Similar medical record determination method and device

Abstract

The invention provides a similar medical record determination method and device, and relates to the technical field of medical treatment. When the method is executed, a core treatment item corresponding to each disease is determined firstly, then different core treatment items are classified to obtain historical medical records under each disease phenotype, the different core treatment items correspond to different disease phenotypes, and then the core treatment items are classified to obtain the historical medical records under each disease phenotype. Determining a feature set of each disease phenotype and a weight value corresponding to each feature in the feature set, finally obtaining a target medical record, and determining a similar medical record corresponding to the target medical record according to the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set. Thus, the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set are determined, and the similar medical record corresponding to the target medical record is determined through the features and the weight values corresponding to the features, so that the accuracy of determining the similar medical record can be improved.

Description

Technical Field

[0001] The present application relates to the field of medical technology, and in particular to a method and device for determining similar medical records. Background Art

[0002] The clinical auxiliary diagnosis system includes auxiliary diagnosis, auxiliary diagnosis and treatment, similar medical record recommendation, differential diagnosis recommendation and other functions. Among them, similar medical record recommendation helps auxiliary doctors analyze the treatment plans of similar medical records in the past, so as to better understand the patient's condition and improve the accuracy and efficiency of auxiliary diagnosis and treatment decisions. However, there are many different disease phenotypes under the same disease, and each disease phenotype may correspond to different symptoms, physiological indicators, clinical characteristics, etc. In the prior art, the similar medical record determination method mainly includes the similar medical record determination method based on medical record text, but the medical record data has many types of patient diseases, complex diagnoses, and complex related feature record data, and not all record data are related to the disease treatment phenotype, such as routine examination and test items for admission, etc., which interferes with the determination of similar medical records, resulting in low accuracy in the determination of similar medical records.

[0003] In summary, how to improve the accuracy of determining similar medical records is an urgent problem that technical personnel in this field need to solve. Summary of the invention

[0004] In view of this, the present application provides a method and device for determining similar medical records, aiming to improve the accuracy of determining similar medical records.

[0005] In a first aspect, the present application provides a method for determining similar medical records, comprising:

[0006] Determine the core treatment items corresponding to each disease;

[0007] Classify different core treatment items to obtain historical medical records under each disease phenotype, wherein the different core treatment items correspond to different disease phenotypes;

[0008] Determine a feature set of each disease phenotype and a weight value corresponding to each feature in the feature set;

[0009] Obtain target medical records;

[0010] A similar medical record corresponding to the target medical record is determined according to the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set.

[0011] Optionally, the determining of the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set includes:

[0012] extracting features from medical records of each of the disease phenotypes;

[0013] Calculating the log-likelihood ratio of each feature's presence in the medical records of each disease phenotype;

[0014] According to each feature and the log-likelihood ratio of each feature appearing in the medical records of each disease phenotype, a feature set of each disease phenotype and a weight value corresponding to each feature in the feature set are determined, wherein the feature set includes multiple features, and the weight value corresponding to each feature is the log-likelihood ratio of each feature appearing in the medical records of each disease phenotype.

[0015] Optionally, after determining the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set, the method further includes:

[0016] Using a chi-square test to determine the correlation between each feature and the disease phenotype corresponding to each feature;

[0017] The features in the feature set whose relevance is lower than the relevance threshold are deleted.

[0018] Optionally, before extracting features from the medical records of each disease phenotype, the method further includes:

[0019] The medical record NPL of each disease phenotype is structured.

[0020] Optionally, the determination of the core treatment items corresponding to each disease includes:

[0021] Identify the core repository;

[0022] The core treatment items corresponding to each disease are determined according to the core treatment library.

[0023] Optionally, determining the core repository includes:

[0024] Include the drug classification code ATC, the national essential drug classification, and the generic names of drugs in the drug database of the State Drug Administration into the core disposal library;

[0025] Establishing a mapping relationship between the generic name of the drug and the ICD-10 code;

[0026] Determine the knowledge weight coefficient corresponding to each generic name of the drug according to the source and source ratio of the generic name of the drug;

[0027] Incorporating the names of surgical procedures in the DIP classification catalog into the core repository;

[0028] Establishing a mapping relationship between the surgical operation name and the ICD-10 code;

[0029] Determine the knowledge weight coefficient for each surgical procedure name.

