WO2022222943A1 - 科室推荐方法、装置、电子设备及存储介质 - Google Patents

科室推荐方法、装置、电子设备及存储介质 Download PDF

Info

Publication number
WO2022222943A1
WO2022222943A1 PCT/CN2022/087819 CN2022087819W WO2022222943A1 WO 2022222943 A1 WO2022222943 A1 WO 2022222943A1 CN 2022087819 W CN2022087819 W CN 2022087819W WO 2022222943 A1 WO2022222943 A1 WO 2022222943A1
Authority
WO
WIPO (PCT)
Prior art keywords
disease
disease entity
entity
text
consultation
Prior art date
Application number
PCT/CN2022/087819
Other languages
English (en)
French (fr)
Inventor
赵璐偲
Original Assignee
康键信息技术(深圳)有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 康键信息技术(深圳)有限公司 filed Critical 康键信息技术(深圳)有限公司
Publication of WO2022222943A1 publication Critical patent/WO2022222943A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33Querying
    • G06F16/335Filtering based on additional data, e.g. user or group profiles
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/33Querying
    • G06F16/3331Query processing
    • G06F16/334Query execution
    • G06F16/3344Query execution using natural language analysis
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/36Creation of semantic tools, e.g. ontology or thesauri
    • G06F16/367Ontology
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00Handling natural language data
    • G06F40/20Natural language analysis
    • G06F40/279Recognition of textual entities
    • G06F40/289Phrasal analysis, e.g. finite state techniques or chunking
    • G06F40/295Named entity recognition
    • GPHYSICS
    • G16INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16HHEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H40/00ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
    • G16H40/20ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management or administration of healthcare resources or facilities, e.g. managing hospital staff or surgery rooms

