to the candidate medical object node, wherein the candidate reasoning rule includes candidate condition patient information as a candidate reasoning condition, and candidate conclusion patient information as a candidate reasoning conclusion; a second determination module, used to determine the confidence of the candidate reasoning rule based on a first co-occurrence number of the candidate condition patient information in the patient information of multiple medical objects; and a third determination module, used to determine the target reasoning rule for providing patient medical reference information based on the confidence of the candidate reasoning rules corresponding to each of the multiple candidate medical object nodes.

[0014] A third aspect of the present invention provides an electronic device, comprising: one or more processors; a memory for storing one or more computer programs, wherein the one or more processors execute the one or more computer programs to implement the steps of the above method.

[0015] The fourth aspect of the present invention further provides a computer-readable storage medium on which a computer program or instruction is stored, and the steps of the above method are implemented when the above computer program or instruction is executed by a processor.

[0016] The fifth aspect of the present invention also provides a computer program product, including a computer program or instructions, which implement the steps of the above method when executed by a processor.

[0017] According to an embodiment of the present invention, the medical knowledge graph is traversed to obtain multiple medical object nodes and multiple neighbor nodes corresponding to each medical object node; based on multiple candidate patient information related to the candidate medical object node, a candidate reasoning rule is determined; according to the first co-occurrence number of the candidate conditional patient information in the patient information of multiple medical objects, the confidence of the candidate reasoning rule is determined to determine the target reasoning rule for providing patient medical reference information. Since a technical means of determining the candidate reasoning rule among multiple candidate patient information is adopted to avoid the centralized characteristics of the medical knowledge graph, a ring-structured rule is obtained. In addition, the confidence is determined according to the first co-occurrence number of the candidate conditional patient information in the patient information of multiple medical objects to obtain the target reasoning rule, avoiding the problem that the existing knowledge graph rule extraction technology is difficult to extract accurate rules that conform to medical common sense, and can adapt to a variety of diseases and provide decision support for clinicians. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The above contents and other objects, features and advantages of the present invention will become more apparent through the following description of the embodiments of the present invention with reference to the accompanying drawings, in which:

[0019] Figure 1 A schematic diagram of an application scenario of a rule extraction method, apparatus, device, medium, and program product applicable to a medical knowledge graph according to an embodiment of the present invention is shown;

[0020] Figure 2 A flowchart of a rule extraction method applicable to a medical knowledge graph according to an embodiment of the present invention is schematically shown;

[0021] Figure 3 A schematic diagram schematically shows a rule extraction method applicable to a medical knowledge graph according to an embodiment of the present invention;

[0022] Figure 4 A structural block diagram of a rule extraction device applicable to a medical knowledge graph according to an embodiment of the present invention is schematically shown;

[0023] Figure 5 A block diagram of an electronic device suitable for implementing a rule extraction method applicable to a medical knowledge graph according to an embodiment of the present invention is schematically shown. DETAILED DESCRIPTION

[0024] Below, embodiments of the present invention will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present invention. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of embodiments of the present invention. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of concepts of the present invention.

[0025] The terms used herein are only for describing specific embodiments and are not intended to limit the present invention. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0026] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0027] When using expressions such as "at least one of A, B, and C, etc.", they should generally be interpreted according to the meaning of the expression commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0028] In the technical solution of the present invention, the user information (including but not limited to user personal information, user image information, user device information, such as location information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved are all information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, invention and application of the relevant data comply with relevant laws, regulations and standards, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entrances for users to choose to authorize or refuse.

[0029] In the scenario of using personal information for automated decision-making, the methods, devices, and systems provided by the embodiments of the present invention provide users with corresponding operation portals for users to choose to agree or reject the automated decision-making results; if the user chooses to reject, the expert decision-making process will be entered. The expression "automated decision-making" here refers to the activity of automatically analyzing and evaluating an individual's behavioral habits, interests and hobbies, or economic, health, credit status, etc. through computer programs, and making decisions. The expression "expert decision-making" here refers to the activity of making decisions by people who specialize in a certain field, have specialized experience, knowledge and skills, and have reached a certain level of professionalism.

