Method and apparatus for natural language processing of medical texts expressed in Chinese

Abstract

A method for processing unstructured medical text expressed in Chinese, the method comprising: using an attention-based named entity recognition (NER) model to identify medical entities in the unstructured medical text expressed in Chinese; using a multidimensional entity understanding framework to structure the identified medical entities; using a medical knowledge graph to normalize the structured medical entities; and outputting the normalized medical entities.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to U.S. Patent Application No. 16 / 395,439, filed in the U.S. Patent and Trademark Office on April 26, 2019, which is incorporated herein by reference in its entirety. Technical Field

[0003] The present disclosure relates to a natural language processing (NLP) framework for processing and understanding medical-related content expressed in Chinese. Background Art

[0004] In recent years, electronic health record (EHR) systems and electronic medical record (EMR) systems have been increasingly adopted in hospitals around the world. EHR systems can collect a wide range of medical data, including structured and unstructured data, text and images. More specifically, most text-based clinical data are still collected and stored in the form of unstructured natural language. Although great efforts have been made in structuring and formalizing medical content, only a small amount of medical content, such as laboratory test results, drug orders, is stored in a structured form. On the contrary, many important medical-related text contents, such as doctors' and nurses' notes, reports, treatment plans, discharge summaries and books, still use "free text" as their representation. These unstructured data and semi-structured data are difficult to use in the development of modern medical artificial intelligence systems, such as clinical decision support systems.

[0005] In addition, understanding medical texts in Chinese may be more difficult than understanding medical texts in English. For example, there are no established standards or guidelines for processing and understanding medical content in Chinese. Second, although there are some existing medical text processing frameworks in English, such as the Unified Medical Language System (UMLS) and the International Statistical Classification of Diseases and Related Health Problems, 10th Revision (ICD-10), these frameworks cannot be directly translated into Chinese because many linguistic elements are significantly different. Summary of the invention

[0006] In one embodiment, a method for processing unstructured medical text expressed in Chinese is provided, the method comprising: using an attention-based named entity recognition (NER) model to identify one or more medical entities in the unstructured medical text expressed in Chinese; structuring the identified medical entities using a multidimensional entity understanding framework; normalizing the structured medical entities using a medical knowledge graph; and outputting the normalized medical entities.

[0007] In one embodiment, a device is provided, comprising: at least one memory configured to store program code; and at least one processor configured to read the program code and operate according to the instructions of the program code, wherein the program code comprises: a recognition code configured to cause the at least one processor to use an attention-based named entity recognition (NER) model to recognize one or more medical entities in an unstructured medical text expressed in Chinese; a structuring code configured to cause the at least one processor to structure the recognized medical entities using a multidimensional entity understanding framework; a normalization code configured to cause the at least one processor to normalize the structured medical entities using a medical knowledge graph; and an output code configured to cause the at least one processor to output the normalized medical entities.

[0008] In one embodiment, a non-transitory computer-readable medium is provided, which stores instructions, including one or more instructions, which, when executed by one or more processors of a device, cause the one or more processors to: use an attention-based named entity recognition (NER) model to identify one or more medical entities in an unstructured medical text expressed in Chinese; structure the identified medical entities using a multidimensional entity understanding framework; normalize the structured medical entities using a medical knowledge graph; and output the normalized medical entities. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 is a diagram of an example of a natural language processing framework according to an embodiment;

[0010] Figure 2 is a diagram of an example environment in which the systems and / or methods described herein may be implemented;

[0011] Figure 3 yes Figure 2 diagram of example components of one or more devices;

[0012] Figure 4 is a diagram of an example of a named entity recognition model according to an embodiment;

[0013] Figure 5 is a diagram of an example of a multidimensional entity understanding framework according to an embodiment;

[0014] Figure 6 is a flow diagram of an example process for implementing a natural language processing framework, according to an embodiment. DETAILED DESCRIPTION

[0015] In the medical field, a large number of documents are based on and use free or unstructured text as their representation. However, the application of artificial intelligence technology in the medical field may require processing, structuring and understanding of medical related entities. The embodiment of the present disclosure relates to a natural language processing (NLP) framework 100 for understanding medical content such as medical text data 104 expressed in Chinese. The NLP framework 100 may include an attention-based deep named entity recognition (NER) model 101, which is used together with a Chinese medical dictionary to identify medical related entities and their categories in unstructured medical text data 104. The multidimensional entity understanding framework 102 can be used to structure the free text content by determining a series of attributes for describing the corresponding core medical entity. In addition, the medical knowledge graph 103 can be used to perform medical entity normalization to output a normalized entity 105. Therefore, the NLP framework 100 can provide a feasible way for processing unstructured medical text content and semi-structured medical text content expressed in Chinese.

