Electronic medical record query method, device, computer equipment and storage medium

Abstract

The present application relates to the field of big data, and provides an electronic medical record query method, device, computer equipment and storage medium. The method includes: obtaining a query condition text and parsing it to obtain a query entity, attribute name and attribute value; querying a target entity having the same normalized entity as the query entity from an entity information table; querying a data record containing the target entity from an attribute information table to obtain a candidate data record set; each candidate data record in the candidate data record set includes an attribute name, an attribute value and a medical record identifier; matching the attribute name and attribute value in the candidate data record with the attribute name and attribute value corresponding to the query entity, and determining the medical record identifier in the successfully matched candidate data record as the target medical record identifier; querying the corresponding target electronic medical record according to the target medical record identifier and sending it to the terminal. The use of this method can save memory resources when querying electronic medical records.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to an electronic medical record query method, device, computer equipment and storage medium. Background Art

[0002] In today's era of big data explosion, electronic medical record query is a common application function in the medical industry, whether in clinical diagnosis or scientific research.

[0003] In traditional technology, electronic medical records generated by hospitals exist in the form of semi-free text. When searching electronic medical records, simple text matching is usually performed based on keywords. When performing text matching, full-text retrieval of the electronic medical records is required. In this way, a large amount of electronic medical record text needs to be loaded into the memory, resulting in a large amount of memory resources being consumed during the query. Summary of the invention

[0004] Based on this, it is necessary to provide an electronic medical record query method, device, computer equipment and storage medium that can save memory resources in response to the above technical problems.

[0005] An electronic medical record query method, the method comprising:

[0006] Receiving the query condition text sent by the terminal, parsing the query condition text, and obtaining the query entity corresponding to the query condition text, the attribute name and attribute value corresponding to the query entity;

[0007] Querying a target entity corresponding to the query entity from a pre-established entity information table; the target entity and the query entity have the same normalized entity;

[0008] Querying the data record containing the target entity from the pre-established attribute information table to obtain a candidate data record set; each candidate data record in the candidate data record set includes an attribute name, an attribute value and a medical record identifier;

[0009] Respectively matching the attribute name and attribute value in each of the candidate data records with the attribute name and attribute value corresponding to the query entity;

[0010] Determine the medical record identifier corresponding to the successfully matched candidate data record as the target medical record identifier;

[0011] The corresponding target electronic medical record is queried according to the target medical record identifier, and the queried target electronic medical record is sent to the terminal.

[0012] In one embodiment, before querying a target entity corresponding to the query entity from a pre-established entity information table, the method further includes:

[0013] Acquire a historical electronic medical record set, perform entity recognition on the historical electronic medical records in the historical electronic medical record set, save the recognized entities into an entity information table, and obtain an initial entity information table;

[0014] Normalizing each entity in the initial entity information table to obtain a normalized entity corresponding to each entity;

[0015] Each entity is associated with its corresponding normalized entity and stored to obtain an established entity information table.

[0016] In one embodiment, the entity identification of the electronic medical records in the historical electronic medical record set includes:

[0017] Obtain a training medical record set, perform entity labeling on each training medical record in the training medical record set, and obtain entity labeling results corresponding to each training medical record;

[0018] Train the entity recognition model based on each training medical record and the corresponding entity labeling results;

[0019] According to the trained entity recognition model, entity recognition is performed on the electronic medical records in the historical electronic medical record set.

[0020] In one embodiment, normalizing each entity in the initial entity information table includes:

[0021] Obtain standard disease classification information forms;

[0022] The similarity between each entity in the initial entity information table and each disease noun in the standard disease classification information table is obtained, and the disease nouns whose similarity with each entity exceeds a preset threshold are respectively determined as normalized entities corresponding to each entity.

[0023] In one of the embodiments, before querying the data record containing the target entity from the pre-established attribute information table, the method further includes:

[0024] Acquire a historical electronic medical record set, and segment the historical electronic medical records in the historical electronic medical record set into sentences to obtain short sentences to be processed;

[0025] Extracting the attribute name to be processed corresponding to the short sentence to be processed and the attribute value corresponding to the attribute name to be processed;

[0026] Extract all entities corresponding to the short sentence to be processed to obtain an entity set;

[0027] When it is determined according to the attribute name to be processed that the corresponding attribute value is a part attribute, the attribute name to be processed and the attribute value corresponding to the attribute name to be processed are respectively associated with each entity in the entity set;

[0028] When it is determined according to the attribute name to be processed that the corresponding attribute value is a non-part attribute, obtaining the position distance between the attribute value corresponding to the attribute name to be processed and each entity in the entity set, and associating the attribute name to be processed, the attribute value corresponding to the attribute name to be processed and the entity with the smallest position distance;

[0029] Each entity in the entity set and its associated attribute name and attribute value are stored in the attribute information table in correspondence.