Definitions

  • the present application relates to the field of artificial intelligence, and in particular, to a department recommendation method, apparatus, electronic device, and computer-readable storage medium.
  • the continuous development and improvement of artificial intelligence technology has greatly enriched and facilitated people's daily life.
  • many hospitals are currently equipped with machines that include intelligent guidance services.
  • the purpose is to use the Internet technology platform and machine learning and statistical learning methods to realize the online intelligent guidance function in a data-driven way, and recommend to patients.
  • Departments and doctors so as to maximize the subjective initiative of the patients, so that they can effectively and accurately find the medical services they need.
  • the online consultation platform also has an urgent need for intelligent consultation.
  • the number of consultation users is huge, while the number of online doctors is insufficient and the background is different. According to the results of intelligent consultation, it can help users to more accurately match the corresponding doctor, so It can improve the efficiency of online consultation.
  • the inventor realizes that although the application of the intelligent guidance service has achieved some results in improving the efficiency of medical treatment, there are still some problems in the design mode of most systems at present.
  • the asymmetry of information makes it more difficult to guide a doctor;
  • the current guide service is usually based on statistics or text classification methods to recommend departments and doctors, but statistical methods cannot provide personalized recommendation strategies based on patients' own symptoms. , and the latter is difficult to distinguish the noisy data in the text and extract the key medical factors (diseases, symptoms, etc.), which will also increase the difficulty of the guidance.
  • a department recommendation method provided by this application includes:
  • the disease entity in the standard text is identified by the pre-trained disease entity recognition model, and an entity relationship map is constructed for the identified disease entity.
  • the disease entity of the preset condition is obtained, and the first disease entity is obtained;
  • Dimensionality reduction is performed on the first disease entity and the second disease entity respectively, and the first disease entity and the second disease entity after dimensionality reduction are aggregated to obtain a target disease entity;
  • the degree of association between the target disease entity and the department in the pre-built medical department library is calculated, and a department whose degree of association is greater than a preset degree of association is selected from the medical department to obtain a target department.
  • the present application also provides a department recommendation device, the device comprising:
  • the cleaning module is used to clean the consultation data in the consultation text to obtain the standard text
  • the screening module is used to identify the disease entity in the standard text by using the disease entity recognition model completed by pre-training, and construct an entity relationship map for the identified disease entity, according to the entity relationship map, from the identified disease entity
  • the disease entities that meet the preset conditions are screened out from the entities, and the first disease entity is obtained;
  • the selection module is used for screening disease entities from the standard text by using a pre-built regular expression of disease entities, and calculating the matching degree of the screened disease entities with the disease entities in the preset disease entity dictionary library, from the screened disease entities. Selecting a disease entity whose matching degree is greater than a preset threshold from the disease entities to obtain a second disease entity;
  • an aggregation module configured to perform dimension reduction on the first disease entity and the second disease entity respectively, and summarize the first disease entity and the second disease entity after dimension reduction to obtain a target disease entity;
  • the calculation module is configured to calculate the degree of association between the target disease entity and the department in the pre-built medical department library, and select a department whose degree of association is greater than a preset degree of association from the medical department to obtain the target department.
  • the present application also provides an electronic device, the electronic device comprising:
  • the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to implement the department recommendation method as follows:
  • the disease entity in the standard text is identified by the pre-trained disease entity recognition model, and an entity relationship map is constructed for the identified disease entity.
  • the disease entity of the preset condition is obtained, and the first disease entity is obtained;
  • Dimensionality reduction is performed on the first disease entity and the second disease entity respectively, and the first disease entity and the second disease entity after dimensionality reduction are aggregated to obtain a target disease entity;
  • the degree of association between the target disease entity and the department in the pre-built medical department library is calculated, and a department whose degree of association is greater than a preset degree of association is selected from the medical department to obtain a target department.
  • the present application also provides a computer-readable storage medium, where at least one computer program is stored, and the at least one computer program is executed by a processor in an electronic device to implement the department recommendation method described below :
  • the disease entity in the standard text is identified by the pre-trained disease entity recognition model, and an entity relationship map is constructed for the identified disease entity.
  • the disease entity of the preset condition is obtained, and the first disease entity is obtained;
  • Dimensionality reduction is performed on the first disease entity and the second disease entity respectively, and the first disease entity and the second disease entity after dimensionality reduction are aggregated to obtain a target disease entity;
  • the degree of association between the target disease entity and the department in the pre-built medical department library is calculated, and a department whose degree of association is greater than a preset degree of association is selected from the medical department to obtain a target department.
  • FIG. 1 is a schematic flowchart of a department recommendation method provided by an embodiment of the present application
  • FIG. 2 is a detailed flowchart of one step of the department recommendation method provided in FIG. 1 in the first embodiment of the present application;
  • FIG. 3 is a schematic diagram of a module of a department recommendation device provided by an embodiment of the present application.
  • FIG. 4 is a schematic diagram of the internal structure of an electronic device for implementing a department recommendation method provided by an embodiment of the present application
  • the embodiments of the present application provide a method for department recommendation.
  • the execution subject of the department recommendation method includes, but is not limited to, at least one of electronic devices that can be configured to execute the method provided by the embodiments of the present application, such as a server and a terminal.
  • the department recommendation method can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform.
  • the server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, and the like.
  • the department recommendation method includes:
  • the consultation text can be understood as a data summary generated in a business scenario, such as a summary of the user's basic data and behavior data in a medical consultation scenario, and the consultation data refers to the user's input needs Consulted medical data, including: disease symptoms, disease category, and disease name, etc.
  • the data generated in the actual business scenario is complex and the amount of data is huge.
  • this application cleans the consultation data in the consultation text. , so as to filter out the useless data of the consultation text and improve the efficiency of subsequent text processing.
  • the cleaning of the inquiry data in the inquiry text to obtain the standard text includes: performing a deduplication operation on the inquiry data in the inquiry text, and detecting the deduplicated inquiry data. Whether there is a data missing value in the medical text; if there is no data missing value, the deduplicated medical consultation text is used as the standard text; if there is a data missing value, the data missing value is filled with data to obtain the standard text. text.
  • the deduplication operation on the consultation data in the consultation text includes: calculating the similarity of any two consultation data in the consultation text, if the similarity is greater than a preset similarity If the similarity is not greater than the preset similarity, then delete one of the two consultation data in the consultation text.
  • the embodiment of the present application further includes: using a hash algorithm to convert the consultation data in the consultation text into a corresponding hash value , in order to realize the calculation of the similarity of the follow-up consultation data.
  • the following method is used to calculate the similarity of any two consultation data in the consultation text:
  • d represents the similarity of any two consultation data in the consultation text
  • w 1j and w 2j represent the hash values of any two consultation data in the consultation text.
  • the input data may be missing.
  • the user's ID card or mobile phone number is missing one or more, resulting in incomplete user information, which affects the The integrity of the entire inquiry text, so the present application implements the filling of the existing data missing values by detecting whether there are missing data values in the inquiry text after deduplication.
  • the detection of the missing data value may be implemented by a detection function in a currently known missing data value detection tool, such as the missmapfunction detection function in the Ameliapackage tool.