[0016] Figure 2 2 is a diagram of an example environment 200 in which the systems and / or methods described herein may be implemented. Figure 2 As shown, environment 200 may include user device 210, platform 220, and network 230. The devices of environment 200 may be interconnected via wired connections, wireless connections, or a combination of wired and wireless connections.

[0017] The user device 210 includes one or more devices capable of receiving, generating, storing, processing, and / or providing information associated with the platform 220. For example, the user device 210 may include a computing device (e.g., a desktop computer, a laptop computer, a tablet computer, a handheld computer, a smart speaker, a server, etc.), a mobile phone (e.g., a smart phone, a wireless phone, etc.), a wearable device (e.g., a pair of smart glasses or a smart watch), or the like. In some implementations, the user device 210 may receive information from the platform 220 and / or send information to the platform 220.

[0018] The platform 220 includes one or more devices capable of implementing the NLP framework 100, as described elsewhere herein. In some implementations, the platform 220 may include a cloud server or a group of cloud servers. In some implementations, the platform 220 may be designed as a modular platform so that certain software components may be swapped in or out according to specific needs. Therefore, the platform 220 may be easily and / or quickly reconfigured for different uses.

[0019] In some implementations, as shown, the platform 220 can be hosted in a cloud computing environment 222. It should be noted that although the implementations described herein describe the platform 220 as being hosted in a cloud computing environment 222, in some implementations, the platform 220 is not cloud-based (i.e., can be implemented outside of a cloud computing environment) or can be partially cloud-based.

[0020] Cloud computing environment 222 includes an environment hosting platform 220. Cloud computing environment 222 can provide computing, software, data access, storage, and other services that do not require end users (e.g., user devices 210) to know the physical location and configuration of the systems and / or devices hosting platform 220. As shown, cloud computing environment 222 may include a set of computing resources 224 (this set of computing resources is collectively referred to as "computing resources 224", and a single computing resource is referred to as "computing resource 224").

[0021] The computing resources 224 include one or more personal computers, workstation computers, server devices, or other types of computing and / or communication devices. In some implementations, the computing resources 224 may control the platform 220. Cloud resources may include computing instances running in the computing resources 224, storage devices provided in the computing resources 224, data transmission devices provided by the computing resources 224, etc. In some implementations, the computing resources 224 may communicate with other computing resources 224 via wired connections, wireless connections, or a combination of wired and wireless connections.

[0022] Further Figure 2 As shown, the computing resources 224 include a set of cloud resources, such as one or more applications ("APP") 224-1, one or more virtual machines ("VM") 224-2, virtualized storage ("VS") 224-3, one or more hypervisors ("HYP") 224-4, etc.

[0023] The applications 224-1 include one or more software applications that can be provided to or accessed by the user device 210 and / or the sensor device 220. The applications 224-1 can eliminate the need to install and run software applications on the user device 210. For example, the applications 224-1 can include software associated with the platform 220 and / or any other software that can be provided through the cloud computing environment 222. In some implementations, one application 224-1 can send information to / receive information from one or more other applications 224-1 through the virtual machine 224-2.

[0024] Virtual machine 224-2 includes a software implementation of a machine (e.g., a computer) that runs a program, similar to a physical machine. Depending on the degree and purpose of the virtual machine 224-2 to any actual machine, virtual machine 224-2 can be a system virtual machine or a process virtual machine. A system virtual machine can provide a complete system platform that supports the operation of a complete operating system ("OS"). A process virtual machine can run a single program and can support a single process. In some implementations, virtual machine 224-2 can run on behalf of a user (e.g., user device 210) and can manage the infrastructure of cloud computing environment 222, such as data management, synchronization, or long-term data transfer.

