A method and device for generating structured medical data

Abstract

The present disclosure is about a method and device for generating structured medical data, which includes: receiving a medical text to be processed, and segmenting the medical text to be processed to obtain multiple words; identifying multiple second medical named entities from the multiple words in combination with multiple first medical named entities; establishing logical relationships between the multiple second medical named entities based on the logical relationships between the multiple first medical named entities and the natural language entity relationships; and generating structured medical data in combination with the second medical named entities and the logical relationships between the second medical named entities. The method generates structured medical data in combination with medical named entities and the logical relationships between medical named entities, thereby achieving data structuring of massive medical texts, improving processing speed, and improving accuracy.

Description

[0001] Cross-references

[0002] This case is a divisional application of the parent case with application number: 201610862821.5, application date: September 28, 2016, and invention name: A method and device for generating structured medical data. Technical Field

[0003] The present disclosure relates to the technical field of natural language processing of medical texts, and in particular, to a structured medical data generating method and a structured medical data generating device. Background Art

[0004] Medical data mainly includes patients' medical records, doctor's orders, nursing documents, examination findings, examination conclusions, etc. These data reflect the patients' basic information, clinical diagnosis, treatment process and results. With the establishment and improvement of medical system informatization, more and more medical data are converted from manual records to electronic entry. Clinical information such as medical records, doctor's orders, nursing documents, examination reports, etc. is mainly written by medical staff in natural language, and the information structure is relatively complex. How to process, analyze and mine a large amount of this information is an important issue in the construction of medical informatization.

[0005] Medical text structuring is a process of extracting and converting (or encoding) text information. Specifically, it is the automatic conversion of unstructured natural language information into a data structure that computers can "understand" and easily process. The resulting structured data can be used for information retrieval, discovery of medical records, patient information management, in-depth analysis of medical data, etc.

[0006] Traditional methods for structuring medical texts mostly rely on medical practitioners to manually process the text content of pathology reports based on their experience. The process essentially relies on the medical knowledge of medical personnel to manually extract the specimens and the values ​​of their various indicators contained in the pathology text data. However, this manual processing method is not only time-consuming and labor-intensive, but also difficult to guarantee the accuracy. In addition, some researchers have tried to perform structured processing through traditional natural language processing and other means. However, the writing style of medical text information is very different from that of ordinary written texts. There is often no specific subject-predicate or subject-predicate-object structure, which makes it difficult to process through syntactic analysis.

[0007] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention

[0008] The purpose of the present disclosure is to provide a structured medical data generating method and a structured medical data generating device, thereby overcoming one or more problems caused by limitations and defects of related technologies at least to a certain extent.

[0009] According to one aspect of the present disclosure, a method for generating structured medical data is provided, comprising:

[0010] Receiving a medical text to be processed, and performing word segmentation on the medical text to be processed to obtain a plurality of words;

[0011] identifying a plurality of second medical named entities from the plurality of words in combination with the plurality of first medical named entities;

[0012] Establishing a logical relationship between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities and the natural language entity relationship;

[0013] Structured medical data is generated by combining the second medical named entities and the logical relationships between the second medical named entities.

[0014] In an exemplary embodiment of the present disclosure, the medical text to be processed is segmented according to a hidden Markov model.

[0015] In an exemplary embodiment of the present disclosure, identifying a plurality of second medical named entities from the plurality of words comprises:

[0016] performing exact matching on the plurality of words based on the plurality of first medical named entities to identify a first portion of the second medical named entities from the plurality of words; and,

[0017] Fuzzy matching is performed on the multiple words based on a preset rule to identify a second part of the second medical named entities from the multiple words.

[0018] In an exemplary embodiment of the present disclosure, establishing a logical relationship between the plurality of second medical named entities includes:

[0019] Determining whether there may be a logical relationship between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities;

[0020] When determining whether a logical relationship may exist between the plurality of second medical named entities, whether the logical relationship actually exists is confirmed in combination with the natural language entity relationship.

[0021] In an exemplary embodiment of the present disclosure, confirming whether the logical relationship actually exists in combination with the natural language entity relationship includes:

[0022] Whether the logical relationship actually exists is confirmed based on one or more of artificial prior knowledge, data statistics, and a conditional random field (CRF) algorithm.