  • the filling in the missing values of the data includes: acquiring a missing position of the data to be filled, a preset filling parameter at the missing position, and calculating a missing value probability of the filling parameter, according to The missing position, the filling parameter and the missing value probability are used to obtain the corresponding filling missing value of the data.
  • L( ⁇ ) represents the filled data missing value
  • xi represents the missing position of the ith data missing value
  • represents the filling parameter corresponding to the filled data missing value
  • n represents the deduplication in the consultation text.
  • ⁇ ) represents the probability of missing values for the filled parameter.
  • the disease entity recognition model includes a BERT neural network, which is used to identify disease entities in the standard text, wherein the disease entities refer to the disease name and disease symptoms in the inquiry text,
  • the names of the diseases include: gallstones, pulmonary nodules, colds and fever, etc.
  • the symptoms of the diseases include nausea, dizziness, weakness, and the like.
  • the embodiment of the present application identifies disease entities in the standard text through a disease entity recognition model completed by pre-training, so as to ensure the premise of subsequent target department matching.
  • the disease entity recognition model completed by pre-training to identify disease entities in the standard text includes:
  • the encoding layer includes Embedding
  • the S20 includes: using the encoding layer to perform index encoding on the characters in the standard sample to obtain a character encoding index; using the encoding layer to encode the characters Convert into a corresponding character vector to obtain an initial character vector; combine the character code index and the character vector to generate a character vector.
  • the feedforward attention mechanism includes: a self-attention module, a convolution module, and an encoder
  • the S21 includes: querying the character vector, using the convolution module in the feedforward attention mechanism to perform feature extraction on the queried character vector to obtain a feature character vector, and using the encoder in the feedforward attention mechanism to extract the feature character
  • the information sequence of the vector, the feature sequence vector is obtained.
  • the disease entity identification module includes: a fully connected layer and an activation function
  • the step S22 includes: using the fully connected layer to detect disease entity location information in the feature sequence vector, and using the activation The function outputs the location information of the disease entity to obtain the disease entity.
  • an entity relationship graph for the identified disease entity includes: using a translation model to perform relationship vector modeling on the identified disease entity to obtain the entity relationship graph vector space; convert the entity relationship graph vector space into an entity relationship graph of a visual interface to obtain the entity relationship graph.
  • Trans includes: multi-relational data embedding (TransE for short), knowledge embedding into a hyperplane (TransH for short), separate embedding of entities and relationships (TransR), embedding through dynamic mapping matrix (TransD) and Adaptive metric function (TransA).
  • TransE multi-relational data embedding
  • TransH knowledge embedding into a hyperplane
  • TransR separate embedding of entities and relationships
  • TransD embedding through dynamic mapping matrix
  • Adaptive metric function TransA
  • the transformation of the visualization interface of the entity relationship graph vector space is implemented by the currently known TensorBoard tool.
  • the embodiment of the present application can determine the association relationship between disease entities through the entity relationship map, so the embodiment of the present application selects the disease entities that meet the preset conditions from the identified disease entities according to the entity relationship map. disease entity, obtain the first disease entity.
  • the preset conditions may be set according to actual business scenarios, such as setting whether the disease entity has two or more associated disease entities, and if so, the corresponding disease entity is screened to generate the first disease.
  • the screening of the disease entity may be implemented by a currently preset entity screening script, and the preset entity screening script may be compiled by a currently known JavaScript script language.
  • the pre-built regular expression of disease entities may be constructed by using currently known entity naming recognition (Named Entity Recognition, NER) tools, such as finding high-frequency words describing main disease entities according to statistical analysis or Disease entities automatically selected by the system front-end.
  • NER Named Entity Recognition
  • the second disease entity is obtained by calculating the matching degree between the screened disease entity and the disease entity in the preset disease entity dictionary library, and selecting disease entities whose matching degree is greater than a preset threshold from the screened disease entities , to ensure the extraction accuracy of disease entities.
  • the disease entities in the disease entity dictionary library are obtained by downloading from a professional website and consulting professional medical personnel.
  • the disease entity dictionary library is constructed by a database, and is used to realize fast storage and reading of data.
  • the following method is used to calculate the degree of matching between the screened disease entity and the disease entity in the preset disease entity dictionary library:
  • T(x, y) represents the degree of matching
  • xi represents the ith disease entity in the screened disease entities
  • yi represents the ith disease entity in the disease entity dictionary library.
  • the preset threshold is 0.9, which can also be set according to actual business scenarios.
  • the first disease entity and the second disease entity are dimensionally reduced, so as to map the first disease entity and the second disease entity into the interval [0, 1], and improve the Processing speed of subsequent disease entities.
  • the first disease entity is dimensionally reduced by the following method:
  • x' represents the first disease entity after dimensionality reduction
  • X represents the first disease entity
  • X_max represents the disease entity with the largest dimension among the first disease entities
  • X_min represents the disease entity with the smallest dimension among the first disease entities.
  • the first disease entity and the second disease entity after dimensionality reduction are summarized to obtain the target disease entity, so as to ensure the completeness of the information of the disease entity when matching with the target department in the future, so as to improve the target disease entity. Accuracy of department matching.
  • the target disease entity can also be stored in a blockchain node.
  • S5. Calculate the degree of association between the target disease entity and a department in the pre-built medical department library, and select a department whose degree of association is greater than a preset degree of association from the medical department to obtain a target department.
  • the department in the medical department library refers to the location information of doctors in the medical field, such as: dermatology department, chest X-ray department, and electrocardiography department, etc.
  • the medical department library is also constructed based on the database, Easy access to data quickly.
  • the target department is obtained by calculating the degree of association between the target disease entity and the department in the medical department library, and selecting from the medical department a department whose degree of association is greater than a preset degree of association.
  • Target department matching for disease entities For the calculation method of the correlation degree, refer to the calculation method of the matching degree in the above S3, which will not be further elaborated here.
  • the optional preset correlation degree is 0.85, which can also be set according to actual business scenarios.
  • the embodiment of the present application first cleans the data in the acquired medical consultation text to obtain standard text, which can improve the processing speed of the data in the subsequent medical consultation text; Regular expressions screen disease entities from the standard text respectively, which ensures the comprehensiveness of disease entity acquisition, which can solve the problems that medical information does not match in the process of disease entities, and the noise data in the text is indistinguishable, thereby improving the problem.
  • the accuracy of clinic recommendation further, in the embodiment of the present application, the disease entity is dimensionally reduced to improve the speed of subsequent calculation, and according to the degree of association between the target disease entity and the department in the medical department database, from the medical department database Among the departments, the department with the correlation degree greater than the preset correlation degree is selected, so as to further improve the accuracy rate of the recommendation of the inquiring department. Therefore, a department recommendation method, device, electronic device and storage medium proposed in this application can reduce the difficulty of department recommendation.
  • FIG. 3 it is a functional block diagram of the department recommendation device of the present application.
  • the department recommendation device 100 described in this application may be installed in an electronic device.
  • the department recommendation device may include a cleaning module 101 , a screening module 102 , a selection module 103 , a summarizing module 104 and a computing module 105 .
  • the modules described in the present invention can also be called units, which refer to a series of computer program segments that can be executed by the electronic device processor and can perform fixed functions, and are stored in the memory of the electronic device.
  • each module/unit is as follows:
  • the cleaning module 101 is used to clean the consultation data in the consultation text to obtain a standard text
  • the screening module 102 is used to identify the disease entity in the standard text by using the disease entity recognition model completed by pre-training, and construct an entity relationship map for the identified disease entity, according to the entity relationship map, from the identified disease entity. Screening out disease entities that meet preset conditions from the disease entities to obtain a first disease entity;