[0025] Virtualized memory 224-3 includes one or more storage systems and / or one or more devices using virtualization technology within the storage system or device of computing resource 224. In some implementations, in the environment of the storage system, the type of virtualization may include block virtualization and file virtualization. Block virtualization may refer to the abstraction (or separation) of logical storage from physical storage, so that the storage system can be accessed without considering physical storage or heterogeneous structures. Separation allows administrators of storage systems to have flexibility in how administrators manage storage for end users. File virtualization can eliminate the dependency between data accessed at the file level and the location where the file is physically stored. This can optimize storage usage, server consolidation and / or non-interference file migration performance.

[0026] Hypervisor 224-4 may provide hardware virtualization technology that allows multiple operating systems (e.g., "guest operating systems") to run simultaneously on a host computer such as computing resource 224. Hypervisor 224-4 may present a virtual operating platform to the guest operating systems and may manage the operation of the guest operating systems. Multiple instances of each operating system may share virtualized hardware resources.

[0027] The network 230 includes one or more wired networks and / or wireless networks. For example, the network 230 may include a cellular network (e.g., a fifth generation (5G) network, a long term evolution (LTE) network, a third generation (3G) network, a code division multiple access (CDMA) network, etc.), a public land mobile network (PLMN), a local area network (LAN), a wide area network (WAN), a metropolitan area network (MAN), a telephone network (e.g., a public switched telephone network (PSTN)), a private network, an ad hoc network, an intranet, the Internet, a fiber-based network, etc., and / or a combination of these or other types of networks.

[0028] Figure 2 The number and arrangement of devices and networks shown are provided as examples. In practice, there may be additional devices and / or networks, fewer devices and / or networks, different devices and / or networks, or different Figure 2The devices and / or networks shown are arranged differently. Figure 2 Two or more of the devices shown may be implemented in a single device, or Figure 2 The single device shown may be implemented as multiple distributed devices. Additionally or alternatively, one set of devices (eg, one or more devices) of environment 200 may perform one or more functions described as being performed by another set of devices of environment 200.

[0029] Figure 3 300 is a diagram of example components of device 300. Device 300 may correspond to user device 210 and / or platform 220. Figure 3 As shown, device 300 may include a bus 310 , a processor 320 , a memory 330 , a storage component 340 , an input component 350 , an output component 360 , and a communication interface 370 .

[0030] The bus 310 includes components that allow communication between components of the device 300. The processor 320 is implemented in hardware, firmware, or a combination of hardware and software. The processor 320 is a central processing unit (CPU), a graphics processing unit (GPU), an accelerated processing unit (APU), a microprocessor, a microcontroller, a digital signal processor (DSP), a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), or another type of processing component. In some implementations, the processor 320 includes one or more processors that can be programmed to perform functions. The memory 330 includes a random access memory (RAM), a read-only memory (ROM), and / or another type of dynamic or static storage device (e.g., flash memory, magnetic memory, and / or optical memory) that stores information and / or instructions for use by the processor 320.

[0031] Storage component 340 stores information and / or software related to the operation and use of device 300. For example, storage component 340 may include a hard disk (e.g., a magnetic disk, an optical disk, a magneto-optical disk, and / or a solid-state disk), a compact disk (CD), a digital versatile disk (DVD), a floppy disk, a cassette, a magnetic tape, and / or another type of non-transitory computer-readable medium, and a corresponding drive.

[0032] Input components 350 include components that allow device 300 to receive information, such as input through user input (e.g., a touch screen display, a keyboard, a keypad, a mouse, buttons, switches, and / or a microphone). In addition or alternatively, input components 350 may include sensors for sensing information (e.g., a global positioning system (GPS) component, an accelerometer, a gyroscope, and / or an actuator). Output components 360 include components that provide output information from device 300 (e.g., a display, a speaker, and / or one or more light emitting diodes (LEDs)).