[0030] Optionally, after determining the core repository, the method further includes:

[0031] Through data query, the number of times the core disposal items under each ICD-10 code are used is counted;

[0032] The core disposal items under various ICD-10 codes are optimized according to the number of uses.

[0033] In a second aspect, the present application provides a similar medical record determination device, comprising:

[0034] The first determination module is used to determine the core treatment items corresponding to each disease;

[0035] A classification module is used to classify different core treatment items to obtain historical medical records under various disease phenotypes, wherein different core treatment items correspond to different disease phenotypes;

[0036] A second determination module is used to determine a feature set of each disease phenotype and a weight value corresponding to each feature in the feature set;

[0037] An acquisition module, used to acquire target medical records;

[0038] The third determination module is used to determine the similar medical record corresponding to the target medical record according to the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set.

[0039] Optionally, the second determining module includes:

[0040] An extraction unit, used for extracting features from the medical records of each disease phenotype;

[0041] a calculation unit, configured to calculate a log-likelihood ratio of each feature appearing in the medical records of each disease phenotype;

[0042] The first determination unit is used to determine the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set according to each feature and the log-likelihood ratio of each feature appearing in the medical records of each disease phenotype, wherein the feature set includes multiple features, and the weight value corresponding to each feature is the log-likelihood ratio of each feature appearing in the medical records of each disease phenotype.

[0043] Optionally, the second determining module includes:

[0044] A fourth determination module is used to determine the correlation between each feature and the disease phenotype corresponding to each feature by using a chi-square test;

[0045] The deleting module is used to delete the features in the feature set whose relevance is lower than a relevance threshold.

[0046] Optionally, the second determining module includes:

[0047] A structuring unit is used to structure the medical record NPL of each disease phenotype.

[0048] Optionally, the first determining module includes:

[0049] A second determination unit is used to determine a core repository;

[0050] The third determining unit is used to determine the core treatment items corresponding to each disease according to the core treatment library.

[0051] Optionally, the second determining unit includes:

[0052] The first inclusion subunit is used to include the drug classification code ATC, the national essential drug classification, and the generic names of drugs in the drug database of the State Drug Administration into the core disposal library;

[0053] The first establishing subunit is used to establish a mapping relationship between the generic name of the drug and the ICD-10 code;

[0054] A first determination subunit is used to determine the knowledge weight coefficient corresponding to each of the generic names of the drugs according to the sources and source ratios of the generic names of the drugs;

[0055] The second inclusion subunit is used to include the surgical operation names in the DIP classification catalog into the core disposal library;

[0056] A second establishing subunit is used to establish a mapping relationship between the surgical operation name and the ICD-10 code;

[0057] The second determination subunit is used to determine the knowledge weight coefficient of each surgical operation name.

[0058] Optionally, the second determining unit includes:

[0059] The query statistics subunit is used to count the number of times the core disposal items under each ICD-10 code are used through data query;

[0060] The optimization subunit is used to optimize the core treatment items under various ICD-10 codes according to the number of uses.

[0061] In a third aspect, an embodiment of the present application provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements any similar medical record determination method as described in the first aspect of the present application.

[0062] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which instructions are stored. When the instructions are executed on a terminal device, the terminal device executes any similar medical record determination method as described in the first aspect of the present application.

[0063] The present application provides a method for determining similar medical records. When executing the method, the core treatment items corresponding to each disease are first determined, and then the different core treatment items are classified to obtain historical medical records under each disease phenotype, wherein different core treatment items correspond to different disease phenotypes, and then the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set are determined. Finally, the target medical record is obtained, and the similar medical record corresponding to the target medical record is determined based on the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set. In this way, by determining the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set, and determining the similar medical record corresponding to the target medical record by the feature and its corresponding weight value, the accuracy of determining similar medical records can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the technical solutions in this embodiment or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0065] Figure 1 A flowchart of a method for determining similar medical records provided in an embodiment of the present application;

[0066] Figure 2 A schematic diagram of the structure of a similar medical record determination device provided in an embodiment of the present application;

[0067] Figure 3 A schematic diagram of the structure of a computer device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0068] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. The present application provides a method and device for determining similar medical records, which are used in the field of medical technology. The above is only an example and does not limit the application field of the method and device provided in the present application.