[0030] An embodiment of the present invention provides a rule extraction method suitable for a medical knowledge graph, which traverses the medical knowledge graph to obtain multiple medical object nodes and multiple neighbor nodes corresponding to each medical object node, wherein the neighbor nodes represent the patient information of the medical object; for a candidate medical object node among the multiple medical object nodes, based on multiple candidate patient information related to the candidate medical object node, a candidate reasoning rule is determined, the candidate reasoning rule includes candidate condition patient information as a candidate reasoning condition, and candidate conclusion patient information as a candidate reasoning conclusion; according to the first co-occurrence number of the candidate condition patient information in the patient information of the multiple medical objects, the confidence of the candidate reasoning rule is determined; based on the confidence of the candidate reasoning rule corresponding to each of the multiple candidate medical object nodes, a target reasoning rule for providing patient medical reference information is determined.

[0031] Figure 1 An application scenario diagram of a rule extraction method, apparatus, device, medium, and program product applicable to a medical knowledge graph according to an embodiment of the present invention is schematically shown.

[0032] like Figure 1 As shown, the application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is used to provide a medium for a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired, wireless communication links, or optical fiber cables, etc.

[0033] The user can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (only for example).

[0034] The first terminal device 101, the second terminal device 102, and the third terminal device 103 may be various electronic devices having display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0035] The server 105 may be a server that provides various services, such as a background management server (only as an example) that provides support for websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103. The background management server may analyze and process the received data such as the medical knowledge graph, and feed back the processing results (such as target inference rules for providing patient medical reference information, etc.) to the terminal device.

[0036] It should be noted that the rule extraction method applicable to the medical knowledge graph provided in the embodiment of the present invention can generally be executed by the server 105. Accordingly, the rule extraction device applicable to the medical knowledge graph provided in the embodiment of the present invention can generally be set in the server 105. The rule extraction method applicable to the medical knowledge graph provided in the embodiment of the present invention can also be executed by a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Correspondingly, the rule extraction device applicable to the medical knowledge graph provided in the embodiment of the present invention can also be set in a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105.

[0037] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0038] Figure 2 A flowchart of a rule extraction method applicable to a medical knowledge graph according to an embodiment of the present invention is schematically shown.

[0039] like Figure 2 As shown, the rule extraction method applicable to the medical knowledge graph of this embodiment includes operations S210 to S240.

[0040] In operation S210, the medical knowledge graph is traversed to obtain a plurality of medical object nodes and a plurality of neighbor nodes corresponding to each medical object node, wherein the neighbor nodes represent patient information of the medical object.

[0041] According to an embodiment of the present invention, the patient information includes symptom information and treatment information.

[0042] According to an embodiment of the present invention, the medical knowledge graph is a structured semantic knowledge base, which supports various applications such as data governance, clinical decision support, and treatment plan recommendations by converting massive information in the medical field into a machine-readable structured form. The medical knowledge graph includes multiple nodes, and there are edge relationships between the nodes. Specifically, the nodes can be entities such as patients, diseases, symptoms, and drugs, and the edge relationships can be the relationships between the entities. For example, the edge relationship between the patient A node and the drug A node can be that patient A takes drug A.

[0043] According to an embodiment of the present invention, the medical object node may be a node representing the above-mentioned patient entity.

[0044] According to an embodiment of the present invention, the multiple neighbor nodes corresponding to each medical object node may be nodes having an edge relationship with the medical object node, representing the above-mentioned disease, symptom, drug and other disease information and treatment information.

[0045] In operation S220, for a candidate medical object node among multiple medical object nodes, a candidate reasoning rule is determined based on multiple candidate patient information related to the candidate medical object node, wherein the candidate reasoning rule includes candidate condition patient information as a candidate reasoning condition, and candidate conclusion patient information as a candidate reasoning conclusion.

[0046] According to an embodiment of the present invention, the candidate medical object node may be sampled from a plurality of medical object nodes.

[0047] According to an embodiment of the present invention, a candidate inference rule is constructed based on part of the candidate patient information among a plurality of candidate patient information related to the candidate medical object node, wherein the candidate inference rule is a ring structure.

[0048] According to an embodiment of the present invention, the ring structure of the candidate inference rule can be to infer the candidate conclusion patient information from the candidate conditional patient information. For example, when there is candidate conditional patient information of taking drug A, it can be inferred that the patient suffers from disease A.

[0049] In operation S230 , the confidence of the candidate inference rule is determined according to the first co-occurrence number of the candidate condition patient information in the patient information of the plurality of medical objects.