[0033] The communication interface 370 includes transceiver-like components (e.g., a transceiver and / or a separate receiver and transmitter) that enable the device 300 to communicate with other devices, such as through a wired connection, a wireless connection, or a combination of a wired connection and a wireless connection. The communication interface 370 may allow the device 300 to receive information from another device and / or provide information to another device. For example, the communication interface 370 may include an Ethernet interface, an optical interface, a coaxial interface, an infrared interface, a radio frequency (RF) interface, a universal serial bus (USB) interface, a Wi-Fi interface, a cellular network interface, etc.

[0034] Device 300 may perform one or more processes described herein. Device 300 may perform these processes in response to processor 320 executing software instructions stored by a non-transitory computer-readable medium such as memory 330 and / or storage component 340. In this document, a computer-readable medium is defined as a non-transitory memory device. A memory device includes memory space within a single physical storage device or memory space distributed across multiple physical storage devices.

[0035] The software instructions may be read into the memory 330 and / or storage component 340 from another computer-readable medium or from another device through the communication interface 370. When executed, the software instructions stored in the memory 330 and / or storage component 340 may cause the processor 320 to perform one or more processes described herein. Additionally or alternatively, hardwired circuitry may be used in place of or in combination with software instructions to perform one or more processes described herein. Therefore, the implementations described herein are not limited to any specific combination of hardware circuitry and software. Figure 3 The number and arrangement of components shown are provided as examples. In practice, device 300 may include additional components, fewer components, different components, or different components. Figure 3 Additionally or alternatively, one set of components (eg, one or more components) of device 300 may perform one or more functions described as being performed by another set of components of device 300.

[0036] Refer again Figure 1 According to an embodiment of the present disclosure, an NLP framework 100 can be used to understand unstructured or semi-structured medical texts expressed in Chinese, such as medical text data 104. For example, the NLP framework 100 can solve three problems of Chinese medical NLP. First, an attention-based deep NER model 101 can be used to extract medical-related entities. Second, a multidimensional entity understanding framework 102 can be used to characterize key features of medical entities in an environment. Third, for entity normalization, a medical knowledge graph 103 can be used to identify potential entity synonyms.

[0037] In an embodiment, the NLP framework 100 can be used to formalize medical content expressed in Chinese. For example, a series of attributes can be defined to fully describe the characteristics of medical entities. In addition, for medical-related NER expressed in Chinese, attention-based character-level and word-level bidirectional LSTM-CRF deep learning models can be used. This model can recognize medical entities outside of vocabulary. In addition, a knowledge graph KG-based system can be used for medical entity normalization.

[0038] In an embodiment, the NLP framework 100 can solve various problems of processing and understanding medical texts expressed in Chinese. The NLP framework 100 can process medical-related free texts expressed in Chinese, including but not limited to doctors' and nurses' notes, reports, treatment plans, discharge summaries, and books. The NLP framework 100 may include a multidimensional medical entity understanding framework 102, which can solve the key features of how to fully characterize medical entity items. The NLP framework 100 may also include a NER model 101, which uses an attention mechanism and a bidirectional long short-term memory conditional random field (LSTM-CRF) NER model. The attention mechanism can improve the accuracy of sequence labeling. The NLP framework 100 may include a knowledge graph 103 for discovering synonyms and normalization of medical entities.

[0039] Figure 4 An example of a NER model 101 according to an embodiment is shown. NER may be a subtask of information extraction that seeks to locate and classify named entities in a text into predefined categories. The NER model 101 may involve medically related named entities, such as diseases, symptoms, surgeries, etc. In order to detect the boundaries and categories of named entities, a BIO tagging system (B: start of entity, I: middle of entity, O: outside entity) may be used together with categories.

[0040] NER Model 101 can be used with the classic LSTM-CRF model with character-level attention. The example structure of NER Model 101 is Figure 4 As shown in . Word-level information and character-level information are used to represent each word. These two levels of embedding can be pre-trained using a skip-gram model with a large unlabeled corpus. For each word, the pre-trained character embedding can be sent to a character-level bidirectional LSTM, and the forward hidden state and the backward hidden state can be concatenated as the character-level output.