  • the selection module 103 is used for screening disease entities from the standard text by using a pre-built disease entity regular expression, and calculating the matching degree between the screened disease entities and the disease entities in the preset disease entity dictionary library, Select disease entities whose matching degree is greater than a preset threshold from the screened disease entities to obtain a second disease entity;
  • the summarizing module 104 is configured to perform dimension reduction on the first disease entity and the second disease entity respectively, and summarize the first disease entity and the second disease entity after dimension reduction to obtain a target disease entity;
  • the calculation module 105 is configured to calculate the degree of association between the target disease entity and a department in a pre-built medical department library, and select a department whose degree of association is greater than a preset degree of association from the medical department to obtain the target department.
  • modules in the department recommendation device 100 in the embodiments of the present application use the same technical means as the department recommendation method described in the above-mentioned FIG. 1 and FIG. 2 , and can generate the same technology The effect will not be repeated here.
  • FIG. 4 it is a schematic structural diagram of an electronic device for implementing the department recommendation method of the present application.
  • the electronic device 1 may include a processor 10, a memory 11 and a bus, and may also include a computer program stored in the memory 11 and executed on the processor 10, such as a department recommendation program 12.
  • the memory 11 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, mobile hard disk, multimedia card, card-type memory (for example: SD or DX memory, etc.), magnetic memory, magnetic disk, Optical discs, etc., the computer-readable storage medium may be non-volatile or volatile.
  • the memory 11 may be an internal storage unit of the electronic device 1 in some embodiments, such as a mobile hard disk of the electronic device 1 . In other embodiments, the memory 11 may also be an external storage device of the electronic device 1, such as a pluggable mobile hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) equipped on the electronic device 1. card, flash memory card (FlashCard), etc.
  • the memory 11 may also include both an internal storage unit of the electronic device 1 and an external storage device.
  • the memory 11 can not only be used to store application software installed in the electronic device 1 and various types of data, such as codes of department recommendation programs 12, etc., but also can be used to temporarily store data that has been output or will be output.
  • the processor 10 may be composed of integrated circuits, for example, may be composed of a single packaged integrated circuit, or may be composed of multiple integrated circuits packaged with the same function or different functions, including one or more integrated circuits.
  • Central processing unit Central Processing unit, CPU
  • microprocessor digital processing chip
  • graphics processor and combination of various control chips, etc.
  • the processor 10 is the control core (ControlUnit) of the electronic device, and uses various interfaces and lines to connect the various components of the entire electronic device, by running or executing the program or module (for example, the execution department) stored in the memory 11. recommending programs 12, etc.), and calling data stored in the memory 11 to execute various functions of the electronic device 1 and process data.
  • the bus may be a peripheral component interconnect (PCI for short) bus or an extended industry standard architecture (extended industry standard architecture, EISA for short) bus or the like.
  • PCI peripheral component interconnect
  • EISA extended industry standard architecture
  • the bus can be divided into address bus, data bus, control bus and so on.
  • the bus is configured to implement connection communication between the memory 11 and at least one processor 10 and the like.
  • FIG. 4 only shows an electronic device with components. Those skilled in the art can understand that the structure shown in FIG. 4 does not constitute a limitation on the electronic device 1, and may include fewer or more components than those shown in the drawings. components, or a combination of certain components, or a different arrangement of components.
  • the electronic device 1 may also include a power supply (such as a battery) for powering the various components, preferably, the power supply may be logically connected to the at least one processor 10 through a power management device, so that the power management
  • the device implements functions such as charge management, discharge management, and power consumption management.
  • the power source may also include one or more DC or AC power sources, recharging devices, power failure detection circuits, power converters or inverters, power status indicators, and any other components.
  • the electronic device 1 may further include various sensors, Bluetooth modules, Wi-Fi modules, etc., which will not be repeated here.
  • the electronic device 1 may also include a network interface, optionally, the network interface may include a wired interface and/or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is usually used in the electronic device 1 Establish a communication connection with other electronic devices.
  • a network interface optionally, the network interface may include a wired interface and/or a wireless interface (such as a WI-FI interface, a Bluetooth interface, etc.), which is usually used in the electronic device 1 Establish a communication connection with other electronic devices.
  • the electronic device 1 may further include a user interface, and the user interface may be a display (Display), an input unit (eg, a keyboard (Keyboard)), optionally, the user interface may also be a standard wired interface or a wireless interface.
  • the display may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode, organic light-emitting diode) touch device, and the like.
  • the display may also be appropriately called a display screen or a display unit, which is used for displaying information processed in the electronic device 1 and for displaying a visualized user interface.
  • the department recommendation program 12 stored in the memory 11 in the electronic device 1 is a combination of multiple programs. When running in the processor 10, it can realize:
  • the disease entity in the standard text is identified by the pre-trained disease entity recognition model, and an entity relationship map is constructed for the identified disease entity.
  • the disease entity of the preset condition is obtained, and the first disease entity is obtained;
  • Dimensionality reduction is performed on the first disease entity and the second disease entity respectively, and the first disease entity and the second disease entity after dimensionality reduction are aggregated to obtain a target disease entity;
  • the degree of association between the target disease entity and the department in the pre-built medical department library is calculated, and a department whose degree of association is greater than a preset degree of association is selected from the medical department to obtain a target department.
  • the modules/units integrated in the electronic device 1 may be stored in a non-volatile computer-readable storage medium.
  • the computer-readable storage medium may be volatile or non-volatile.
  • the computer-readable medium may include: any entity or device capable of carrying the computer program code, recording medium, U disk, removable hard disk, magnetic disk, optical disk, computer memory, Read-Only Memory (ROM, Read-Only Memory) ).
  • the present application also provides a computer-readable storage medium, where the readable storage medium stores a computer program, and when executed by a processor of an electronic device, the computer program can realize:
  • the disease entity in the standard text is identified by the pre-trained disease entity recognition model, and an entity relationship map is constructed for the identified disease entity.
  • the disease entity of the preset condition is obtained, and the first disease entity is obtained;
  • Dimensionality reduction is performed on the first disease entity and the second disease entity respectively, and the first disease entity and the second disease entity after dimensionality reduction are aggregated to obtain a target disease entity;
  • the degree of association between the target disease entity and the department in the pre-built medical department library is calculated, and a department whose degree of association is greater than a preset degree of association is selected from the medical department to obtain a target department.
  • modules described as separate components may or may not be physically separated, and components shown as modules may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment.
  • each functional module in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.
  • the above-mentioned integrated units can be implemented in the form of hardware, or can be implemented in the form of hardware plus software function modules.
  • the blockchain referred to in this application is a new application mode of computer technology such as distributed data storage, point-to-point transmission, consensus mechanism, and encryption algorithm.
  • Blockchain essentially a decentralized database, is a series of data blocks associated with cryptographic methods. Each data block contains a batch of network transaction information to verify its Validity of information (anti-counterfeiting) and generation of the next block.
  • the blockchain can include the underlying platform of the blockchain, the platform product service layer, and the application service layer.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Business, Economics & Management (AREA)
  • Business, Economics & Management (AREA)
  • General Health & Medical Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Animal Behavior & Ethology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Epidemiology (AREA)
  • Medical Informatics (AREA)
  • Primary Health Care (AREA)
  • Public Health (AREA)
  • Medical Treatment And Welfare Office Work (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本申请涉及一种人工智能领域,揭露了一种科室推荐方法,包括:对问诊文本中的问诊数据进行清洗,得到标准文本;利用疾病实体识别模型识别标准文本中的疾病实体后构建实体关系图谱,根据实体关系图谱,生成第一疾病实体,利用疾病实体正则表达式从标准文本中筛选疾病实体,计算筛选的疾病实体与疾病实体字典库中疾病实体的匹配度,根据匹配度,生成第二疾病实体;对第一疾病实体和第二疾病实体分别进行降维后汇总,计算汇总后的疾病实体与医疗科室库中科室的关联度,从医疗科室中选取关联度大于预设关联度的科室,得到目标科室。此外,本申请还涉及区块链技术,所述目标疾病实体可存储于区块链中。本申请可以降低科室推荐难度。