[0050] According to an embodiment of the present invention, the first co-occurrence count is obtained by counting the number of times the candidate condition patient information appears in the patient information of multiple medical objects.

[0051] According to an embodiment of the present invention, the confidence of the candidate inference rule is determined based on the first co-occurrence number obtained above, so as to reflect the association strength between the nodes in the candidate inference rule.

[0052] In operation S240 , a target inference rule for providing patient medical reference information is determined based on the confidences of the candidate inference rules corresponding to each of the plurality of candidate medical object nodes.

[0053] According to an embodiment of the present invention, when the confidence of a candidate inference rule meets a preset condition, the candidate inference rule is used as the target inference rule, wherein the preset condition can be that when the confidence reaches a preset confidence threshold, the candidate inference rule corresponding to the confidence is used as the target inference rule; or the confidences of the candidate inference rules corresponding to multiple candidate medical object nodes are sorted, and a preset number of candidate inference rules are selected as the target inference rules based on the sorting.

[0054] According to the embodiments of the present invention, the target reasoning rules are for reference by doctors or other users with patient treatment qualifications, and cannot be used directly as prescriptions or treatment plans to treat patients. Specifically, it can provide decision support for clinicians, assist patients in providing customized treatment plans, promote new drug discovery and disease mechanism research, and promote the realization of smart medical care.

[0055] According to an embodiment of the present invention, the medical knowledge graph is traversed to obtain multiple medical object nodes and multiple neighbor nodes corresponding to each medical object node; based on multiple candidate patient information related to the candidate medical object node, a candidate reasoning rule is determined; according to the first co-occurrence number of the candidate conditional patient information in the patient information of multiple medical objects, the confidence of the candidate reasoning rule is determined to determine the target reasoning rule for providing patient medical reference information. Since a technical means of determining the candidate reasoning rule among multiple candidate patient information is adopted to avoid the centralized characteristics of the medical knowledge graph, a ring-structured rule is obtained. In addition, the confidence is determined according to the first co-occurrence number of the candidate conditional patient information in the patient information of multiple medical objects to obtain the target reasoning rule, avoiding the problem that the existing knowledge graph rule extraction technology is difficult to extract accurate rules that conform to medical common sense, and can adapt to a variety of diseases and provide decision support for clinicians.

[0056] According to an embodiment of the present invention, the first co-occurrence number is determined based on the following operations: comparing multiple patient information corresponding to each of the multiple medical object nodes with the candidate condition patient information to obtain a first comparison result; determining the first co-occurrence number based on the number of first target comparison results that represent a match in the first comparison result.

[0057] According to an embodiment of the present invention, multiple pieces of patient information corresponding to multiple medical object nodes are traversed, and the candidate condition patient information is compared with the multiple pieces of patient information corresponding to each medical object node to obtain a first comparison result.

[0058] According to an embodiment of the present invention, assume that the initial value of the first co-occurrence number m=0, if the first target comparison result exists in the first comparison result (ie, the patient information matches the candidate condition patient information), then m=m+1.

[0059] For example, the candidate conditional patient information is drug A and symptom A, the patient information corresponding to the medical object node B includes drug A, symptom A, symptom B, etc., the patient information corresponding to the medical object node C includes drug C, symptom C, symptom B, etc., and the patient information corresponding to the medical object node D includes drug A, drug D, symptom A, etc. It can be seen that the patient information corresponding to the medical object nodes B and D matches the candidate conditional patient information, then the first co-occurrence number m corresponding to the candidate conditional patient information is 2.

[0060] According to an embodiment of the present invention, if the first co-occurrence number m=0 obtained based on the first target comparison result, the candidate medical object node is re-determined.

[0061] According to an embodiment of the present invention, the relevance of the candidate conditional patient information to the medical object node is extracted by determining the first co-occurrence number of the candidate conditional patient information in the plurality of patient information corresponding to the plurality of medical object nodes.

[0062] According to an embodiment of the present invention, the confidence of a candidate reasoning rule is determined based on a first co-occurrence number of candidate conditional patient information in patient information of multiple medical objects, including: determining the confidence of the candidate reasoning rule based on a ratio between the first co-occurrence number and the second co-occurrence number, wherein the second co-occurrence number is determined based on the following operations: comparing multiple patient information related to an associated medical object node with candidate conclusion patient information to obtain a second comparison result, wherein at least one patient information of the associated medical object matches the candidate conditional patient information; determining the second co-occurrence number based on the number of second target comparison results representing a match in the second comparison result.