[0030] In one embodiment, before sending the queried target electronic medical record to the terminal, the method further includes:

[0031] According to the target medical record identifier, query the corresponding patient basic information from the patient basic information table;

[0032] The step of sending the retrieved target electronic medical record to the terminal includes:

[0033] The target electronic medical record and the basic information of the patient are sent to the terminal.

[0034] An electronic medical record query device, comprising:

[0035] A receiving module, used for receiving a query condition text sent by a terminal, parsing the query condition text, and obtaining a query entity corresponding to the query condition text, and an attribute name and attribute value corresponding to the query entity;

[0036] A first query module, used to query a target entity corresponding to the query entity from a pre-established entity information table; the target entity and the query entity have the same normalized entity;

[0037] A second query module is used to query the data record containing the target entity from a pre-established attribute information table to obtain a candidate data record set; each candidate data record in the candidate data record set includes an attribute name, an attribute value and a medical record identifier;

[0038] A matching module, used to match the attribute name and attribute value in each of the candidate data records with the attribute name and attribute value corresponding to the query entity;

[0039] An identification determination module, used to determine the medical record identification corresponding to the successfully matched candidate data record as the target medical record identification;

[0040] The sending module is used to query the corresponding target electronic medical record according to the target medical record identifier, and send the queried target electronic medical record to the terminal.

[0041] In one embodiment, the device also includes: an entity information table establishment module, which is used to obtain a set of historical electronic medical records, perform entity recognition on the historical electronic medical records in the set of historical electronic medical records, save the recognized entities in the entity information table, and obtain an initial entity information table; normalize each entity in the initial entity information table to obtain a normalized entity corresponding to each entity; and associate each entity with its corresponding normalized entity and store it to obtain an established entity information table.

[0042] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the electronic medical record query method described in any of the above embodiments when executing the computer program.

[0043] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the electronic medical record query method described in any of the above embodiments.

[0044] The above-mentioned electronic medical record query method, device, computer equipment and storage medium, after receiving the query condition text sent by the terminal, first query the target entity with the same normalized entity as the query entity contained in the query text from the pre-established entity information table, and then query the candidate data record containing the target entity from the pre-established attribute information table, and match the attribute name and attribute value in each of the candidate data records with the attribute name and attribute value corresponding to the query entity to determine the target medical record identifier, and finally obtain the queried electronic medical record according to the target medical record identifier. Since the entity information table and attribute information table are established in advance, all electronic medical records are structured, so that all medical record data are stored in a relational database. When performing electronic medical record queries, it is not necessary to load all electronic medical records into the memory, which greatly saves memory resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is an application scenario diagram of an electronic medical record query method in an embodiment;

[0046] Figure 2 is a flowchart of an electronic medical record query method in one embodiment;

[0047] Figure 3 A flowchart of the steps of establishing an entity information table in one embodiment;

[0048] Figure 4A flowchart of steps for establishing an attribute information table in one embodiment;

[0049] Figure 5 is a structural block diagram of an electronic medical record query device in one embodiment;

[0050] Figure 6 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0052] The electronic medical record query method provided in this application can be applied to Figure 1 In the application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. After obtaining the query condition text input by the user, the terminal 102 sends the query condition text to the server, and the server parses the query condition text to obtain the query entity corresponding to the query condition text, the attribute name and attribute value corresponding to the query entity, and further queries the target entity with the same normalized entity as the query entity from the pre-established entity information table, and then queries the data record corresponding to the target entity from the pre-established attribute information table to obtain a candidate data record set, each candidate data record in the candidate data record set includes an attribute name, an attribute value and a medical record identifier, and the server matches the attribute name and attribute value in each candidate data record with the attribute name and attribute value corresponding to the query entity, and determines the medical record identifier corresponding to the successfully matched candidate data record as the target medical record identifier, and the corresponding target electronic medical record can be queried according to the target medical record identifier, and the queried target electronic medical record is returned to the terminal.

[0053] The terminal 102 may be, but is not limited to, various personal computers, laptop computers, smart phones, tablet computers, and portable wearable devices, and the server 104 may be implemented as an independent server or a server cluster consisting of multiple servers.

[0054] In one embodiment, Figure 2 As shown, a method for querying electronic medical records is provided, and the method is applied to Figure 1 The server in the example is used to illustrate the following steps:

[0055] Step 202: receiving the query condition text sent by the terminal, parsing the query condition text, and obtaining the query entity corresponding to the query condition text, and the attribute name and attribute value corresponding to the query entity.