Description

科室推荐方法、装置、电子设备及存储介质
本申请要求于2021年04月21日提交中国专利局、申请号为202110429296.9,发明名称为“科室推荐方法、装置、电子设备及存储介质”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本申请涉及人工智能领域,尤其涉及一种科室推荐方法、装置、电子设备及计算机可读存储介质。
背景技术
人工智能技术的不断发展和完善,极大的丰富和便利了人们的日常生活。例如,目前在医疗领域中,许多医院配置了包含智能导诊服务的机器,目的在于通过互联网技术平台和机器学习、统计学习方法,以数据驱动的方法实现线上智能导诊功能,给患者推荐科室和医生,从而最大发挥就诊人员的主观能动性,使其有效而精准地找到自己需要的医疗服务。同时,线上问诊平台也有迫切的智能导诊需求,问诊用户数量巨大,而在线医生资源数量不足且背景各异,根据智能导诊的结果,可以帮助用户更加精准的匹配对应医生,从而可以提高线上问诊的效率。
发明人意识到,虽然智能导诊服务的投入使用在提高就诊效率方面已经有一些成果,但目前大多数系统的设计模式仍然存在一些问题。第一、由于许多医疗机构针对不同疾病的科室划分趋于细致化、专业化,且不同医院的科室名称存在差异化,而一般患者缺乏专业的医疗知识,因此导致患者知识与医疗专业术语和医院信息的不对称,给导诊增加了难度;第二、目前导诊服务通常是基于统计或文本分类的方法进行科室和医生推荐,但统计方法无法根据患者自身的症状情况提供个性化的推荐策略,而后者又难以区分文本中的噪声数据,提取关键的医疗因素(疾病、症状等),从而也会给导诊增加一定的难度。
发明内容
本申请提供的一种科室推荐方法,包括:
对问诊文本中的问诊数据进行清洗,得到标准文本;
利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体;
利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
本申请还提供一种科室推荐装置,所述装置包括:
清洗模块,用于对问诊文本中的问诊数据进行清洗,得到标准文本;
筛选模块,用于利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述 疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体;
选取模块,用于利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
汇总模块,用于对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
计算模块,用于计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
本申请还提供一种电子设备,所述电子设备包括:
至少一个处理器;以及,
与所述至少一个处理器通信连接的存储器;其中,
所述存储器存储有可被所述至少一个处理器执行的计算机程序,所述计算机程序被所述至少一个处理器执行,以实现如下所述的科室推荐方法:
对问诊文本中的问诊数据进行清洗,得到标准文本;
利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体;
利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
本申请还提供一种计算机可读存储介质,所述计算机可读存储介质中存储有至少一个计算机程序,所述至少一个计算机程序被电子设备中的处理器执行以实现如下所述的科室推荐方法:
对问诊文本中的问诊数据进行清洗,得到标准文本;
利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体;
利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
附图说明
图1为本申请一实施例提供的科室推荐方法的流程示意图;
图2为本申请第一实施例中图1提供的科室推荐方法其中一个步骤的详细流程示意图;
图3为本申请一实施例提供的科室推荐装置的模块示意图;
图4为本申请一实施例提供的实现科室推荐方法的电子设备的内部结构示意图;
本申请目的的实现、功能特点及优点将结合实施例,参照附图做进一步说明。
具体实施方式
应当理解,此处所描述的具体实施例仅仅用以解释本申请,并不用于限定本申请。
本申请实施例提供一种科室推荐方法。所述科室推荐方法的执行主体包括但不限于服务端、终端等能够被配置为执行本申请实施例提供的该方法的电子设备中的至少一种。换言之,所述科室推荐方法可以由安装在终端设备或服务端设备的软件或硬件来执行,所述软件可以是区块链平台。所述服务端包括但不限于:单台服务器、服务器集群、云端服务器或云端服务器集群等。
参照图1所示,为本申请一实施例提供的科室推荐方法的流程示意图。在本申请实施例中,所述科室推荐方法包括:
S1、对所述问诊文本中的问诊数据进行清洗,得到标准文本。
本申请实施例中,所述问诊文本可以理解为在业务场景中产生的数据汇总,比如在医疗问诊场景中用户的基本数据和行为数据的汇总,所述问诊数据是指用户输入需要咨询的医疗数据,其包括:疾病症状、疾病类别以及疾病名称等。进一步地,应该了解,在实际业务场景中产生的数据错综复杂且数据量极大,为了更好的对所述问诊文本进行分析处理,本申请对所述问诊文本中的问诊数据进行清洗,以筛选出所述问诊文本的无用数据,提高后续文本处理的效率。
详细地,所述对所述问诊文本中的问诊数据进行清洗,得到标准文本,包括:对所述问诊文本中的问诊数据进行去重操作,并检测去重后的所述问诊文本中是否存在数据缺失值;若不存在数据缺失值,则将去重后的所述问诊文本作为标准文本;若存在数据缺失值,则对所述数据缺失值进行数据填充,得到标准文本。
详细地,所述对所述问诊文本中的问诊数据进行去重操作,包括:计算所述问诊文本中任意两个问诊数据的相似度,若所述相似度大于预设相似度,则同时保留所述问诊文本中任意两个问诊数据,若所述相似度不大于预设相似度,则删除所述问诊文本中任意两个问诊数据的一个问诊数据。
需要说明的是,本申请实施例在计算所述问诊文本中任意两个问诊数据的相似度之前,还包括:利用hash算法将所述问诊文本中的问诊数据转换成对应hash值,以实现后续问诊数据相似度的计算。
一个可选实施例中,利用下述方法计算所述问诊文本中任意两个问诊数据的相似度:
Figure PCTCN2022087819-appb-000001
其中,d表示问诊文本中任意两个问诊数据的相似度,w 1j和w 2j表示问诊文本中任意两个问诊数据的hash值。
进一步地,应该了解在实际业务场景中,用户在进行医疗咨询时,会存在输入的数据存在缺失的情况,比如用户的身份证或者手机号码少输一个或多个,导致用户信息不全,从而影响整个问诊文本的完整性,因此本申请通过检测去重后的所述问诊文本中是否存在数据缺失值,以实现对存在的数据缺失值填补。
一个可选实施例中,所述数据缺失值的检测可以通过当前已知的数据缺失值检测工具中的检测函数实现,如Ameliapackage工具中的missmapfunction检测函数实现。
一个可选实施例中,所述对所述数据缺失值进行填充包括:获取待填充数据的缺失位置,在所述缺失位置预设的填充参数,并计算所述填充参数的缺失值概率,根据所述缺失位置、填充参数以及缺失值概率,得到对应填充的数据缺失值。可选的,利用下述公式对所述数据缺失值进行填充:
Figure PCTCN2022087819-appb-000002
其中,L(θ)表示填充的数据缺失值,x i表示第i个数据缺失值的缺失位置,θ表示填充的数据缺失值对应的填充参数,n表示去重后的所述问诊文本中问诊数据的数量,p(x i|θ)表示填充参数的缺失值概率。
S2、利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体。