[0063] According to an embodiment of the present invention, if there is at least one piece of patient information matching the candidate conditional patient information among the multiple pieces of patient information corresponding to a medical object node, the medical object node is taken as an associated medical object node.

[0064] According to an embodiment of the present invention, multiple patient information related to the associated medical object node is traversed, and the candidate conclusion patient information is compared with the multiple patient information corresponding to each associated medical object node to obtain a second comparison result.

[0065] According to an embodiment of the present invention, the initial value of the second co-occurrence number n=0, if there is a second target comparison result in the second comparison result (ie, the patient information matches the candidate conclusion patient information), then n=n+1.

[0066] For example, the candidate conclusion patient information is disease A. The patient information corresponding to the associated medical objects (B medical object node and D medical object node) obtained above is traversed. The patient information corresponding to the B medical object node also includes disease A, etc., and the patient information corresponding to the D medical object node also includes disease D, etc. It can be seen that the patient information corresponding to the B medical object node matches the candidate conclusion patient information, then the second co-occurrence number n corresponding to the candidate conclusion patient information is 1.

[0067] According to an embodiment of the present invention, the confidence of the candidate inference rule may be confirmed based on a ratio n / m between the first co-occurrence number and the second co-occurrence number.

[0068] According to an embodiment of the present invention, the confidence of the candidate inference rule is determined based on the ratio between the first co-occurrence number and the second co-occurrence number, so as to obtain the correlation between the candidate condition patient information and the candidate conclusion patient information, accurately mine the potential rules of the data contained in the medical knowledge graph, and judge whether it conforms to medical common sense, so as to obtain the patient-centered inference rules for the medical knowledge graph that meet the requirements.

[0069] According to an embodiment of the present invention, the above method also includes: determining, from the medical objects in the medical knowledge graph, a medical object node to be processed that is associated with both the target condition patient information and the target conclusion patient information of the target reasoning rule; and determining the target reasoning rule as the medical reasoning rule corresponding to the medical object node to be processed.

[0070] According to an embodiment of the present invention, if the confidence of the obtained candidate inference rule meets the preset conditions, the candidate inference rule is used as the target inference rule, the candidate condition patient information of the candidate inference rule is used as the target condition patient information, and the candidate conclusion patient information of the candidate inference rule is used as the target conclusion patient information.

[0071] According to an embodiment of the present invention, target reasoning rules are used to perform medical reasoning on the medical object nodes to be processed, so as to assist clinicians in making medical decisions.

[0072] For example, if the confidence n / m=0.5 of the above candidate inference rule meets the preset conditions, the candidate conditional patient information A drug and A symptom are used as the target conditional patient information, and the candidate conclusion patient information A disease is used as the target conclusion patient information to obtain the target inference rule, that is, when any medical object node to be processed has at least one of the patient information of taking A drug or having A symptom, the patient corresponding to the medical object node to be processed suffers from disease A.

[0073] According to an embodiment of the present invention, for a candidate medical object node among multiple medical object nodes, a candidate inference rule is determined based on multiple candidate patient information related to the candidate medical object node, including: sampling multiple neighbor nodes corresponding to the candidate medical object node to determine a preset number of candidate patient information; and determining the candidate inference rule based on the preset number of candidate patient information.

[0074] According to an embodiment of the present invention, a plurality of neighboring nodes corresponding to a candidate medical object node are sampled, and a preset number of candidate patient information is randomly selected from the plurality of neighboring nodes, wherein the preset number is preferably 3.

[0075] According to an embodiment of the present invention, candidate inference rules are determined based on a preset number of candidate patient information, including: determining one candidate patient information from the preset number of candidate patient information as candidate conclusion patient information; and using other candidate patient information from the preset number of candidate patient information except the candidate conclusion patient information as candidate conditional patient information.

[0076] According to an embodiment of the present invention, any one piece of candidate patient information is randomly selected from a preset number of candidate patient information as candidate conclusion patient information, and the remaining candidate patient information is used as candidate conditional patient information.