[0023] According to another aspect of the present disclosure, there is provided a structured medical data generating device, comprising:

[0024] A text receiving module is used to receive the medical text to be processed and segment the medical text to be processed to obtain a plurality of words;

[0025] An entity recognition module: used to identify a plurality of second medical named entities from the plurality of words in combination with the plurality of first medical named entities;

[0026] A relationship identification module: used for establishing a logical relationship between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities and the natural language entity relationship;

[0027] Data generation module: used to generate structured medical data by combining the second medical named entities and the logical relationship between the second medical named entities.

[0028] In an exemplary embodiment of the present disclosure, the medical text to be processed is segmented according to a hidden Markov model.

[0029] In an exemplary embodiment of the present disclosure, identifying a plurality of second medical named entities from the plurality of words comprises:

[0030] performing exact matching on the plurality of words based on the plurality of first medical named entities to identify a first portion of the second medical named entities from the plurality of words; and,

[0031] Fuzzy matching is performed on the multiple words based on a preset rule to identify a second part of the second medical named entities from the multiple words.

[0032] In an exemplary embodiment of the present disclosure, establishing a logical relationship between the plurality of second medical named entities includes:

[0033] Determining whether there may be a logical relationship between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities;

[0034] When determining whether a logical relationship may exist between the plurality of second medical named entities, whether the logical relationship actually exists is confirmed in combination with the natural language entity relationship.

[0035] In an exemplary embodiment of the present disclosure, confirming whether the logical relationship actually exists in combination with the natural language entity relationship includes:

[0036] Whether the logical relationship actually exists is confirmed based on one or more of artificial prior knowledge, data statistics, and a conditional random field (CRF) algorithm.

[0037] The structured medical data generation method and device disclosed in the present invention can automatically generate structured medical data based on medical texts by combining medical named entities and logical relationships between medical named entities. Compared with the prior art, it can achieve data structuring of massive medical texts, improve processing speed, and improve accuracy.

[0038] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the specification are used to explain the principles of the present disclosure. Obviously, the accompanying drawings described below are only some embodiments of the present disclosure, and for ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without creative work.

[0040] Figure 1 The flowchart schematically shows a method for generating structured medical data in an exemplary embodiment of the present disclosure.

[0041] Figure 2 The steps of entity recognition in the exemplary embodiment of the present disclosure are schematically illustrated.

[0042] Figure 3 The steps of relationship identification in the exemplary embodiment of the present disclosure are schematically shown.

[0043] Figure 4 The flowchart schematically shows another method for generating structured medical data in an exemplary embodiment of the present disclosure.

[0044] Figure 5 A block diagram schematically shows a structured medical data generating device in an exemplary embodiment of the present disclosure. DETAILED DESCRIPTION

[0045] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as being limited to the examples set forth herein; on the contrary, these embodiments are provided so that the present disclosure will be more comprehensive and complete, and the concepts of the example embodiments are fully conveyed to those skilled in the art. The described features, structures, or characteristics may be combined in one or more embodiments in any suitable manner. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present disclosure. However, those skilled in the art will appreciate that the technical solutions of the present disclosure may be practiced while omitting one or more of the specific details, or other methods, components, devices, steps, etc. may be adopted. In other cases, known technical solutions are not shown or described in detail to avoid obscuring various aspects of the present disclosure.

[0046] In addition, the accompanying drawings are only schematic illustrations of the present disclosure and are not necessarily drawn to scale. The same reference numerals in the figures represent the same or similar parts, and their repeated description will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0047] In this example implementation, a method for generating structured medical data is first provided. Figure 1 As shown, the structured medical data generation method may include the following steps:

[0048] Step S110. receiving a medical text to be processed, and performing word segmentation on the medical text to be processed to obtain a plurality of words;

[0049] Step S120. Combining the plurality of first medical named entities, identifying a plurality of second medical named entities from the plurality of words;

[0050] Step S130. Establishing a logical relationship between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities and the natural language entity relationship;

[0051] Step S140: Generate structured medical data by combining the second medical named entities and the logical relationships between the second medical named entities.