[0069] The clinical auxiliary diagnosis system includes auxiliary diagnosis, auxiliary diagnosis and treatment, similar medical record recommendation, differential diagnosis recommendation and other functions. Among them, similar medical record recommendation helps auxiliary doctors analyze the treatment plans of similar medical records in the past, so as to better understand the patient's condition and improve the accuracy and efficiency of auxiliary diagnosis and treatment decisions. However, there are many different disease phenotypes under the same disease, and each disease phenotype may correspond to different symptoms, physiological indicators, clinical characteristics, etc. In the prior art, the similar medical record determination method mainly includes the similar medical record determination method based on medical record text, but the medical record data has many types of patient diseases, complex diagnoses, and complex related feature record data, and not all record data are related to the disease treatment phenotype, such as routine examination and test items for admission, etc., which interferes with the determination of similar medical records, resulting in low accuracy in the determination of similar medical records.

[0070] After research, the inventor proposed the technical solution of this application. First, the core treatment items corresponding to each disease are determined, and then different core treatment items are classified to obtain historical medical records under each disease phenotype, where different core treatment items correspond to different disease phenotypes. Then, the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set are determined. Finally, the target medical record is obtained, and the similar medical record corresponding to the target medical record is determined based on the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set. In this way, by determining the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set, and determining the similar medical record corresponding to the target medical record by the feature and its corresponding weight value, the accuracy of determining similar medical records can be improved.

[0071] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of the present application. It should be noted that, for ease of description, only the parts related to the relevant invention are shown in the accompanying drawings. In the absence of conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

[0072] See also Figure 1 , Figure 1 A flowchart of a method for determining similar medical records provided in an embodiment of the present application includes:

[0073] S101: Determine the core treatment items corresponding to each disease.

[0074] First, determine the core database. The specific method of determining the core database is as follows: from the drug classification code ATC, the national basic drug classification and the drug database of the State Drug Administration, the vast majority of generic names of drugs currently used in the market are included in the database. The drug code contains a total of 5 levels. As shown below:

[0075] Level 1: The highest level, which classifies drugs according to their main anatomical organizational systems. There are 14 major categories, represented by A to V, for example, A is the digestive system and metabolism, B is blood and blood organs, C is the cardiovascular system, D is skin and subcutaneous tissue, ..., V is auxiliary drugs for various systems.

[0076] Tier 2: Based on Tier 1, drugs are further classified according to their therapeutic properties. Each major category has several therapeutic subcategories.

[0077] Level 3: This level classifies drugs based on their pharmacological properties, such as their mechanism of action, pharmacological effects, etc.

[0078] Level 4: Based on Level 3, the drugs are classified according to their chemical structure and composition.

[0079] Tier 5: This is the lowest level and is used to specifically identify the chemical substance of a drug product.

[0080] For example: The complete ATC code for a certain drug is as follows: The ATC code for systemic anti-infective drugs is J, the ATC code for systemic antibacterial drugs is J01, (β-lactam antibiotics, penicillin) is J01C, the ATC code for broad-spectrum penicillin is J01CA, the ATC code for ampicillin is J01CA01, the ATC code for amoxicillin is J01CA04, etc.

[0081] The surgery and operation section includes currently common surgeries according to ICD-9-CM3 (National Medical Insurance Version 2.0). This code includes four levels of surgery, for example: the first level is digestive system surgery, the second level is appendix surgery, the third level is appendectomy, and the fourth level is laparoscopic appendectomy and other appendectomy.

[0082] Next, the mapping relationship between disease codes and drugs and surgical operations and the knowledge weight coefficient are determined. The mapping relationship between disease codes and drugs is constructed as follows: First, the original medical prescription medication data in the historical medical records is upgraded to the third-level pharmacological action level according to the hierarchical relationship in the core disposal library, and the original diagnosis information in the historical medical records is converted into ICD-10 codes (National Medical Insurance Version 2.0).