本申请实施例中,所述疾病实体识别模型包括BERT神经网络,用于识别所述标准文本中的疾病实体,其中,所述疾病实体是指所述问诊文本中的疾病名称和疾病症状,所述疾病名称包括:胆结石、肺结节以及感冒发烧等,所述疾病症状包括;恶心、眩晕以及无力等。为了更好的了解所述问诊文本中需要查询的目标科室,本申请实施例通过预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,以保障后续目标科室匹配的前提。
详细地,参阅图2所示,所述利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,包括:
S20、利用所述疾病实体识别模型中的编码层对所述标准样本中字符进行位置向量编码,生成字符向量;
S21、利用所述疾病实体识别模型中的前馈注意力机制对所述字符向量进行特征序列提取,得到特征序列向量;
S22、利用所述疾病实体识别模型中疾病实体识别模块的对所述特征序列向量进行疾病实体识别,得到疾病实体。
一个可选实施例中,所述编码层包括Embedding,所述S20包括:利用所述编码层对所述标准样本中的字符进行索引编码,得到字符编码索引;利用所述编码层将所述字符转换成对应的字符向量,得到初始字符向量;将所述字符编码索引和所述字符向量进行组合,生成字符向量。
一个可选实施例中,所述前馈注意力机制包括:自注意力模块、卷积模块以及编码器,所述S21包括:利用所述前馈注意力机制中的自注意力模块查询所述字符向量,利用所述前馈注意力机制中的卷积模块对查询到的所述字符向量进行特征提取,得到特征字符向量,利用所述前馈注意力机制中的编码器提取所述特征字符向量的信息序列,得到特征序列向量。
一个可选实施例中,所述疾病实体识别模块包括:全连接层和激活函数,所述S22包括:利用所述全连接层检测所述特征序列向量中的疾病实体位置信息,利用所述激活函数输出所述疾病实体位置信息,得到所述疾病实体。
需要说明的是,所述疾病实体识别模型的训练过程属于当前较为成熟的技术,在此不做进一步地阐述。
进一步地,应该了解,在识别出的疾病实体中,会存在一定的实体关系,比如疾病症状与疾病名称对应关系(如感冒包含无力),为了更加直观了解到识别的疾病实体之间的关系,本申请实施例对识别的所述疾病实体构建实体关系图谱,详细地,所述对识别的所述疾病实体构建实体关系图谱,包括:利用翻译模型对识别的所述疾病实体进行关系向量建模,得到所述实体关系图谱向量空间;将所述实体关系图谱向量空间转换为可视化界面的实体关系图谱,得到所述实体关系图谱。
其中,所述利用翻译模型(Trans)包括:多元关系数据嵌入(简称TransE)、将知识嵌入到超平面(简称TransH)、实体和关系分开嵌入(TransR)、通过动态映射矩阵嵌入 (TransD)以及自适应的度量函数(TransA)。需要声明的是,利用所述Trans实现实体-关系向量的建模属于当前较为成熟的技术,在此不做进一步地阐述。
一个可选实施例中,所述实体关系图谱向量空间的可视化界面的转换通过当前已知的TensorBoard工具实现。
进一步地,本申请实施例通过所述实体关系图谱,可以确定疾病实体之间的关联关系,因此本申请实施例根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体。其中,所述预设条件可以根据实际业务场景设置,比如设置疾病实体是否具有两个及两个以上关联的疾病实体,若具有,则筛选出对应疾病实体,生成第一疾病。
一个可选实施例中,所述疾病实体的筛选可以通过当前预设的实体筛选脚本实现,所述预设的实体筛选脚本可以由当前已知的JavaScript脚本语言编译。
S3、利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体。
本申请实施例中,所述预构建的疾病实体正则表达式可以通过当前已知的实体命名识别(NamedEntityRecognition,NER)工具进行构建,比如根据统计分析找到高频描述主要疾病实体词语或者根据用户在系统前端自动选择的疾病实体。
进一步地,应该了解,在实际业务场景中,由于用户的表达习惯不同等因素不同,会导致通过疾病实体正则表达式提取的疾病实体并不满足行业内标准的疾病实体,因此,本申请实施例通过计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,并从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体,以保障疾病实体的提取准确率。其中,所述疾病实体字典库中疾病实体通过从专业网站下载和咨询专业医疗人员得到,可选的,所述疾病实体字典库通过数据库构建,用于实现数据的快速存储和读取。
一个可选实施例中,利用下述方法计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度:
Figure PCTCN2022087819-appb-000003
其中,T(x,y)表示匹配度,x i表示筛选的所述疾病实体中第i个疾病实体,y i表示疾病实体字典库中第i个疾病实体。
一个可选实施例中,所述预设阈值为0.9,也可以根据实际业务场景设置。
S4、对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体。
本申请实施例中,将所述第一疾病实体和所述第二疾病实体进行降维,以将所述第一疾病实体和所述第二疾病实体映射到区间[0,1]中,提高后续疾病实体的处理速度。
一个可选实施例中,利用下述方法对所述第一疾病实体进行降维:
x'=(X-X_min)/(X_max-X_min)
其中,x'表示降维后的所述第一疾病实体,X表示第一疾病实体,X_max表示第一疾病实体中维度最大的疾病实体,X_min表示第一疾病实体中维度最小的疾病实体。
进一步地,所述第二疾病实体的降维方法可以参阅所述第一疾病实体的降维方法,在此不作进一步的赘述。
进一步地,本申请实施例将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体,以确保后续与目标科室匹配时疾病实体的信息全面性,以提高目标科室匹配的准确率。
进一步地,为保障所述目标疾病实体的复用性和隐私性,所述目标疾病实体还可存储于一区块链节点中。
S5、计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
本申请实施例中,所述医疗科室库中科室是指在医疗领域中的医生诊疗位置信息,比如:皮肤科室、胸透科室以及心电图科室等等,所述医疗科室库也基于数据库进行构建,方便数据的快速存取。
进一步地,本申请实施例通过计算所述目标疾病实体与医疗科室库中科室的关联度,并从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室,以实现疾病实体的目标科室匹配。其中,所述关联度的计算方法可以参阅上述S3中匹配度的计算方法,在此不做进一步阐述,可选的所述预设关联度为0.85,也可以根据实际业务场景设置。
本申请实施例首先对获取的问诊文本中数据进行清洗,得到标准文本,可以提高后续问诊文本中数据的处理速度;其次,本申请实施例利用预训练完成的疾病实体识别模型以及疾病实体正则表达式分别从所述标准文本中筛选疾病实体,保障了疾病实体获取的全面性,从而可以解决在疾病实体过程中医疗信息不匹配,及文本中出现噪声数据难以区分的问题,进而提高问诊科室推荐的准确性;进一步地,本申请实施例通过对疾病实体进行降维,以提高后续计算的速度,并根据所述目标疾病实体与医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,以进一步地提高问诊科室推荐的准确率。因此,本申请提出的一种科室推荐方法、装置、电子设备以及存储介质可以降低科室推荐难度。
如图3所示,是本申请科室推荐装置的功能模块图。
本申请所述科室推荐装置100可以安装于电子设备中。根据实现的功能,所述科室推荐装置可以包括清洗模块101、筛选模块102、选取模块103、汇总模块104以及计算模块105。本发所述模块也可以称之为单元,是指一种能够被电子设备处理器所执行,并且能够完成固定功能的一系列计算机程序段,其存储在电子设备的存储器中。
在本实施例中,关于各模块/单元的功能如下:
所述清洗模块101,用于对问诊文本中的问诊数据进行清洗,得到标准文本;