[0052] The structured medical data generation method in this example embodiment can automatically generate structured medical data based on medical text by combining medical named entities and logical relationships between medical named entities. Compared with the prior art, it can achieve data structuring of massive medical texts, improve processing speed, and improve accuracy.

[0053] Below, each step of the method for generating structured medical data in this example implementation will be further described in detail.

[0054] In step S110, a medical text to be processed is received, and the medical text to be processed is segmented to obtain a plurality of words.

[0055] In this technical field, word segmentation refers to the process of recombining continuous word sequences into word sequences according to certain specifications. For example, in this example implementation, known medical named entities and conventional word frequencies of conventional texts can be combined to perform word segmentation according to a hidden Markov model (HMM). Among them, the hidden Markov model (HMM) is a statistical model that can be used to describe a Markov process containing implicit unknown parameters, and then use these parameters for further analysis. However, it is easy to understand that in other exemplary embodiments of the present disclosure, other methods can also be used for word segmentation, and this exemplary embodiment does not specifically limit this.

[0056] In this example implementation, the above-mentioned known medical named entities can come from a medical knowledge graph. The medical knowledge graph is a medical knowledge database maintained according to actual structural needs. In this example implementation, the medical knowledge graph can include a medical named entity vocabulary and a medical named entity classification relationship logic table, which can be understood as a knowledge set abstracted from actual medical knowledge; the medical named entity vocabulary is composed of medical named entities and corresponding classifications, such as a medical named entity can be fever (classified as a manifestation), and its function is to recall medical named entities in the text; the medical named entity relationship logic table is composed of the relationship between medical named entities, and its function is to recall the potential logical relationship between medical named entities in the text, such as the head (classified as an anatomical part) and fever (classified as a manifestation) There is a logical relationship, etc. In this example implementation, the medical knowledge graph can be generated by medical personnel through a medical terminology dictionary combined with mining actual text.

[0057] In step S120, a plurality of second medical named entities are identified from the plurality of words in combination with the plurality of first medical named entities. Figure 2 As shown, in this example implementation, step S120 may include the following steps S122 to S124. Among them:

[0058] In step S122, the multiple words are accurately matched based on the multiple first medical named entities to identify the first part of the second medical named entities from the multiple words. For example, the results of word segmentation may include: elderly, children, 68 years old, female, none, asthma, blood pressure, blood sugar, cough, lung cancer, diabetes, etc., which can be accurately matched directly according to the words in the medical knowledge graph.

[0059] In step S124, fuzzy matching is performed on the multiple words based on preset rules to identify the second part of the second medical named entity from the multiple words. For example, if the results of word segmentation include: date, drug dosage, etc., then matching can be performed by fuzzy matching. The fuzzy matching method may include: identifying patterns appearing in the text by means of regular expressions. For example, if the word segmentation result of the date December 11, 2010 appears, it can be identified by the regular expression of (\d+year\d+month\d+day), but the present disclosure is not limited to this. In addition, in other exemplary embodiments of the present disclosure, matching can also be performed in other ways according to the situation, and this is not particularly limited in the present exemplary embodiment.

[0060] In step S130, a logical relationship between the plurality of second medical named entities is established based on the logical relationship between the plurality of first medical named entities and the natural language entity relationship. Figure 3 As shown, in this example implementation, step S130 may include the following steps S132 to S134. Among them:

[0061] In step S132, it is determined whether there may be a logical relationship between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities.

[0062] The establishment of the above relationship is mainly established by medical personnel based on medical knowledge, such as whether there may be a logical relationship between the chemotherapy regimen corresponding to the drug and the time when the chemotherapy regimen occurs, but the present disclosure is not limited to this. In addition, in other exemplary embodiments of the present disclosure, it is also possible to judge whether the logical relationship exists in other ways according to the situation, and this exemplary embodiment is not particularly limited to this.

[0063] In step S134, when determining that a logical relationship may exist between the plurality of second medical named entities, whether the logical relationship actually exists is confirmed in combination with the natural language entity relationship.