[0041] As an alternative to directly concatenating the character-level output and the pre-trained word embedding, the network can allow the NER model 101 to decide how to combine the information of each specific word. The two vectors can be added using a weighted sum and sent to a two-layer fully connected network. A logistic function can be added to the output to make the attention value in the range of [0, 1]. Attention can be used as a weight to combine the character-level output and the pre-trained word embedding. The weighted sum can be used as the final word representation.

[0042] The final word representation can be sent to the word-level bidirectional LSTM, and for each word, the forward hidden state and the backward hidden state can be concatenated together. Then, the shared weight matrix can project each word into a predefined label.

[0043] Figure 5 An example of a multidimensional medical entity understanding framework 102 according to an embodiment is shown. For medical-related entity terms, the multidimensional medical entity understanding framework 102 may use or include one or more parsers and analyzers to extract a complete description of the entity.

[0044] Examples of embodiments of these parsers and analyzers may include Figure 5 One or more elements shown. For example, the positive / negative entity analyzer 501 can identify whether an entity is rejected (negated), especially whether a negative word appears in a Chinese word. The intensity analyzer 502 can identify intensity adjectives, such as a little, very. The causal analyzer 503 can identify what causes a sign, such as a symptom or a disease. The post-condition analyzer 504 can analyze medical results. The pre-condition analyzer 505 can identify some specific conditions of a certain sign, such as suddenness, stimulation. The change pattern analyzer 506 can extract the changes of medical signs over time. The time analyzer 507 can extract the time when the medical sign occurs and the duration of the medical sign. The frequency analyzer 508 can extract frequency-related items, such as three times a day. The body part analyzer 509 can identify one or more body parts associated with the medical sign.

[0045] Another challenge in Chinese medical NLP is the general rules for named entity normalization. For example, medical facts often have dozens or hundreds of formal and informal descriptions and expressions in Chinese, and many medical entities expressed in Chinese are foreign words and phrases, so the interpretation of foreign entity words can be varied.

[0046] In an embodiment, the NLP framework 100 may normalize medical entities using the medical knowledge graph 103. For example, the medical entities may first be extracted using the NER model 101 and a dictionary-based tokenizer, and then the NLP framework 100 may use the multidimensional medical entity understanding framework 102 to obtain the core terms of the medical entities, but without obtaining any adjective description terms. Next, the medical knowledge graph 103 may be used to identify the normalized entities or centroids of the medical entities. The medical knowledge graph 103 may include relationships between medical entity aliases and their centroid terms, and also provide heuristic methods for identifying potential synonymous entities.

[0047] Figure 6 is a flow chart of an example process 600 for implementing the NLP framework 100. In some implementations, the platform 220 may perform Figure 6 In some implementations, the process may be performed by another device or group of devices (e.g., user device 210) that is separate from or includes platform 220. Figure 6 One or more process blocks.

[0048] like Figure 6 As shown, the process 600 may include using the attention-based NER model 101 to identify one or more medical entities in an unstructured medical text represented in Chinese (block 610 ).

[0049] Further Figure 6 As shown, process 600 may include structuring the identified medical entities using the multidimensional entity comprehension framework 102 (block 620 ).

[0050] Further Figure 6 As shown, the process 600 may include normalizing the structured medical entities using the medical knowledge graph 103 (block 630 ).

[0051] Further Figure 6 As shown, process 600 may include outputting a normalized medical entity (block 640).

[0052] In an embodiment, the unstructured medical text in Chinese may include at least one of a doctor's note, a nurse's note, a report, a treatment plan, a discharge summary, or a book.

[0053] In embodiments, a medical entity may include at least one of a disease, a symptom, or a medical procedure.

[0054] In an embodiment, the NER model 101 may be used with a Long Short-Term Memory Conditional Random Field (LSTM-CRF) model to recognize medical entities.

[0055] In an embodiment, each word of a medical entity may be represented by word-level information and character-level information.

[0056] In an embodiment, the recognizing may further include: concatenating the word-level embedding with the character-level embedding using the attention value as a weighted sum.

[0057] In an embodiment, the weighted sum may be sent to a word-level long short-term memory (LSTM), and a shared weight matrix may be used to project each word into one or more predefined tags.