所述筛选模块102,用于利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体;
所述选取模块103,用于利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
所述汇总模块104,用于对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
所述计算模块105,用于计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
详细地,本申请实施例中所述科室推荐装置100中的所述各模块在使用时采用与上述的图1和图2中所述的科室推荐方法一样的技术手段,并能够产生相同的技术效果,这里不再赘述。
如图4所示,是本申请实现科室推荐方法的电子设备的结构示意图。
所述电子设备1可以包括处理器10、存储器11和总线,还可以包括存储在所述存储器11中并可在所述处理器10上运行的计算机程序,如科室推荐程序12。
其中,所述存储器11至少包括一种类型的可读存储介质,所述可读存储介质包括闪存、移动硬盘、多媒体卡、卡型存储器(例如:SD或DX存储器等)、磁性存储器、磁盘、光盘等,所述计算机可读存储介质可以是非易失性的,也可以是易失性的。所述存储器11在一些实施例中可以是电子设备1的内部存储单元,例如该电子设备1的移动硬盘。所述存储器11在另一些实施例中也可以是电子设备1的外部存储设备,例如电子设备1上配备的插接式移动硬盘、智能存储卡(SmartMediaCard,SMC)、安全数字(SecureDigital,SD)卡、闪存卡(FlashCard)等。进一步地,所述存储器11还可以既包括电子设备1的内部存储单元也包括外部存储设备。所述存储器11不仅可以用于存储安装于电子设备1的应用软件及各类数据,例如科室推荐程序12的代码等,还可以用于暂时地存储已经输出或者将要输出的数据。
所述处理器10在一些实施例中可以由集成电路组成,例如可以由单个封装的集成电路所组成,也可以是由多个相同功能或不同功能封装的集成电路所组成,包括一个或者多个中央处理器(CentralProcessingunit,CPU)、微处理器、数字处理芯片、图形处理器及各种控制芯片的组合等。所述处理器10是所述电子设备的控制核心(ControlUnit),利用各种接口和线路连接整个电子设备的各个部件,通过运行或执行存储在所述存储器11内的程序或者模块(例如执行科室推荐程序12等),以及调用存储在所述存储器11内的数据,以执行电子设备1的各种功能和处理数据。
所述总线可以是外设部件互连标准(peripheralcomponentinterconnect,简称PCI)总线或扩展工业标准结构(extendedindustrystandardarchitecture,简称EISA)总线等。该总线可以分为地址总线、数据总线、控制总线等。所述总线被设置为实现所述存储器11以及至少一个处理器10等之间的连接通信。
图4仅示出了具有部件的电子设备,本领域技术人员可以理解的是,图4示出的结构并不构成对所述电子设备1的限定,可以包括比图示更少或者更多的部件,或者组合某些部件,或者不同的部件布置。
例如,尽管未示出,所述电子设备1还可以包括给各个部件供电的电源(比如电池),优选地,电源可以通过电源管理装置与所述至少一个处理器10逻辑相连,从而通过电源管理装置实现充电管理、放电管理、以及功耗管理等功能。电源还可以包括一个或一个以上的直流或交流电源、再充电装置、电源故障检测电路、电源转换器或者逆变器、电源状态指示器等任意组件。所述电子设备1还可以包括多种传感器、蓝牙模块、Wi-Fi模块等,在此不再赘述。
进一步地,所述电子设备1还可以包括网络接口,可选地,所述网络接口可以包括有线接口和/或无线接口(如WI-FI接口、蓝牙接口等),通常用于在该电子设备1与其他电子设备之间建立通信连接。
可选地,该电子设备1还可以包括用户接口,用户接口可以是显示器(Display)、输入单元(比如键盘(Keyboard)),可选地,用户接口还可以是标准的有线接口、无线接口。可选地,在一些实施例中,显示器可以是LED显示器、液晶显示器、触控式液晶显示器以及OLED(OrganicLight-EmittingDiode,有机发光二极管)触摸器等。其中,显示器也可以适当的称为显示屏或显示单元,用于显示在电子设备1中处理的信息以及用于显示可视化的用户界面。
应该了解,所述实施例仅为说明之用,在专利申请范围上并不受此结构的限制。
所述电子设备1中的所述存储器11存储的科室推荐程序12是多个程序的组合,在所述处理器10中运行时,可以实现:
对问诊文本中的问诊数据进行清洗,得到标准文本;
利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出 满足预设条件的疾病实体,得到第一疾病实体;
利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
具体地,所述处理器10对上述程序的具体实现方法可参考图1对应实施例中相关步骤的描述,在此不赘述。
进一步地,所述电子设备1集成的模块/单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时,可以存储在一个非易失性计算机可读取存储介质中。所述计算机可读存储介质可以是易失性的,也可以是非易失性的。例如,所述计算机可读介质可以包括:能够携带所述计算机程序代码的任何实体或装置、记录介质、U盘、移动硬盘、磁碟、光盘、计算机存储器、只读存储器(ROM,Read-OnlyMemory)。
本申请还提供一种计算机可读存储介质,所述可读存储介质存储有计算机程序,所述计算机程序在被电子设备的处理器所执行时,可以实现:
对问诊文本中的问诊数据进行清洗,得到标准文本;
利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体;
利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
在本申请所提供的几个实施例中,应该理解到,所揭露的设备,装置和方法,可以通过其它的方式实现。例如,以上所描述的装置实施例仅仅是示意性的,例如,所述模块的划分,仅仅为一种逻辑功能划分,实际实现时可以有另外的划分方式。
所述作为分离部件说明的模块可以是或者也可以不是物理上分开的,作为模块显示的部件可以是或者也可以不是物理单元,即可以位于一个地方,或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。
另外,在本申请各个实施例中的各功能模块可以集成在一个处理单元中,也可以是各个单元单独物理存在,也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现,也可以采用硬件加软件功能模块的形式实现。
对于本领域技术人员而言,显然本申请不限于上述示范性实施例的细节,而且在不背离本申请的精神或基本特征的情况下,能够以其他的具体形式实现本申请。
因此,无论从哪一点来看,均应将实施例看作是示范性的,而且是非限制性的,本申请的范围由所附权利要求而不是上述说明限定,因此旨在将落在权利要求的等同要件的含义和范围内的所有变化涵括在本申请内。不应将权利要求中的任何附关联图标记视为限制所涉及的权利要求。
本申请所指区块链是分布式数据存储、点对点传输、共识机制、加密算法等计算机技术的新型应用模式。区块链(Blockchain),本质上是一个去中心化的数据库,是一串使用 密码学方法相关联产生的数据块,每一个数据块中包含了一批次网络交易的信息,用于验证其信息的有效性(防伪)和生成下一个区块。区块链可以包括区块链底层平台、平台产品服务层以及应用服务层等。
此外,显然“包括”一词不排除其他单元或步骤,单数不排除复数。系统权利要求中陈述的多个单元或装置也可以由一个单元或装置通过软件或者硬件来实现。第二等词语用来表示名称,而并不表示任何特定的顺序。
最后应说明的是,以上实施例仅用以说明本申请的技术方案而非限制,尽管参照较佳实施例对本申请进行了详细说明,本领域的普通技术人员应当理解,可以对本申请的技术方案进行修改或等同替换,而不脱离本申请技术方案的精神和范围。