[0064] For example, in a medical document, the specific text content is: PET-CT reexamination on 2015-12-11 showed no disease progression, and CIK cell immunotherapy was performed on 2016-01-16; among them, the entity 2015-12-11, the entity 2016-01-16 and the entity CIK cell immunotherapy all have potential relationships, but only 2016-01-16 is a real modifier. However, it is easy for those skilled in the art to understand that in other exemplary embodiments of the present disclosure, other methods can also be used to determine whether the logical relationship actually exists, and this exemplary embodiment does not specifically limit this.

[0065] In step S140, structured medical data is generated by combining the second medical named entities and the logical relationships between the second medical named entities.

[0066] In step S130, the result generated is a completely structured result, but actual needs may require a more general data structure, such as: csv format or json format, but the present disclosure is not limited to this, and users can choose according to their needs; the present disclosure also designs different data extraction modules according to different actual needs.

[0067] The structured medical data generation method and device disclosed in the present invention generate structured medical data by combining medical named entities and the logical relationships between medical named entities, thereby achieving data structuring of massive medical texts, improving processing speed and accuracy.

[0068] In other embodiments of the present disclosure, the above-mentioned confirmation of whether the logical relationship actually exists in combination with the natural language entity relationship includes: confirming whether the logical relationship actually exists based on one or more of artificial prior knowledge, data statistics, and conditional random field (CRF) algorithm, but the present disclosure is not limited thereto. In addition, in other exemplary embodiments of the present disclosure, it is also possible to confirm whether the logical relationship actually exists in other ways according to the situation, and this exemplary embodiment does not make any special limitation on this.

[0069] In some embodiments of the present disclosure, the above conditional random field is a typical discriminant model, and its joint probability can be written in the form of a multiplication of several potential functions.

[0070] In other embodiments of the present disclosure, reference Figure 4 As shown, another structured medical data generation method is disclosed, including steps S410 to S440, wherein:

[0071] In step S410, a medical text to be processed is received, and the medical text to be processed is segmented to obtain a plurality of words.

[0072] The above steps are the same as step S110 and thus will not be described in detail.

[0073] In step S420, medical entities in the medical text are recalled through the medical vocabulary list in the medical knowledge graph.

[0074] After word segmentation is completed, the words appearing in the medical named entity vocabulary are recalled according to the classification in the medical named entity vocabulary; for entities that cannot be accurately and completely defined in the vocabulary, they are recalled through fuzzy matching.

[0075] In step S430, the logical relationships between the recalled entities are recalled through the entity rule strategy in the medical vocabulary list in the medical knowledge graph.

[0076] This step includes the following two steps: first, the possible logical relationship between the recalled entities is determined through the logical relationship between the subject classifications in the medical knowledge graph; second, after the possible relationship between the recalled subjects, it is necessary to determine whether the above logical relationship actually exists based on the text semantic relationship.

[0077] In step S440, according to actual needs, feature extraction is performed through entities and the relationships between entities to meet actual retrieval, comparison, analysis and other needs.

[0078] The following are embodiments of the device of the present invention, which can be used to implement the method embodiments of the present invention. For details not disclosed in the device embodiments of the present invention, please refer to the method embodiments of the present invention.

[0079] This example embodiment also provides a structured medical data generation device, which is a device based on medical knowledge graphing to achieve data structuring of massive medical texts. Figure 5 As shown, the structured medical data generating device may include: a text receiving module 510, an entity recognition module 520, a relationship recognition module 530 and a data generating module 540; wherein:

[0080] The text receiving module 510 may be used to receive a medical text to be processed, and perform word segmentation on the medical text to be processed to obtain a plurality of words;

[0081] The entity recognition module 520 may be used to identify a plurality of second medical named entities from the plurality of words in combination with the plurality of first medical named entities;

[0082] The relationship identification module 530 may be used to establish a logical relationship between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities and the natural language entity relationship;

[0083] The data generating module 540 may be configured to generate structured medical data in combination with the second medical named entities and the logical relationships between the second medical named entities.

[0084] In some other embodiments of the present disclosure, the medical text to be processed is segmented according to a hidden Markov model.