[0077] According to an embodiment of the present invention, for a candidate medical object node among multiple medical object nodes, a candidate inference rule is determined based on multiple candidate patient information related to the candidate medical object node, and also includes: when the number of multiple candidate patient information related to the candidate medical object node does not meet a preset number, the multiple medical object nodes are resampled to determine new candidate patient information.

[0078] According to an embodiment of the present invention, if the number of multiple candidate patient information related to the candidate medical object node does not meet the preset number, the candidate inference rule cannot be constructed, and the multiple medical object nodes need to be resampled to extract the candidate patient information that meets the preset number.

[0079] Figure 3 A schematic diagram of a rule extraction method applicable to a medical knowledge graph according to an embodiment of the present invention is schematically shown.

[0080] According to an embodiment of the present invention, Figure 3 As shown, the rule extraction method applicable to the medical knowledge graph first samples multiple neighbor nodes corresponding to the candidate medical object node based on operation S310 to determine the candidate patient information. Then, operation S320 is performed to determine whether the candidate patient information obtained above meets the preset number. If not, operation S310 is performed to determine new candidate patient information. If it is satisfied, operation S330 is performed to determine the candidate reasoning rule based on the preset number of candidate patient information. Operation S340 is performed to compare the multiple patient information corresponding to each of the multiple medical object nodes with the candidate condition patient information to determine the first co-occurrence number. Operation S350 is performed to compare the multiple patient information related to the associated medical object node with the candidate conclusion patient information to obtain the second co-occurrence number. Operation S360 is performed to determine the confidence of the candidate reasoning rule based on the ratio between the first co-occurrence number and the second co-occurrence number. Operation S370 is performed to determine the target reasoning rule based on the confidence of the candidate reasoning rule corresponding to each of the multiple candidate medical object nodes.

[0081] Based on the above rule extraction method applicable to medical knowledge graph, the present invention also provides a rule extraction device applicable to medical knowledge graph. Figure 4 The device is described in detail.

[0082] Figure 4 The structural block diagram of a rule extraction device applicable to a medical knowledge graph according to an embodiment of the present invention is schematically shown.

[0083] like Figure 4 As shown, the rule extraction device 400 applicable to the medical knowledge graph of this embodiment includes an acquisition module 410, a first determination module 420, a second determination module 430 and a third determination module 440.

[0084] The acquisition module 410 is used to traverse the medical knowledge graph to obtain multiple medical object nodes and multiple neighbor nodes corresponding to each medical object node, where the neighbor nodes represent the patient information of the medical object. In one embodiment, the acquisition module 410 can be used to perform the operation S210 described above, which will not be repeated here.

[0085] The first determination module 420 is used to determine a candidate reasoning rule for a candidate medical object node among the multiple medical object nodes based on multiple candidate patient information related to the candidate medical object node, wherein the candidate reasoning rule includes candidate condition patient information as a candidate reasoning condition and candidate conclusion patient information as a candidate reasoning conclusion. In one embodiment, the first determination module 420 can be used to perform the operation S220 described above, which will not be repeated here.

[0086] The second determination module 430 is used to determine the confidence of the candidate inference rule according to the first co-occurrence number of the candidate condition patient information in the patient information of multiple medical objects. In one embodiment, the second determination module 430 can be used to perform the operation S230 described above, which will not be repeated here.

[0087] The third determination module 440 is used to determine the target inference rule for providing patient medical reference information based on the confidence of the candidate inference rules corresponding to the plurality of candidate medical object nodes. In one embodiment, the third determination module 440 can be used to perform the operation S240 described above, which will not be described in detail here.

[0088] According to an embodiment of the present invention, the medical knowledge graph is traversed to obtain multiple medical object nodes and multiple neighbor nodes corresponding to each medical object node; based on multiple candidate patient information related to the candidate medical object node, a candidate reasoning rule is determined; according to the first co-occurrence number of the candidate conditional patient information in the patient information of multiple medical objects, the confidence of the candidate reasoning rule is determined to determine the target reasoning rule for providing patient medical reference information. Since a technical means of determining the candidate reasoning rule among multiple candidate patient information is adopted to avoid the centralized characteristics of the medical knowledge graph, a ring-structured rule is obtained. In addition, the confidence is determined according to the first co-occurrence number of the candidate conditional patient information in the patient information of multiple medical objects to obtain the target reasoning rule, avoiding the problem that the existing knowledge graph rule extraction technology is difficult to extract accurate rules that conform to medical common sense, and can adapt to a variety of diseases and provide decision support for clinicians.