[0056] The query condition text refers to the query text entered by the user, for example, "diarrhea for more than two weeks", "hepatitis patients with hypertension for 3 years". Entities refer to medical terms such as diseases, symptoms, examinations and tests, allergy history, etc. The entity obtained by parsing the query condition text is the query entity. The attribute name corresponding to the query entity is used to indicate the attribute type of the query entity, including the size, time, length, etc. of the entity. The attribute value corresponding to the query entity refers to the specific attribute information corresponding to the attribute name, for example, the attribute value corresponding to the time attribute is three days.

[0057] Specifically, the terminal interface can display a query control, such as a query input box. When the query text entered by the user is obtained through the query control and the corresponding user-triggered query operation is received, such as clicking the corresponding query button, the terminal uses the query text entered by the user as the query condition text, and generates a query request carrying the query condition text, and sends it to the server. The server receives the query request, first parses the query request, obtains the query condition text carried therein, further parses the query condition text, extracts the query entity in the query condition text, and the attribute name and attribute value corresponding to the query entity. For example, for the query condition text "diarrhea for more than two weeks", the query entity "diarrhea" is parsed, the attribute name is the time attribute, and the corresponding attribute value is "more than two weeks".

[0058] Step 204: search for a target entity corresponding to the query entity from a pre-established entity information table; the target entity and the query entity have the same normalized entity.

[0059] Among them, the entity information table stores the correspondence between each entity and its corresponding normalized entity. The normalized entity of an entity refers to the entity obtained after normalization of the entity. The normalization here means that entities expressing the same meaning are represented by a unified entity noun. For example, diarrhea, diarrhea, and diarrhea can be uniformly represented by diarrhea, so diarrhea is the normalized entity corresponding to the three entities of diarrhea, diarrhea, and diarrhea.

[0060] In one embodiment, the normalized entity may be a noun in a standard disease classification information table, and the standard disease classification information table may be, for example, ICD-10.

[0061] Specifically, the server can query the entity information table according to the query entity, obtain the normalized entity corresponding to the query entity, and then obtain all entities that have a corresponding relationship with the normalized entity as the target entity. For example, when the query entity is "diarrhea", the corresponding normalized entity queried is "diarrhea", and the entities that have a corresponding relationship with the normalized entity include: "diarrhea" and "diarrhea", then the two entities "diarrhea" and "diarrhea" are entities with the same normalized entity as the query entity "diarrhea", that is, the target entity corresponding to the query entity "diarrhea".

[0062] Step 206, querying the data record containing the target entity from the pre-established attribute information table to obtain a candidate data record set; each candidate data record in the candidate data record set includes an attribute name, an attribute value and a medical record identifier.

[0063] The attribute information table is a data table established based on the correspondence between the entity, the attribute information corresponding to the entity, and the medical record identifier corresponding to the medical record to which the entity belongs. A data record refers to a row of data in the attribute information table. Each data record in the attribute information table includes an entity field, an attribute name field, an attribute value field, and a medical record identifier field.

[0064] Specifically, for each target entity, the server queries the data record containing the target entity from the attribute information table, obtains the data record corresponding to the target entity as a candidate data record, and multiple candidate data records constitute a candidate data record set. It can be understood that each candidate data record in the candidate data record set includes the attribute name, attribute value and medical record identifier corresponding to the target entity.

[0065] It can also be understood that when the query condition text contains multiple query entities, different query entities correspond to different target entities. Then for each query entity, the candidate data records obtained by querying the target entity corresponding to the query entity also correspond to the query entity, that is, different query entities will correspond to different sets of candidate data records.

[0066] Step 208: Match the attribute name and attribute value in each candidate data record with the attribute name and attribute value corresponding to the query entity.

[0067] Step 210: determine the medical record identifier corresponding to the successfully matched candidate data record as the target medical record identifier.

[0068] Specifically, the server matches the attribute name and attribute value in each candidate data record with the attribute name and attribute value corresponding to the query entity. When the attribute name and attribute value in any candidate data record successfully match the attribute name and attribute value corresponding to the query entity, the medical record identifier in the successfully matched candidate data record is determined as the target medical record identifier. The match here means that the attribute name in the candidate record is the same as the attribute name in the query entity, and the attribute value corresponding to the same attribute name in the candidate data record is the same as the attribute value corresponding to the same attribute name in the query entity, or the attribute value corresponding to the same attribute name in the candidate data record satisfies the condition defined by the attribute value corresponding to the same attribute name in the query entity. For example, if the query condition text is diarrhea for more than two weeks, the query entity is diarrhea, the corresponding attribute name is time, and the attribute value is more than two weeks. When matching, if the time attribute value in the candidate data record is more than 14 days, the match is successful.