[0083] The knowledge weight between disease codes and drugs is constructed as follows: Use the association rule mining algorithm Apriori algorithm to discover frequent association rules between the main diagnosis and drugs. These association rules represent the correlation between the main diagnosis and certain drugs, which can help determine which drugs often appear with a specific main diagnosis. Finally, the mined association rules are subjected to frequency analysis, and the frequency of occurrence of the drug is used as the knowledge weight coefficient between the disease. Those skilled in the art should know that the above-mentioned Apriori algorithm is not used to limit the present application, and other algorithms for counting the frequency of occurrence of frequent association rules between the main diagnosis and drugs are within the scope of protection of the present application, for example, multiple matching counting method, that is, statistics are performed on the association rules between any main diagnosis and the drug at intervals of a period of time, and the number of occurrences of the association rules between the main diagnosis and the drug within a cycle are counted.

[0084] The mapping relationship between disease codes and surgical operations is established with reference to the DIP classification catalog. The method for constructing the knowledge weight coefficient between disease codes and surgical operations is the same as the method for constructing the knowledge weight between disease codes and drugs, and will not be repeated here. The knowledge weight coefficient is used as a decision coefficient to determine whether the item is used as a core treatment item. For example, the mapping relationships between disease codes and drugs, and disease codes and surgical operations established for historical medical records diagnosed with appendicitis are shown in the following table:

[0085] Table 1 Core database example table

[0086]

[0087]

[0088] According to the core database, the generic name of the drug or the name of the surgical operation whose knowledge weight coefficient corresponding to the disease code exceeds the knowledge weight coefficient threshold is determined as the core treatment item corresponding to the disease. The knowledge weight coefficient threshold is set by technicians in this field based on experience, and the specific value depends on the specific situation. For example, if the knowledge weight coefficient threshold is 0.8, (β-lactam antibiotics, penicillin), appendectomy and appendiceal abscess drainage are selected from the above table as core treatment items.

[0089] In some implementations, the number of times the core treatment items under each ICD-10 code are used in the historical data can be counted through data query, and the core treatment items with a number of uses below a number threshold can be deleted to optimize the core treatment items under various ICD-10 codes according to the number of uses. The number threshold is set by technicians in this field based on experience, and the specific value depends on the specific situation. S102: Different core treatment items are classified into core treatment categories.

[0090] Different core treatment categories correspond to different disease phenotypes. For example, the three core treatment items (β-lactam antibiotics, penicillin), appendectomy, and appendiceal abscess drainage are classified into six core treatment categories: (β-lactam antibiotics, penicillin); (β-lactam antibiotics, penicillin) + appendectomy; (β-lactam antibiotics, penicillin) + appendiceal abscess drainage; appendectomy; appendiceal abscess drainage; appendectomy + appendiceal abscess drainage. Then, the number of historical medical records under each category is counted based on the historical medical records. In this process, the categories with fewer historical medical records under the core treatment category are deleted. Finally, the core treatment categories under K35 (appendicitis) can be divided into three categories, namely acute appendicitis (β-lactam antibiotics, penicillin), periappendiceal abscess ((β-lactam antibiotics, penicillin) + appendiceal abscess drainage) and appendiceal perforation ((β-lactam antibiotics, penicillin) + appendectomy).

[0091] S103: Determine a feature set of each disease phenotype and a weight value corresponding to each feature in the feature set.

[0092] Features are extracted from the medical records of each disease phenotype, including but not limited to symptoms, diseases, signs, abnormal test values, etc. in the medical records. First, a disease phenotype is selected as the target disease phenotype from multiple disease phenotypes, and then the conditional probability value of each feature in the target disease phenotype is calculated. Then, the conditional probability value of each feature in other disease phenotypes except the target disease phenotype is calculated, and the weight value of each feature in each disease phenotype is determined according to the difference between the logarithms of the two conditional probabilities (also called the log-likelihood ratio). The specific calculation formula is as follows:

[0093] W ij =log(P(F i =1|D j =1))-log(P(F i =1|D j =0));

[0094] Among them, W ij Disease phenotype D j Medium Feature F i The weight of feature Fi In disease phenotype D j The specificity in P(F i =1|D j =1) is the disease phenotype D j Medium Feature F i The conditional probability value of the disease phenotype D j There is feature F in i The probability of i =1|D j =0) is the disease phenotype D j Other disease phenotypes other than i The conditional probability value is the value of the disease phenotype D j The presence of characteristic F in other disease phenotypes i The probability of i =1 indicates feature F i Existence, D j =1 means the jth disease phenotype exists, D j = 0 represents the presence or absence of the jth disease phenotype, (F i =1|D j =1) indicates disease phenotype D j Medium Feature F i Existence, (F i =1|D j =0) indicates that the disease phenotype D j Other disease phenotypes i exist.