[0089] According to an embodiment of the present invention, any multiple modules among the acquisition module 410, the first determination module 420, the second determination module 430 and the third determination module 440 can be combined into one module for implementation, or any one of the modules can be split into multiple modules. Alternatively, at least part of the functions of one or more of these modules can be combined with at least part of the functions of other modules and implemented in one module. According to an embodiment of the present invention, at least one of the acquisition module 410, the first determination module 420, the second determination module 430 and the third determination module 440 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware and firmware or in a proper combination of any of them. Alternatively, at least one of the acquisition module 410 , the first determination module 420 , the second determination module 430 , and the third determination module 440 may be at least partially implemented as a computer program module, and when the computer program module is executed, a corresponding function may be performed.

[0090] According to an embodiment of the present invention, the rule extraction device 400 applicable to the medical knowledge graph also includes a first comparison module and a fourth determination module.

[0091] The first comparison module is used to compare the multiple patient information corresponding to the multiple medical object nodes with the candidate condition patient information to obtain a first comparison result.

[0092] The fourth determination module is used to determine a first co-occurrence number according to the number of first target comparison results that represent a match in the first comparison result.

[0093] According to an embodiment of the present invention, the second determining module 430 includes a first determining submodule.

[0094] The first determination submodule is used to determine the confidence of the candidate inference rule according to the ratio between the first co-occurrence number and the second co-occurrence number.

[0095] Among them, the rule extraction device 400 applicable to the medical knowledge graph also includes a second comparison module and a fifth determination module.

[0096] The second comparison module is used to compare multiple patient information related to the associated medical object node with the candidate conclusion patient information to obtain a second comparison result, wherein at least one patient information of the associated medical object matches the candidate condition patient information.

[0097] The fifth determination module is used to determine the second co-occurrence number according to the number of second target comparison results that represent a match in the second comparison results.

[0098] According to an embodiment of the present invention, the rule extraction device 400 applicable to the medical knowledge graph also includes a sixth determination module and a seventh determination module.

[0099] The sixth determination module is used to determine, from the medical objects in the medical knowledge graph, the medical object nodes to be processed that are associated with both the target condition patient information and the target conclusion patient information of the target reasoning rule.

[0100] The seventh determination module is used to determine the target reasoning rule as the medical reasoning rule corresponding to the medical object node to be processed.

[0101] According to an embodiment of the present invention, the first determining module 420 includes a second determining submodule and a third determining submodule.

[0102] The second determination submodule is used to sample multiple neighboring nodes corresponding to the candidate medical object node to determine a preset number of candidate patient information.

[0103] The third determination submodule is used to determine candidate inference rules based on a preset number of candidate patient information.

[0104] According to an embodiment of the present invention, the third determining submodule includes a first determining unit and a second determining unit.

[0105] The first determining unit is used to determine one candidate patient information from a preset number of candidate patient information as candidate conclusion patient information.

[0106] The second determining unit is used to use other candidate patient information except the conclusion patient information among the preset number of candidate patient information as candidate conditional patient information.

[0107] According to an embodiment of the present invention, the first determining module 420 further includes a fourth determining submodule.

[0108] The fourth determination submodule is used to resample the multiple medical object nodes to determine new candidate patient information when the number of multiple candidate patient information related to the candidate medical object node does not meet a preset number.

[0109] Figure 5 A block diagram of an electronic device suitable for implementing a rule extraction method applicable to a medical knowledge graph according to an embodiment of the present invention is schematically shown.

[0110] like Figure 5 As shown, the electronic device 500 according to an embodiment of the present invention includes a processor 501, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 502 or a program loaded from a storage part 508 to a random access memory (RAM) 503. The processor 501 may include, for example, a general-purpose microprocessor (such as a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (for example, an application-specific integrated circuit (ASIC)), etc. The processor 501 may also include an onboard memory for caching purposes. The processor 501 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present invention.