[0069] It can be understood that when the query condition text contains multiple query entities, since each query entity corresponds to a candidate data record set, when matching attribute names and attribute values, the attribute names and attribute values ​​of the candidate data records in each candidate data record set can be matched with the attribute names and attribute values ​​of the corresponding query entities.

[0070] S212, querying the corresponding target electronic medical record according to the target medical record identifier, and sending the queried target electronic medical record to the terminal.

[0071] Specifically, the medical record identifier can uniquely identify an electronic medical record. Therefore, after obtaining the target medical record identifier, the server can query from the document library storing the electronic medical record according to the target medical record identifier, use the queried electronic medical record as the target electronic medical record, and send it to the terminal, thereby realizing the query of the electronic medical record.

[0072] In the above-mentioned electronic medical record query method, after receiving the query condition text sent by the terminal, the target entity having the same normalized entity as the query entity contained in the query text is first queried from the pre-established entity information table, and then the candidate data record containing the target entity is queried from the pre-established attribute information table, and the attribute name and attribute value in each candidate data record are matched with the attribute name and attribute value corresponding to the query entity to determine the target medical record identifier, and finally the queried electronic medical record can be obtained according to the target medical record identifier. Since the entity information table and attribute information table are established in advance, all electronic medical records are structured, so that all medical record data are stored in a relational database. When performing electronic medical record queries, it is not necessary to load all electronic medical records into the memory, which greatly saves memory resources.

[0073] Furthermore, since the target medical record identifier is obtained based on the target entity, and the target entity is an entity that has a normalized entity with the query entity, the target medical record identifier obtained based on the target entity can cover more electronic medical records, avoid missing certain electronic medical records due to differences in the expression of medical terms, and improve the comprehensiveness of electronic medical record queries.

[0074] In one embodiment, before step 204, that is, searching the pre-established entity information table for the target entity corresponding to the query entity, the electronic medical record query method further includes the step of establishing the entity information table. Figure 3 As shown, the entity information table establishment step specifically includes the following steps 302-306.

[0075] Step 302, obtain a historical electronic medical record set, perform entity recognition on the historical electronic medical records in the historical electronic medical record set, save the recognized entities into an entity information table, and obtain an initial entity information table.

[0076] The historical electronic medical record set refers to a set of historical electronic medical records. The historical electronic medical records here refer to electronic medical records generated before the current time. The historical electronic medical record set can be stored in a server or in a database of other computer devices.

[0077] Specifically, entity recognition refers to the process of identifying all entities contained in historical electronic medical records. After obtaining the historical electronic medical record set, the server performs entity recognition on the historical electronic medical records in the historical electronic medical record set, and saves all recognized entities into the entity information table to obtain an initial entity information table.

[0078] In one embodiment, a pre-trained entity recognition model can be obtained, and entities can be identified from historical electronic medical records through the trained entity recognition model. The entity recognition model here refers to a machine learning model that can be used for entity recognition. The machine learning model here is a mathematical model that has a certain ability by performing machine learning on samples. The full name of machine learning in English is Machine Learning, abbreviated as ML. The machine learning model can adopt a neural network model, a support vector machine, a logistic regression model, a random forest model, or a gradient boosting tree model. The learning method of machine learning can be supervised learning, unsupervised learning, or reinforcement learning. The entity recognition model can be trained and stored locally by a computer device, or it can be obtained by a server from other computer devices through the network.

[0079] Step 304: normalize each entity in the initial entity information table to obtain a normalized entity corresponding to each entity. Step 306: associate each entity with its corresponding normalized entity and store it to obtain a well-established entity information table.

[0080] Specifically, since the entities in the initial entity information table are identified from different historical electronic medical records, and due to differences in medical terminology habits, the same meaning is often expressed differently in different electronic medical records. For example, "diarrhea" can be expressed by multiple different words such as diarrhea and diarrhoea. In this embodiment, after obtaining the initial entity information table, the server can normalize the entities in the entity information table to obtain the normalized entities corresponding to each entity. After obtaining the normalized entities of each entity, the server can associate and store each entity with its corresponding normalized entity to obtain a well-established entity information table.

[0081] It can be understood that since the entity information table is obtained by associating and storing each entity with its corresponding normalized entity, the associative storage of each entity and its corresponding normalized entity will save the association relationship between each entity and its corresponding normalized entity. When querying in the entity information table through a certain entity, its associated normalized entity can be queried. Similarly, when querying in the entity information table through a certain normalized entity, all non-normalized entities corresponding to it can be queried.

[0082] In the above embodiment, each entity is normalized to obtain a normalized entity, and the association relationship between each entity and the normalized entity is stored to obtain an entity information table. Through the entity information table, all entities that have the same meaning as a certain entity can be queried, which can avoid missing certain electronic medical records due to differences in the expression of medical terms, thereby improving the comprehensiveness of electronic medical record queries.