[0095] When the feature F i Only in disease phenotype D j When the feature F appears in the test set with a high probability, its weight value is large, indicating that the feature is a specific indicator of the disease phenotype. i If a feature appears in multiple disease phenotypes with similar frequencies, its weight value in each disease phenotype will be relatively small, which means that the feature contributes less to distinguishing different disease phenotypes.

[0096] In this way, a feature set can be constructed for each disease phenotype, which contains multiple features and their corresponding weight values. For example, acute appendicitis (white blood cell count less than 1.5*10 ^4 weight value 0.25, abdominal pain weight value 0.25, nausea weight value 0.15, vomiting weight value 0.13, fever weight value 0.12, etc.), periappendiceal abscess (white blood cell count higher than 1.5*10 ^4weight value 0.4, persistent abdominal pain weight value 0.25, diarrhea weight value 0.2, fever weight value 0.1 and other multiple features) and appendicitis perforation (abdominal X-ray showed air-fluid level 0.3, board-like abdomen 0.2, abdominal tenderness 0.15, rebound pain 0.15, fever 0.12 and other multiple features).

[0097] In some embodiments, before extracting features from the medical records of each disease phenotype, the medical records NPL of each disease phenotype can be structured. NPL structuring generally refers to the structural processing of unstructured data in natural language processing (Natural Language Processing). Structural processing is the conversion of unstructured or semi-structured natural language data into a structured data format to facilitate better computer understanding and processing. This structured processing generally involves operations such as word segmentation, part-of-speech tagging, named entity recognition, and relationship extraction of the text, thereby extracting key information from the text and storing and representing it in a structured form.

[0098] In some embodiments, the chi-square test is used to determine the correlation between each feature and the disease phenotype corresponding to each feature, and the features in the feature set with a correlation lower than the correlation threshold are deleted to ensure that each feature in the feature set is significantly correlated with the disease phenotype. The correlation threshold is set by those skilled in the art based on experience, and the specific value depends on the specific situation.

[0099] S104: Obtain target medical records.

[0100] Obtain a target medical record, where the target medical record is a medical record to be determined to be similar to the target medical record.

[0101] S105: Determine similar medical records corresponding to the target medical record based on the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set.

[0102] The target medical record is feature extracted to obtain the features of the target medical record, which include but are not limited to symptoms, diseases, signs, abnormal test values, etc. in the medical record. The feature set consisting of the features of the target medical record is weighted and summed with the feature sets of each disease phenotype under the diagnosis to obtain the disease phenotype medical record with the highest similarity, and the disease phenotype medical record is determined as the similar medical record corresponding to the target medical record. For example, the target medical record contains a feature set consisting of multiple features such as white blood cell 1.8*10^4, abdominal pain, and fever. Each feature in the feature set is respectively calculated with the feature sets of the three disease phenotypes under appendicitis. The weight sum of the intersection is calculated as the similarity value, among which the weight sum of acute appendicitis is 0.37, the weight sum of periappendiceal abscess is 0.75, and the weight sum of appendix perforation is 0.27. Therefore, the target medical record is the periappendiceal abscess phenotype under appendicitis.

[0103] In the embodiment of the present application, the core treatment items corresponding to each disease are first determined, and then the different core treatment items are classified to obtain the historical medical records under each disease phenotype, wherein different core treatment items correspond to different disease phenotypes, and then the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set are determined. Finally, the target medical record is obtained, and the similar medical record corresponding to the target medical record is determined according to the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set. In this way, by determining the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set, and determining the similar medical record corresponding to the target medical record by the feature and its corresponding weight value, the accuracy of determining similar medical records can be improved.

[0104] The above are some specific implementations of the method for determining similar medical records provided in the embodiments of the present application. Based on this, the present application also provides a corresponding device. The device provided in the embodiments of the present application will be introduced below from the perspective of functional modularization.