[0111] In RAM 503, various programs and data required for the operation of electronic device 500 are stored. Processor 501, ROM 502 and RAM 503 are connected to each other via bus 504. Processor 501 performs various operations of the method flow according to the embodiment of the present invention by executing the program in ROM 502 and / or RAM 503. It should be noted that the program can also be stored in one or more memories other than ROM 502 and RAM 503. Processor 501 can also perform various operations of the method flow according to the embodiment of the present invention by executing the program stored in the one or more memories.

[0112] According to an embodiment of the present invention, the electronic device 500 may further include an input / output (I / O) interface 505, which is also connected to the bus 504. The electronic device 500 may further include one or more of the following components connected to the input / output (I / O) interface 505: an input portion 506 including a keyboard, a mouse, etc.; an output portion 507 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage portion 508 including a hard disk, etc.; and a communication portion 509 including a network interface card such as a LAN card, a modem, etc. The communication portion 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the input / output (I / O) interface 505 as needed. A removable medium 511, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 510 as needed, so that the computer program read therefrom is installed into the storage portion 508 as needed.

[0113] The present invention also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiment; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present invention is implemented.

[0114] According to an embodiment of the present invention, the computer-readable storage medium may be a non-volatile computer-readable storage medium, for example, may include but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, the computer-readable storage medium may be any tangible medium containing or storing a program, which may be used by or in combination with an instruction execution system, an apparatus or a device. For example, according to an embodiment of the present invention, the computer-readable storage medium may include the ROM 502 and / or RAM 503 described above and / or one or more memories other than ROM 502 and RAM 503.

[0115] The embodiment of the present invention also includes a computer program product, which includes a computer program, and the computer program includes a program code for executing the method shown in the flowchart. When the computer program product is run in a computer system, the program code is used to enable the computer system to implement the rule extraction method applicable to the medical knowledge graph provided by the embodiment of the present invention.

[0116] The computer program executes the above functions defined in the system / device of the embodiment of the present invention when it is executed by the processor 501. According to the embodiment of the present invention, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0117] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices, magnetic storage devices, etc. In another embodiment, the computer program may also be transmitted and distributed in the form of signals on a network medium, and downloaded and installed through the communication part 509, and / or installed from the removable medium 511. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0118] In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 509, and / or installed from the removable medium 511. When the computer program is executed by the processor 501, the above functions defined in the system of the embodiment of the present invention are performed. According to the embodiment of the present invention, the system, device, means, module, unit, etc. described above can be implemented by a computer program module.

[0119] According to an embodiment of the present invention, the program code for executing the computer program provided by the embodiment of the present invention can be written in any combination of one or more programming languages. Specifically, these computing programs can be implemented using high-level process and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, Java, C++, python, "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on the remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (e.g., using an Internet service provider to connect through the Internet).

[0120] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of a code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flow chart, and the combination of the boxes in the block diagram or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0121] It will be appreciated by those skilled in the art that the features described in the various embodiments of the present invention may be combined and / or combined in various ways, even if such combinations or combinations are not explicitly described in the present invention. In particular, without departing from the spirit and teachings of the present invention, the features described in the various embodiments of the present invention may be combined and / or combined in various ways. All of these combinations and / or combinations fall within the scope of the present invention.

[0122] The embodiments of the present invention are described above. However, these embodiments are only for the purpose of illustration, and are not intended to limit the scope of the present invention. Although each embodiment is described above, it does not mean that the measures in each embodiment cannot be used in combination advantageously. Without departing from the scope of the present invention, those skilled in the art may make various substitutions and modifications, which should all fall within the scope of the present invention.