[0083] In one embodiment, entity recognition is performed on electronic medical records in a historical electronic medical record set, including: obtaining a training medical record set, performing entity labeling on each training medical record in the training medical record set, and obtaining entity labeling results corresponding to each training medical record; training an entity recognition model based on each training medical record and the corresponding entity labeling results; and performing entity recognition on the electronic medical records in the historical electronic medical record set based on the trained entity recognition model.

[0084] Specifically, the server can obtain historical electronic medical records as training medical records, and multiple training medical records form a training medical record set. The server further performs entity annotation on the training medical records. Entity annotation is the process of annotating entity nouns from the training medical records. After the entity annotation is completed, the entity annotation results corresponding to each training medical record can be obtained. The entity annotation results include the entity nouns and entity positions contained in the training medical records.

[0085] In one embodiment, when an entity is annotated, it can be annotated in BIO mode, where O represents a useless word, B represents the beginning of an entity, and I represents the middle word of an entity. Specifically, the training medical records can be segmented, and the words obtained by the segmentation can be matched with medical vocabulary in a pre-established medical dictionary, and the words that are successfully matched are determined as entity nouns.

[0086] Furthermore, the server can use each training medical record as an input sample, and use the entity annotation results corresponding to the training medical record as training labels to train the entity recognition model. The training process is the process of continuously adjusting the model parameters. When the training conditions are met, a trained entity recognition model is obtained. In one embodiment, a biLSTM-CRF model can be trained through each training medical record and the corresponding entity annotation results to obtain an entity recognition model, wherein biLSTM refers to bidirectional LSTM; CRF refers to conditional random field. It can be understood that the present application can adopt various training methods provided by the prior art to perform model training to obtain an entity recognition model. The present application does not limit the model training method.

[0087] After the server obtains the trained entity recognition model, when performing entity recognition on electronic medical records in the historical electronic medical record set, the electronic medical records in the historical electronic medical record set can be input into the entity recognition model, thereby identifying the entities in the electronic medical record.

[0088] In this embodiment, the entity recognition model obtained by training medical records and corresponding entity annotation results is used to perform entity recognition on electronic medical records in a historical electronic medical record set, which can improve the efficiency and accuracy of entity recognition.

[0089] In one embodiment, each entity in the initial entity information table is normalized, including: obtaining a standard disease classification information table; obtaining the similarity between each entity in the initial entity information table and each disease noun in the standard disease classification information table, and determining the disease nouns whose similarity with each entity exceeds a preset threshold as normalized entities corresponding to each entity.

[0090] Specifically, for each entity in the initial entity information table, the server calculates the similarity between the entity and each disease noun in the standard disease classification information table. When the similarity between the entity and a disease noun in the standard disease classification information table is greater than a preset threshold, the disease noun is determined as the normalized entity corresponding to the entity.

[0091] In one embodiment, the server may use a variety of similarity calculation methods to calculate the similarity, and perform weighted summation of the similarities calculated by various similarity calculation methods to obtain the final similarity, so that the calculated similarity can be made more accurate, thereby obtaining an accurate normalized entity. For example, when calculating the similarity between an entity in the initial entity information table and a disease noun in the standard disease classification information table, the server may calculate the similarity between the entity and the disease noun based on the synonym dictionary, calculate the similarity between the entity and the disease noun based on the minimum edit distance (Levenshtein Distance), and calculate the similarity between the entity and the disease noun based on the vector space distance, multiply each similarity by the corresponding weight, and then sum them up to obtain the final similarity between the entity and the disease noun.

[0092] In one embodiment, the standard disease classification information table is ICD10.

[0093] In this embodiment, the normalized entities of each entity are obtained by obtaining the similarity between each entity in the initial entity information table and each disease noun in the standard disease classification information table, so that the obtained normalized entities are more universal.

[0094] In one embodiment, before searching for data records containing a target entity from a pre-established attribute information table, the electronic medical record query method further includes an attribute information table establishment step, such as Figure 4 As shown, the steps of establishing the attribute information table include:

[0095] Step 402, obtaining a historical electronic medical record set, segmenting the historical electronic medical records in the historical electronic medical record set into sentences, and obtaining short sentences to be processed.

[0096] Specifically, the server may divide the historical electronic medical records into sentences according to punctuation marks such as periods and exclamation marks used to separate sentences, and obtain short sentences to be processed.

[0097] Step 404: extract the attribute name to be processed corresponding to the short sentence to be processed and the attribute value corresponding to the attribute name to be processed.

[0098] It is understandable that for a historical electronic medical record, after sentence segmentation, multiple short sentences to be processed are usually obtained. Then, for each short sentence to be processed, the server needs to extract the attribute name in the short sentence to be processed. The attribute name extracted from the short sentence to be processed is the attribute name to be processed, and each attribute name to be processed has a corresponding attribute value.