[0105] like Figure 2 As shown, Figure 2 The schematic diagram of the structure of a similar medical record determination device provided in an embodiment of the present application is as follows. The similar medical record determination device 200 includes:

[0106] The first determination module 210 is used to determine the core treatment items corresponding to each disease;

[0107] A classification module 220 is used to classify different core treatment items to obtain historical medical records under various disease phenotypes, wherein different core treatment items correspond to different disease phenotypes;

[0108] A second determination module 230 is used to determine a feature set of each disease phenotype and a weight value corresponding to each feature in the feature set;

[0109] An acquisition module 240 is used to acquire a target medical record;

[0110] The third determination module 250 is used to determine similar medical records corresponding to the target medical record according to the feature sets of each disease phenotype and the weight value corresponding to each feature in the feature set.

[0111] Optionally, the second determining module 230 includes:

[0112] An extraction unit, used for extracting features from the medical records of each disease phenotype;

[0113] a calculation unit, configured to calculate a log-likelihood ratio of each feature appearing in the medical records of each disease phenotype;

[0114] The first determination unit is used to determine the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set according to each feature and the log-likelihood ratio of each feature appearing in the medical records of each disease phenotype, wherein the feature set includes multiple features, and the weight value corresponding to each feature is the log-likelihood ratio of each feature appearing in the medical records of each disease phenotype.

[0115] Optionally, the second determining module 230 includes:

[0116] The fourth determination module is used to determine the correlation between each feature and the disease phenotype corresponding to each feature by using a chi-square test;

[0117] The deleting module is used to delete the features in the feature set whose relevance is lower than a relevance threshold.

[0118] Optionally, the second determining module 230 includes:

[0119] A structuring unit is used to structure the medical record NPL of each disease phenotype.

[0120] Optionally, the first determining module 210 includes:

[0121] A second determination unit is used to determine a core repository;

[0122] The third determining unit is used to determine the core treatment items corresponding to each disease according to the core treatment library.

[0123] Optionally, the second determining unit includes:

[0124] The first inclusion subunit is used to include the drug classification code ATC, the national essential drug classification, and the generic names of drugs in the drug database of the State Drug Administration into the core disposal library;

[0125] The first establishing subunit is used to establish a mapping relationship between the generic name of the drug and the ICD-10 code;

[0126] A first determination subunit is used to determine the knowledge weight coefficient corresponding to each of the generic names of the drugs according to the sources and source ratios of the generic names of the drugs;

[0127] The second inclusion subunit is used to include the surgical operation names in the DIP classification catalog into the core disposal library;

[0128] A second establishing subunit is used to establish a mapping relationship between the surgical operation name and the ICD-10 code;

[0129] The second determination subunit is used to determine the knowledge weight coefficient of each surgical operation name.

[0130] Optionally, the second determining unit includes:

[0131] The query statistics subunit is used to count the number of times the core disposal items under each ICD-10 code are used through data query;

[0132] The optimization subunit is used to optimize the core treatment items under various ICD-10 codes according to the number of uses.

[0133] The embodiments of the present application also provide corresponding devices and computer storage media for implementing the solutions provided by the embodiments of the present application.

[0134] like Figure 3 As shown, the computer device 01 is in the form of a general computing device. The components of the computer device 01 may include but are not limited to: one or more processors or processing units 03, a system memory 08, and a bus 04 connecting different system components (including the system memory 08 and the processing unit 03).

[0135] Bus 04 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor or a local bus using any of a variety of bus structures. For example, these architectures include but are not limited to Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus and Peripheral Component Interconnect (PCI) bus.

[0136] The computer device 01 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the computer device 01, including volatile and non-volatile media, removable and non-removable media.

[0137] The system memory 08 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 09 and / or cache memory 10. The computer device 01 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 11 may be used to read and write non-removable, non-volatile magnetic media ( Figure 3 not shown, usually called a "hard drive"). Although Figure 3Not shown in the figure, a disk drive for reading and writing a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 04 via one or more data medium interfaces. The memory 08 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the various embodiments of the present invention.

[0138] A program / utility 12 having a set (at least one) of program modules 13 may be stored, for example, in the memory 08, such program modules 13 including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination thereof may include an implementation of a network environment. The program modules 13 generally perform the functions and / or methods of the embodiments described herein.