[0085] In some other embodiments of the present disclosure, identifying a plurality of second medical named entities from the plurality of words comprises:

[0086] performing exact matching on the plurality of words based on the plurality of first medical named entities to identify a first portion of the second medical named entities from the plurality of words; and,

[0087] Fuzzy matching is performed on the multiple words based on a preset rule to identify a second part of the second medical named entities from the multiple words.

[0088] In some other embodiments of the present disclosure, establishing a logical relationship between the plurality of second medical named entities includes:

[0089] Determining whether there may be a logical relationship between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities;

[0090] When determining whether a logical relationship may exist between the plurality of second medical named entities, whether the logical relationship actually exists is confirmed in combination with the natural language entity relationship.

[0091] In some other embodiments of the present disclosure, confirming whether the logical relationship actually exists in combination with the natural language entity relationship includes:

[0092] Whether the logical relationship actually exists is confirmed based on one or more of artificial prior knowledge, data statistics, and a conditional random field (CRF) algorithm.

[0093] Since the functional modules of the structured medical data generating device of the embodiment of the present disclosure are the same as those in the above-mentioned method invention embodiment, they will not be described again here.

[0094] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. On the contrary, the features and functions of one module or unit described above can be further divided into multiple modules or units to be embodied.

[0095] In addition, although the steps of the method in the present disclosure are described in a specific order in the drawings, this does not require or imply that the steps must be performed in this specific order, or that all the steps shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps, etc.

[0096] Through the description of the above implementation, it is easy for those skilled in the art to understand that the example implementation described here can be implemented by software, or by software combined with necessary hardware. Therefore, the technical solution according to the implementation of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, including several instructions to enable a computing device (which can be a personal computer, a server, a mobile terminal, or a network device, etc.) to execute the method according to the implementation of the present disclosure.

[0097] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any modification, use or adaptation of the present disclosure, which follows the general principles of the present disclosure and includes common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the appended claims.

Claims

1. A method for generating structured medical data, characterized in that: include: Receiving a medical text to be processed, and performing word segmentation on the medical text to be processed to obtain a plurality of words; Combining the plurality of first medical named entities to identify the plurality of second medical named entities from the plurality of words, comprising: performing precise matching on the plurality of words based on the plurality of first medical named entities to identify the first part of the second medical named entities from the plurality of words; and performing fuzzy matching on the plurality of words based on a preset rule to identify the second part of the second medical named entities from the plurality of words; Establishing a logical relationship between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities and the natural language entity relationship, including: judging whether a logical relationship may exist between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities; and when judging whether a logical relationship may exist between the plurality of second medical named entities, confirming whether the logical relationship actually exists in combination with the natural language entity relationship; When it is determined that the logical relationship exists, structured medical data is generated in combination with the second medical named entities and the logical relationship between the second medical named entities.

2. The method for generating structured medical data according to claim 1, characterized in that: The medical text to be processed is segmented according to a hidden Markov model.

3. The method for generating structured medical data according to claim 1, characterized in that: Confirming whether the logical relationship actually exists in combination with the natural language entity relationship includes: Whether the logical relationship actually exists is confirmed based on one or more of artificial prior knowledge, data statistics, and a conditional random field (CRF) algorithm.

4. The method for generating structured medical data according to claim 1, characterized in that: The preset rule includes a regular expression.

5. A structured medical data generating device, characterized in that: include: A text receiving module, used for receiving a medical text to be processed, and performing word segmentation on the medical text to be processed to obtain a plurality of words; An entity recognition module, configured to accurately match the plurality of words based on the plurality of first medical named entities to identify a first part of second medical named entities from the plurality of words; and to fuzzily match the plurality of words based on a preset rule to identify a second part of second medical named entities from the plurality of words; A relationship identification module, for determining whether a logical relationship may exist between the plurality of second medical named entities based on the logical relationship between the plurality of first medical named entities; when determining whether a logical relationship may exist between the plurality of second medical named entities, confirming whether the logical relationship actually exists in combination with a natural language entity relationship; A data generation module is used to generate structured medical data in combination with the second medical named entities and the logical relationship between the second medical named entities when it is determined that the logical relationship exists.

Images