[0099] Specifically, the server can extract attribute names and attribute values ​​from the short sentence to be processed by regular expressions, enumeration, etc. The extracted attribute values ​​are divided into two types: part attributes and non-part attributes, where the part attributes are used to describe the parts of human organs.

[0100] Step 406: extract all entities corresponding to the short sentence to be processed to obtain an entity set.

[0101] In one embodiment, the server may segment the short sentence to be processed, match the words obtained by segmentation with entities in a pre-established medical dictionary, extract the matched words as entities corresponding to the short sentence to be processed, and obtain an entity set.

[0102] Step 408 , judging whether the corresponding attribute value is a part attribute according to the attribute name to be processed, if so, proceeding to step 410 ; if not, proceeding to step 412 .

[0103] Step 410: associate the attribute name to be processed and the attribute value corresponding to the attribute name to be processed with each entity in the entity set.

[0104] Step 412, obtaining the position distance between the attribute value corresponding to the attribute name to be processed and each entity in the entity set, and associating the attribute name to be processed, the attribute value corresponding to the attribute name to be processed and the entity with the smallest position distance.

[0105] For example, the entities extracted from "gastroscopy showed an ulcer about 1.8*1.5cm in size at the gastric fundus" include: examination entity: gastroscopy; symptom entity: ulcer;

[0106] Attributes include: stomach fundus (location), 1.8*1.5 (size),

[0107] Among them, the distance between "1.8*1.5" and "gastroscope" is farther than the distance between "1.8*1.5" and "ulcer", so "1.8*1.5" is associated with "ulcer", and the location attribute is gastric fundus, and it is associated with both the ulcer and gastroscope entities.

[0108] Step 414: Store each entity in the entity set and its associated attribute name and attribute value in the attribute information table in correspondence.

[0109] It can be understood that after associating the attribute name to be processed and the attribute value corresponding to the attribute name to be assigned to a certain entity, the attribute name to be processed and the attribute value corresponding to the attribute name to be assigned to the attribute name to be assigned are used as the attribute information of the entity, so when storing, the entity and the attribute information of the entity can be stored in correspondence. In this way, when searching the attribute information table according to the entity, the data record containing the entity will also contain the attribute information of the entity.

[0110] It can also be understood that, since the entity is extracted from the historical electronic medical record, there is a corresponding relationship between the entity and the historical electronic medical record. Therefore, the medical record identifier of the historical electronic medical record corresponding to the entity can also be stored in the data record corresponding to the entity.

[0111] After the entity, the attribute information corresponding to the entity, and the medical record identifier corresponding to the entity are stored in the same data record, an established attribute information table is obtained.

[0112] In this embodiment, entities, attribute names, and attribute values ​​are extracted from the short sentence to be processed, and it is determined whether the corresponding attribute value is a part attribute based on the attribute name to be processed. The attribute name, attribute value, and entity are correspondingly associated based on the judgment result. Finally, an attribute information table is established based on the association result. Since different associations are performed based on part attributes and non-part attributes, the obtained attribute information table is more accurate.

[0113] In one embodiment, before sending the queried target electronic medical record to the terminal, the electronic medical record query method also includes: querying the corresponding patient basic information from the patient basic information table according to the target medical record identifier; sending the queried target electronic medical record to the terminal, including: sending the target electronic medical record and the patient's basic information to the terminal.

[0114] Specifically, the patient information table stores the correspondence between the medical record identifier and the patient's basic information. Therefore, the server can query the corresponding patient's basic information from the patient's basic information table according to the target medical record identifier. After querying the patient's basic information, the patient's basic information and the target electronic medical record can be sent to the terminal together. The information queried by the terminal is then more comprehensive, thereby better meeting the query requirements.

[0115] It should be understood that although Figure 2-4 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 2-4 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0116] In one embodiment, Figure 5 As shown, an electronic medical record query device 500 is provided, comprising:

[0117] The receiving module 502 is used to receive the query condition text sent by the terminal, parse the query condition text, and obtain the query entity corresponding to the query condition text, the attribute name and attribute value corresponding to the query entity;

[0118] A first query module 504 is used to query a target entity corresponding to the query entity from a pre-established entity information table; the target entity and the query entity have the same normalized entity;

[0119] The second query module 506 is used to query the data records containing the target entity from the pre-established attribute information table to obtain a candidate data record set; each candidate data record in the candidate data record set includes an attribute name, an attribute value and a medical record identifier;

[0120] A matching module 508, for matching the attribute name and attribute value in each candidate data record with the attribute name and attribute value corresponding to the query entity;

[0121] The identifier determination module 510 is used to determine the medical record identifier corresponding to the successfully matched candidate data record as the target medical record identifier.