Claims

1. A rule extraction method suitable for medical knowledge graph, characterized in that: include: Traversing the medical knowledge graph, obtaining a plurality of medical object nodes and a plurality of neighbor nodes corresponding to each of the medical object nodes, wherein the neighbor nodes represent patient information of the medical object; For a candidate medical object node among the plurality of medical object nodes, determining a candidate reasoning rule based on a plurality of candidate patient information related to the candidate medical object node, wherein the candidate reasoning rule includes candidate condition patient information as a candidate reasoning condition, and candidate conclusion patient information as a candidate reasoning conclusion; Determining the confidence of the candidate inference rule according to a ratio between a first co-occurrence number and a second co-occurrence number of the candidate conditional patient information in the patient information of the plurality of medical objects; and Based on the confidences of the candidate inference rules corresponding to the plurality of candidate medical object nodes, respectively, determining a target inference rule for providing patient medical reference information, wherein the target inference rule has a ring structure; Determine, from the medical objects in the medical knowledge graph, a to-be-processed medical object node associated with both the target condition patient information and the target conclusion patient information of the target inference rule; and Determining the target reasoning rule as the medical reasoning rule corresponding to the medical object node to be processed; Wherein, for a candidate medical object node among the plurality of medical object nodes, determining a candidate inference rule based on a plurality of candidate patient information related to the candidate medical object node comprises: Sampling a plurality of neighboring nodes corresponding to the candidate medical object node to determine a preset number of candidate patient information; Determining the candidate inference rule according to the preset number of candidate patient information; The first co-occurrence number is determined based on the following operations: Comparing a plurality of patient information corresponding to each of the plurality of medical object nodes with the candidate condition patient information to obtain a first comparison result; Determining the first co-occurrence count according to the number of first target comparison results representing a match in the first comparison results; The second co-occurrence number is determined based on the following operations: Comparing a plurality of patient information associated with the associated medical object node with the candidate conclusion patient information to obtain a second comparison result, wherein at least one patient information of the associated medical object matches the candidate condition patient information; The second co-occurrence count is determined according to the number of second target comparison results that represent a match in the second comparison results.

2. The method according to claim 1, characterized in that The patient information includes symptom information and treatment information.

3. The method according to claim 1, characterized in that: The step of determining the candidate inference rule according to the preset number of candidate patient information includes: Determining one piece of candidate patient information from the preset number of candidate patient information as the candidate conclusion patient information; The other candidate patient information among the preset number of candidate patient information except the candidate conclusion patient information is used as the candidate conditional patient information.

4. The method according to claim 1, characterized in that: The determining of candidate inference rules for a candidate medical object node among the plurality of medical object nodes based on a plurality of candidate patient information related to the candidate medical object node also includes: When the number of multiple candidate patient information related to the candidate medical object node does not meet the preset number, the multiple medical object nodes are resampled to determine new candidate patient information.

5. A rule extraction device suitable for medical knowledge graph, characterized in that: The device comprises: An acquisition module, used to traverse the medical knowledge graph, acquire multiple medical object nodes and multiple neighbor nodes corresponding to each of the medical object nodes, wherein the neighbor nodes represent patient information of the medical object; A first determination module is used to determine a candidate reasoning rule for a candidate medical object node among the plurality of medical object nodes based on a plurality of candidate patient information related to the candidate medical object node, wherein the candidate reasoning rule includes candidate condition patient information as a candidate reasoning condition and candidate conclusion patient information as a candidate reasoning conclusion; A second determination module is used to determine the confidence of the candidate inference rule according to the ratio between the first co-occurrence number and the second co-occurrence number of the candidate conditional patient information in the patient information of the plurality of medical objects; and A third determination module is used to determine a target reasoning rule for providing patient medical reference information based on the confidence of the candidate reasoning rules corresponding to each of the plurality of candidate medical object nodes, wherein the target reasoning rule is a ring structure; A sixth determination module, used to determine, from the medical objects in the medical knowledge graph, a to-be-processed medical object node associated with both the target condition patient information and the target conclusion patient information of the target inference rule; a seventh determination module, configured to determine the target reasoning rule as a medical reasoning rule corresponding to the medical object node to be processed; Wherein, the device further comprises: A first comparison module, used for comparing a plurality of patient information corresponding to each of the plurality of medical object nodes with the candidate condition patient information to obtain a first comparison result; A fourth determination module, configured to determine the first co-occurrence count according to the number of first target comparison results representing a match in the first comparison results; The first determining module comprises: A second determination submodule is used to sample a plurality of neighbor nodes corresponding to the candidate medical object node to determine a preset number of candidate patient information; A third determination submodule is used to determine the candidate inference rule according to the preset number of candidate patient information; Wherein, the device further comprises: A second comparison module is used to compare multiple patient information related to the associated medical object node with the candidate conclusion patient information to obtain a second comparison result, wherein at least one patient information of the associated medical object matches the candidate condition patient information; The fifth determination module is used to determine the second co-occurrence number according to the number of second target comparison results that represent a match in the second comparison results.

6. An electronic device comprising: one or more processors; a memory for storing one or more computer programs, It is characterized in that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 1 to 5.

Images