[0058] In an embodiment, the multidimensional entity comprehension framework 102 includes a plurality of analyzers.

[0059] In an embodiment, the plurality of analyzers include at least one of a positive / negative entity analyzer, an intensity analyzer, a cause-effect analyzer, a pre-condition analyzer, a change pattern analyzer, a post-condition analyzer, a time analyzer, a frequency analyzer, and a body part analyzer.

[0060] In an embodiment, the medical knowledge graph 103 may be used to identify one or more synonymous medical entities that are synonymous with a medical entity.

[0061] Although Figure 6 Example blocks of process 600 are shown, but in some implementations, process 600 may include additional blocks, fewer blocks, different blocks, or different blocks. Figure 6 The depicted blocks are blocks that are arranged differently. Additionally or alternatively, two or more blocks of process 600 may be performed in parallel.

[0062] The foregoing disclosure provides illustration and description, but is not intended to be exhaustive or to limit the implementations to the precise forms disclosed. Modifications and variations may be made in light of the foregoing disclosure or may be acquired from practice of the implementations.

[0063] As used herein, the term "component" is intended to be broadly interpreted as hardware, firmware, or a combination of hardware and software.

[0064] Obviously, the systems and / or methods described herein can be implemented in various forms of hardware, firmware, or a combination of hardware and software. The actual dedicated control hardware or software code used to implement these systems and / or methods does not limit the implementation. Therefore, the operation and behavior of the systems and / or methods are described herein without reference to specific software code - it should be understood that software and hardware can be designed to implement the systems and / or methods based on the description herein.

[0065] Even though particular combinations of features are recited in the claims and / or disclosed in the specification, these combinations are not intended to limit the disclosure of possible implementations. In fact, many of these features may be combined in ways not specifically recited in the claims and / or disclosed in the specification. Although each dependent claim listed below may only be directly dependent on one claim, the disclosure of possible implementations includes the combination of each dependent claim with every other claim in the claim set.

[0066] Unless explicitly described per se, the elements, actions or instructions used herein should not be understood as being essential or necessary. In addition, as used herein, the articles "one" and "an" are intended to include one or more items, and can be used interchangeably with "one or more". In addition, as used herein, the term "set" is intended to include one or more items (e.g., related items, unrelated items, a combination of related items and unrelated items, etc.), and can be used interchangeably with "one or more". In the case where the intention is only one item, the term "one" or similar language is used. In addition, as used herein, the term "having", "having", "containing" or similar terms are intended to be open terms. In addition, the phrase "based on" means "based on at least in part", unless otherwise explicitly stated.

Claims

1. A method for processing unstructured medical text expressed in Chinese, the method comprising: Using an attention-based named entity recognition (NER) model to identify medical entities in the unstructured medical text represented in Chinese; The NER model is a character-level and word-level bidirectional long short-term memory conditional random field model based on the attention mechanism; each word of the medical entity is represented by word-level information and character-level information; the word-level information and the character-level information are pre-trained by a skip-graph model of an unlabeled corpus; the NER model includes a character-level bidirectional LSTM, attention, a word-level bidirectional LSTM and a shared weight matrix; for each word, the character-level bidirectional LSTM is used to convert the pre-trained character embedding into a character-level output; the attention is used to obtain the weighted sum of the character-level output and the pre-trained word embedding using the attention value, and concatenate the word embedding with the character embedding; the input of the word-level bidirectional LSTM is the weighted sum; the shared weight matrix is ​​used to project the output of the word-level bidirectional LSTM into one or more predefined labels; The identified medical entities were structured using a multidimensional entity understanding framework; Normalize structured medical entities using medical knowledge graphs; Output normalized medical entities.

2. The method according to claim 1, wherein: The unstructured medical text represented in Chinese includes at least one of a doctor's note, a nurse's note, a report, a treatment plan, a discharge summary, or a book.

3. The method according to claim 1 or 2, wherein: The medical entity includes at least one of a disease, a symptom, or a medical procedure.

4. The method according to claim 1 or 2, wherein: The multidimensional entity understanding framework includes a plurality of analyzers.