[0122] The sending module 512 is used to query the corresponding target electronic medical record according to the target medical record identifier, and send the queried target electronic medical record to the terminal.

[0123] In one embodiment, the electronic medical record query device also includes: an entity information table establishment module, which is used to obtain a historical electronic medical record set, perform entity recognition on the historical electronic medical records in the historical electronic medical record set, save the recognized entities in the entity information table, and obtain an initial entity information table; normalize each entity in the initial entity information table respectively to obtain a normalized entity corresponding to each entity; and associate each entity with its corresponding normalized entity and store it to obtain an established entity information table.

[0124] In one embodiment, the entity information table establishment module is also used to obtain a set of training medical records, perform entity annotation on each training medical record in the set of training medical records, and obtain entity annotation results corresponding to each training medical record; train an entity recognition model based on each training medical record and the corresponding entity annotation results; and perform entity recognition on electronic medical records in a set of historical electronic medical records based on the trained entity recognition model.

[0125] In one embodiment, the entity information table establishment module is also used to obtain a standard disease classification information table; obtain the similarity between each entity in the initial entity information table and each disease noun in the standard disease classification information table, and determine the disease nouns whose similarity with each entity exceeds a preset threshold as normalized entities corresponding to each entity.

[0126] In one embodiment, the electronic medical record query device also includes: an attribute information table establishment module, which is used to obtain a historical electronic medical record set, segment the historical electronic medical records in the historical electronic medical record set, and obtain short sentences to be processed; extract the attribute name to be processed corresponding to the short sentence to be processed and the attribute value corresponding to the attribute name to be processed; extract all entities corresponding to the short sentence to be processed to obtain an entity set; when it is determined according to the attribute name to be processed that the corresponding attribute value is a part attribute, associate the attribute name to be processed and the attribute value corresponding to the attribute name to be processed with each entity in the entity set respectively; when it is determined according to the attribute name to be processed that the corresponding attribute value is a non-part attribute, obtain the position distance between the attribute value corresponding to the attribute name to be processed and each entity in the entity set, and associate the attribute name to be processed and the attribute value corresponding to the attribute name to be processed with the entity with the smallest position distance; store each entity in the entity set and its associated attribute name and attribute value in the attribute information table respectively.

[0127] In one embodiment, the electronic medical record query device also includes: an information query module, which is used to query the corresponding patient basic information from the patient basic information table according to the target medical record identifier; the sending module 510 is also used to send the target electronic medical record and the patient basic information to the terminal.

[0128] For the specific definition of the electronic medical record query device, please refer to the definition of the electronic medical record query method above, which will not be repeated here. Each module in the above-mentioned electronic medical record query device can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0129] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 6 As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store data such as entities and attribute information corresponding to the entities. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, an electronic medical record query method is implemented.

[0130] Those skilled in the art will understand that Figure 6 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0131] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the electronic medical record query method described above. The steps of the electronic medical record query method here may be the steps of the electronic medical record query method described in each of the above embodiments.

[0132] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the processor executes the steps of the electronic medical record query method described above. The steps of the electronic medical record query method here can be the steps of the electronic medical record query method in each of the above embodiments.

[0133] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0134] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0135] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A method for querying electronic medical records, the method comprising: Receiving the query condition text sent by the terminal, parsing the query condition text, and obtaining the query entity corresponding to the query condition text, the attribute name and attribute value corresponding to the query entity; Acquire a historical electronic medical record set, perform entity recognition on the historical electronic medical records in the historical electronic medical record set, save the recognized entities into an entity information table, and obtain an initial entity information table; Obtain standard disease classification information forms; Obtaining the similarity between each entity in the initial entity information table and each disease noun in the standard disease classification information table, and determining the disease nouns whose similarity with each entity exceeds a preset threshold as normalized entities corresponding to each entity; Each entity is associated with its corresponding normalized entity and stored to obtain an established entity information table; Querying a target entity corresponding to the query entity from a pre-established entity information table; the target entity and the query entity have the same normalized entity; Querying the data record containing the target entity from the pre-established attribute information table to obtain a candidate data record set; each candidate data record in the candidate data record set includes an attribute name, an attribute value and a medical record identifier; Respectively matching the attribute name and attribute value in each of the candidate data records with the attribute name and attribute value corresponding to the query entity; Determine the medical record identifier corresponding to the successfully matched candidate data record as the target medical record identifier; The corresponding target electronic medical record is queried according to the target medical record identifier, and the queried target electronic medical record is sent to the terminal.