[0139] The computer device 01 may also communicate with one or more external devices 02 (e.g., keyboard, pointing device, display 07, etc.), one or more devices that enable a user to interact with the computer device 01, and / or any device that enables the computer device 01 to communicate with one or more other computing devices (e.g., network card, modem, etc.). Such communication may be performed via input / output (I / O) interface 06. Furthermore, the computer device 01 may also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN) and / or public network, such as the Internet) via network adapter 05. Figure 3 As shown, the network adapter 05 communicates with other modules of the computer device 01 via the bus 04. It should be understood that although Figure 3 Not shown, other hardware and / or software modules may be used in conjunction with computer device 01, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0140] The processor unit 03 executes various functional applications and data processing by running the programs stored in the system memory 08, for example, implementing a similar medical record determination method provided in an embodiment of the present application.

[0141] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article 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, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0142] Through the description of the above implementation methods, it can be known that those skilled in the art can clearly understand that all or part of the steps in the above-mentioned embodiment method can be implemented by means of software plus a general hardware platform. Based on such an understanding, the technical solution of the present application can be embodied in the form of a software product, which can be stored in a storage medium, such as a read-only memory (ROM) / RAM, a magnetic disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network communication device such as a router) to execute the methods described in each embodiment of the present application or some parts of the embodiments.

[0143] Each embodiment in this specification is described in a progressive manner, and the same and similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative work.

[0144] The above description is merely an exemplary embodiment of the present application and is not intended to limit the protection scope of the present application.

Claims

1. A method for determining similar medical records, characterized in that: include: Determine the core treatment items corresponding to each disease; Classify different core treatment items to obtain historical medical records under each disease phenotype, wherein the different core treatment items correspond to different disease phenotypes; Determine a feature set of each disease phenotype and a weight value corresponding to each feature in the feature set; Obtain target medical records; A similar medical record corresponding to the target medical record is determined according to the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set.

2. The method according to claim 1, characterized in that: The step of determining the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set includes: extracting features from medical records of each of the disease phenotypes; Calculating the log-likelihood ratio of each feature's presence in the medical records of each disease phenotype; According to each feature and the log-likelihood ratio of each feature appearing in the medical records of each disease phenotype, a feature set of each disease phenotype and a weight value corresponding to each feature in the feature set are determined, wherein the feature set includes multiple features, and the weight value corresponding to each feature is the log-likelihood ratio of each feature appearing in the medical records of each disease phenotype.

3. The method according to claim 1, characterized in that After determining the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set, the method further includes: Using a chi-square test to determine the correlation between each feature and the disease phenotype corresponding to each feature; The features in the feature set whose relevance is lower than the relevance threshold are deleted.

4. The method according to claim 2, characterized in that: Before extracting features from the medical records of each disease phenotype, the method further includes: The medical record NPL of each disease phenotype is structured.

5. The method according to claim 1, characterized in that The core treatment items corresponding to each disease are determined, including: Identify the core repository; The core treatment items corresponding to each disease are determined according to the core treatment library.

6. The method according to claim 5, characterized in that The determination of the core repository comprises: Include the drug classification code ATC, the national essential drug classification, and the generic names of drugs in the drug database of the State Drug Administration into the core disposal library; Establishing a mapping relationship between the generic name of the drug and the ICD-10 code; Determine the knowledge weight coefficient corresponding to each generic name of the drug according to the source and source ratio of the generic name of the drug; Incorporating the names of surgical procedures in the DIP classification catalog into the core repository; Establishing a mapping relationship between the surgical operation name and the ICD-10 code; Determine the knowledge weight coefficient for each surgical procedure name.

7. The method according to claim 6, characterized in that After determining the core repository, the method further includes: Through data query, the number of times the core disposal items under each ICD-10 code are used is counted; The core disposal items under various ICD-10 codes are optimized according to the number of uses.

8. A method for determining similar medical records, characterized in that: include: The first determination module is used to determine the core treatment items corresponding to each disease; A classification module is used to classify different core treatment items to obtain historical medical records under various disease phenotypes, wherein different core treatment items correspond to different disease phenotypes; A second determination module is used to determine a feature set of each disease phenotype and a weight value corresponding to each feature in the feature set; An acquisition module, used to acquire target medical records; The third determination module is used to determine the similar medical record corresponding to the target medical record according to the feature set of each disease phenotype and the weight value corresponding to each feature in the feature set.

9. A computer device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for determining similar medical records as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed on a terminal device, the terminal device executes the similar medical record determination method according to any one of claims 1 to 7.

Images