5. The method according to claim 4, wherein: The plurality of analyzers include at least one of a positive / negative entity analyzer, an intensity analyzer, a cause-effect analyzer, a pre-condition analyzer, a change pattern analyzer, a post-condition analyzer, a time analyzer, a frequency analyzer, and a body part analyzer.

6. The method according to claim 1 or 2, wherein: The medical knowledge graph is used to identify one or more synonymous medical entities that are synonymous with the medical entity.

7. A device for processing unstructured medical text expressed in Chinese, the device comprising: at least one memory configured to store program code; as well as At least one processor is configured to read the program code and operate according to the instructions of the program code, wherein the program code includes: Identification code, configured to enable the at least one processor to use an attention-based named entity recognition (NER) model to identify medical entities in the unstructured medical text expressed in Chinese; the NER model is a character-level and word-level bidirectional long short-term memory conditional random field model based on an attention mechanism; each word of the medical entity is represented by word-level information and character-level information; the word-level information and the character-level information are pre-trained by a skip-graph model of an unlabeled corpus; the NER model includes a character-level bidirectional LSTM, attention, a word-level bidirectional LSTM and a shared weight matrix; for each word, the character-level bidirectional LSTM is used to convert a pre-trained character embedding into a character-level output; the attention is used to obtain a weighted sum of the character-level output and the pre-trained word embedding using an attention value, and concatenate the word embedding with the character embedding; the input of the word-level bidirectional LSTM is the weighted sum; the shared weight matrix is ​​used to project the output of the word-level bidirectional LSTM into one or more predefined labels; structuring code configured to cause the at least one processor to structure the identified medical entity using a multidimensional entity understanding framework; normalization code configured to cause the at least one processor to normalize the structured medical entities using the medical knowledge graph; and Output code is configured to cause the at least one processor to output a normalized medical entity.

8. The device according to claim 7, wherein: The unstructured medical text represented in Chinese includes at least one of a doctor's note, a nurse's note, a report, a treatment plan, a discharge summary, or a book.

9. The device according to claim 7 or 8, wherein: The medical entity includes at least one of a disease, a symptom, or a medical procedure.

10. The device according to claim 7 or 8, wherein: The multidimensional entity understanding framework includes at least one of the following: a positive / negative entity analyzer, an intensity analyzer, a cause-effect analyzer, a pre-condition analyzer, a change pattern analyzer, a post-condition analyzer, a time analyzer, a frequency analyzer, and a body part analyzer.

11. The device according to claim 7 or 8, wherein: The medical knowledge graph is used to identify one or more synonymous medical entities that are synonymous with the medical entity.

12. A non-transitory computer-readable medium storing instructions, the instructions comprising one or more instructions, which when executed by one or more processors of a device for processing unstructured medical text represented in Chinese, cause the one or more processors to execute the method according to any one of claims 1 to 6.

13. A computer device comprising: Processor; and A memory, the memory being used to store a computer-readable program; The processor is configured to execute the computer-readable program to perform the method according to any one of claims 1 to 6.

14. A device for processing unstructured medical text expressed in Chinese, comprising: A recognition module, configured to use an attention-based named entity recognition (NER) model to recognize medical entities in the unstructured medical text expressed in Chinese; the NER model is a character-level and word-level bidirectional long short-term memory conditional random field model based on an attention mechanism; each word of the medical entity is represented by word-level information and character-level information; the word-level information and the character-level information are pre-trained through a skip-graph model of an unlabeled corpus; the NER model includes a character-level bidirectional LSTM, attention, a word-level bidirectional LSTM and a shared weight matrix; for each word, the character-level bidirectional LSTM is used to convert the pre-trained character embedding into a character-level output; the attention is used to obtain the weighted sum of the character-level output and the pre-trained word embedding using the attention value, and concatenate the word embedding with the character embedding; the input of the word-level bidirectional LSTM is the weighted sum; the shared weight matrix is ​​used to project the output of the word-level bidirectional LSTM into one or more predefined labels; a structuring module configured to structure the identified medical entities using a multidimensional entity understanding framework; a normalization module configured to normalize structured medical entities using a medical knowledge graph; The output module is configured to output normalized medical entities.

Images