2. The method according to claim 1, characterized in that: The performing entity recognition on the electronic medical records in the historical electronic medical record set includes: Obtain a training medical record set, perform entity labeling on each training medical record in the training medical record set, and obtain entity labeling results corresponding to each training medical record; Train the entity recognition model based on each training medical record and the corresponding entity labeling results; According to the trained entity recognition model, entity recognition is performed on the electronic medical records in the historical electronic medical record set.

3. The method according to claim 1, characterized in that Before querying the data record containing the target entity from the pre-established attribute information table, the method further includes: Acquire a historical electronic medical record set, and segment the historical electronic medical records in the historical electronic medical record set into sentences to obtain short sentences to be processed; Extracting the attribute name to be processed corresponding to the short sentence to be processed and the attribute value corresponding to the attribute name to be processed; Extract all entities corresponding to the short sentence to be processed to obtain an entity set; When it is determined according to the attribute name to be processed that the corresponding attribute value is a part attribute, the attribute name to be processed and the attribute value corresponding to the attribute name to be processed are respectively associated with each entity in the entity set; When it is determined according to the attribute name to be processed that the corresponding attribute value is a non-part attribute, obtaining the position distance between the attribute value corresponding to the attribute name to be processed and each entity in the entity set, and associating the attribute name to be processed, the attribute value corresponding to the attribute name to be processed and the entity with the smallest position distance; Each entity in the entity set and its associated attribute name and attribute value are stored in the attribute information table in correspondence.

4. The method according to claim 1, characterized in that: Before sending the queried target electronic medical record to the terminal, the method further includes: According to the target medical record identifier, query the corresponding patient basic information from the patient basic information table; The step of sending the retrieved target electronic medical record to the terminal includes: The target electronic medical record and the basic information of the patient are sent to the terminal.

5. An electronic medical record query device, characterized in that: The device comprises: A receiving module, used for receiving a query condition text sent by a terminal, parsing the query condition text, and obtaining a query entity corresponding to the query condition text, and an attribute name and attribute value corresponding to the query entity; An entity information table establishment module is used to obtain a historical electronic medical record set, perform entity recognition on the historical electronic medical records in the historical electronic medical record set, save the recognized entities in the entity information table, and obtain an initial entity information table; obtain a standard disease classification information table; obtain the similarity between each entity in the initial entity information table and each disease noun in the standard disease classification information table, and determine the disease nouns whose similarity with each entity exceeds a preset threshold as normalized entities corresponding to each entity; associate each entity with its corresponding normalized entity and store it to obtain an established entity information table; A first query module, used to query a target entity corresponding to the query entity from a pre-established entity information table; the target entity and the query entity have the same normalized entity; A second query module is used to query the data record containing the target entity from a pre-established attribute information table to obtain a candidate data record set; each candidate data record in the candidate data record set includes an attribute name, an attribute value and a medical record identifier; A matching module, used to match the attribute name and attribute value in each of the candidate data records with the attribute name and attribute value corresponding to the query entity; An identification determination module, used to determine the medical record identification corresponding to the successfully matched candidate data record as the target medical record identification; The sending module is used to query the corresponding target electronic medical record according to the target medical record identifier, and send the queried target electronic medical record to the terminal.

6. The device according to claim 5, characterized in that The entity information table establishment module is specifically used to: obtain a training medical record set, perform entity labeling on each training medical record in the training medical record set, and obtain entity labeling results corresponding to each training medical record; train an entity recognition model based on each training medical record and the corresponding entity labeling results; and perform entity recognition on the electronic medical records in the historical electronic medical record set based on the trained entity recognition model.

7. The device according to claim 5, characterized in that It also includes an attribute information table establishment module, which is used to: obtain a historical electronic medical record set, separate the historical electronic medical records in the historical electronic medical record set into sentences to obtain short sentences to be processed; extract the attribute names to be processed corresponding to the short sentences to be processed and the attribute values ​​corresponding to the attribute names to be processed; Extract all entities corresponding to the short sentence to be processed to obtain an entity set; when it is determined according to the attribute name to be processed that the corresponding attribute value is a part attribute, associate the attribute name to be processed and the attribute value corresponding to the attribute name to be processed with each entity in the entity set respectively; when it is determined according to the attribute name to be processed that the corresponding attribute value is a non-part attribute, obtain the position distance between the attribute value corresponding to the attribute name to be processed and each entity in the entity set, and associate the attribute name to be processed, the attribute value corresponding to the attribute name to be processed and the entity with the smallest position distance; store each entity in the entity set and its associated attribute name and attribute value in the attribute information table respectively.

8. The device according to claim 5, characterized in that It also includes an information query module, which is used to query the corresponding patient basic information from the patient basic information table according to the target medical record identifier; the sending module is also used to send the target electronic medical record and the patient basic information to the terminal.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 4 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 4 are implemented.

Images