Claims (20)

  1. 一种科室推荐方法,其中,所述方法包括:
    对问诊文本中的问诊数据进行清洗,得到标准文本;
    利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体;
    利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
    对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
    计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
  2. 如权利要求1所述的科室推荐方法,其中,所述对所述问诊文本中的问诊数据进行清洗,得到标准文本,包括:
    对所述问诊文本中的问诊数据进行去重操作,并检测去重后的所述问诊文本中是否存在数据缺失值;
    若不存在数据缺失值,则将去重后的所述问诊文作为标准文本;
    若存在数据缺失值,则对所述数据缺失值进行数据填充,得到标准文本。
  3. 如权利要求2所述的科室推荐方法,其中,所述对所述问诊文本中的问诊数据进行去重操作,包括:
    计算所述问诊文本中任意两个问诊数据的相似度;
    若所述相似度大于预设相似度,则同时保留所述问诊文本中任意两个问诊数据;
    若所述相似度不大于预设相似度,则删除所述问诊文本中任意两个问诊数据的一个问诊数据。
  4. 如权利要求1所述的科室推荐方法,其中,所述利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,包括:
    利用所述疾病实体识别模型中的编码层对所述标准样本中字符进行位置向量编码,生成字符向量;
    利用所述疾病实体识别模型中的前馈注意力机制对所述字符向量进行特征序列提取,得到特征序列向量;
    利用所述疾病实体识别模型中疾病实体识别模块的对所述特征序列向量进行疾病实体识别,得到疾病实体。
  5. 如权利要求4中所述的科室推荐方法,其中,所述利用所述疾病实体识别模型中的前馈注意力机制对所述字符向量进行特征序列提取,得到特征序列向量,包括:
    利用所述前馈注意力机制中的自注意力模块查询所述字符向量;
    利用所述前馈注意力机制中的卷积模块对查询到的所述字符向量进行特征提取,得到特征字符向量;
    利用所述前馈注意力机制中的编码器提取所述特征字符向量的信息序列,得到特征序列向量。
  6. 如权利要求1至5中任一项所述的科室推荐方法,其中,所述对识别的所述疾病实体构建实体关系图谱,包括:
    利用翻译模型对识别的所述疾病实体进行关系向量建模,得到所述实体关系图谱向量空间;
    将所述实体关系图谱向量空间转换为可视化界面的实体关系图谱,得到所述实体关系图谱。
  7. 如权利要求1所述的科室推荐方法,其中,所述计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,包括:
    利用下述方法计算所述匹配度:
    Figure PCTCN2022087819-appb-100001
    其中,T(x,y)表示匹配度,x i表示筛选的所述疾病实体中第i个疾病实体,y i表示疾病实体字典库中第i个疾病实体。
  8. 一种科室推荐装置,其中,所述装置包括:
    清洗模块,用于对问诊文本中的问诊数据进行清洗,得到标准文本;
    筛选模块,用于利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体;
    选取模块,用于利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
    汇总模块,用于对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
    计算模块,用于计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
  9. 一种电子设备,其中,所述电子设备包括:
    至少一个处理器;以及,
    与所述至少一个处理器通信连接的存储器;其中,
    所述存储器存储有可被所述至少一个处理器执行的计算机程序,所述计算机程序被所述至少一个处理器执行,以使所述至少一个处理器能够执行如下所述的科室推荐方法:
    对问诊文本中的问诊数据进行清洗,得到标准文本;
    利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体;
    利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
    对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
    计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
  10. 如权利要求9所述的电子设备,其中,所述对所述问诊文本中的问诊数据进行清洗,得到标准文本,包括:
    对所述问诊文本中的问诊数据进行去重操作,并检测去重后的所述问诊文本中是否存在数据缺失值;
    若不存在数据缺失值,则将去重后的所述问诊文作为标准文本;
    若存在数据缺失值,则对所述数据缺失值进行数据填充,得到标准文本。
  11. 如权利要求10所述的电子设备,其中,所述对所述问诊文本中的问诊数据进行去重操作,包括:
    计算所述问诊文本中任意两个问诊数据的相似度;
    若所述相似度大于预设相似度,则同时保留所述问诊文本中任意两个问诊数据;
    若所述相似度不大于预设相似度,则删除所述问诊文本中任意两个问诊数据的一个问诊数据。
  12. 如权利要求9所述的电子设备,其中,所述利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,包括:
    利用所述疾病实体识别模型中的编码层对所述标准样本中字符进行位置向量编码,生成字符向量;
    利用所述疾病实体识别模型中的前馈注意力机制对所述字符向量进行特征序列提取,得到特征序列向量;
    利用所述疾病实体识别模型中疾病实体识别模块的对所述特征序列向量进行疾病实体识别,得到疾病实体。
  13. 如权利要求12中所述的电子设备,其中,所述利用所述疾病实体识别模型中的前馈注意力机制对所述字符向量进行特征序列提取,得到特征序列向量,包括:
    利用所述前馈注意力机制中的自注意力模块查询所述字符向量;
    利用所述前馈注意力机制中的卷积模块对查询到的所述字符向量进行特征提取,得到特征字符向量;
    利用所述前馈注意力机制中的编码器提取所述特征字符向量的信息序列,得到特征序列向量。
  14. 如权利要求9所述的电子设备,其中,所述计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,包括:
    利用下述方法计算所述匹配度:
    Figure PCTCN2022087819-appb-100002
    其中,T(x,y)表示匹配度,x i表示筛选的所述疾病实体中第i个疾病实体,y i表示疾病实体字典库中第i个疾病实体。
  15. 一种计算机可读存储介质,存储有计算机程序,其中,所述计算机程序被处理器执行时实现如下所述的科室推荐方法:
    对问诊文本中的问诊数据进行清洗,得到标准文本;
    利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,并对识别的所述疾病实体构建实体关系图谱,根据所述实体关系图谱,从识别的所述疾病实体中筛选出满足预设条件的疾病实体,得到第一疾病实体;
    利用预构建的疾病实体正则表达式从所述标准文本中筛选疾病实体,并计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,从筛选的所述疾病实体中选取所述匹配度大于预设阈值的疾病实体,得到第二疾病实体;
    对所述第一疾病实体和所述第二疾病实体分别进行降维,将降维后的所述第一疾病实体和所述第二疾病实体汇总,得到目标疾病实体;
    计算所述目标疾病实体与预构建的医疗科室库中科室的关联度,从所述医疗科室中选取所述关联度大于预设关联度的科室,得到目标科室。
  16. 如权利要求15所述的计算机可读存储介质,其中,所述对所述问诊文本中的问诊数据进行清洗,得到标准文本,包括:
    对所述问诊文本中的问诊数据进行去重操作,并检测去重后的所述问诊文本中是否存在数据缺失值;
    若不存在数据缺失值,则将去重后的所述问诊文作为标准文本;
    若存在数据缺失值,则对所述数据缺失值进行数据填充,得到标准文本。
  17. 如权利要求16所述的计算机可读存储介质,其中,所述对所述问诊文本中的问诊数据进行去重操作,包括:
    计算所述问诊文本中任意两个问诊数据的相似度;
    若所述相似度大于预设相似度,则同时保留所述问诊文本中任意两个问诊数据;
    若所述相似度不大于预设相似度,则删除所述问诊文本中任意两个问诊数据的一个问诊数据。
  18. 如权利要求15所述的计算机可读存储介质,其中,所述利用预训练完成的疾病实体识别模型识别所述标准文本中的疾病实体,包括:
    利用所述疾病实体识别模型中的编码层对所述标准样本中字符进行位置向量编码,生成字符向量;
    利用所述疾病实体识别模型中的前馈注意力机制对所述字符向量进行特征序列提取,得到特征序列向量;
    利用所述疾病实体识别模型中疾病实体识别模块的对所述特征序列向量进行疾病实体识别,得到疾病实体。
  19. 如权利要求18中所述的计算机可读存储介质,其中,所述利用所述疾病实体识别模型中的前馈注意力机制对所述字符向量进行特征序列提取,得到特征序列向量,包括:
    利用所述前馈注意力机制中的自注意力模块查询所述字符向量;
    利用所述前馈注意力机制中的卷积模块对查询到的所述字符向量进行特征提取,得到特征字符向量;
    利用所述前馈注意力机制中的编码器提取所述特征字符向量的信息序列,得到特征序列向量。
  20. 如权利要求15所述的计算机可读存储介质,其中,所述计算筛选的所述疾病实体与预设的疾病实体字典库中疾病实体的匹配度,包括:
    利用下述方法计算所述匹配度:
    Figure PCTCN2022087819-appb-100003
    其中,T(x,y)表示匹配度,x i表示筛选的所述疾病实体中第i个疾病实体,y i表示疾病实体字典库中第i个疾病实体。
PCT/CN2022/087819 2021-04-21 2022-04-20 科室推荐方法、装置、电子设备及存储介质 WO2022222943A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110429296.9A CN113111162A (zh) 2021-04-21 2021-04-21 科室推荐方法、装置、电子设备及存储介质
CN202110429296.9 2021-04-21

Publications (1)

Publication Number Publication Date
WO2022222943A1 true WO2022222943A1 (zh) 2022-10-27

Family

ID=76719376

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/087819 WO2022222943A1 (zh) 2021-04-21 2022-04-20 科室推荐方法、装置、电子设备及存储介质

Country Status (2)

Country Link
CN (1) CN113111162A (zh)
WO (1) WO2022222943A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113111162A (zh) * 2021-04-21 2021-07-13 康键信息技术(深圳)有限公司 科室推荐方法、装置、电子设备及存储介质
CN113488159A (zh) * 2021-08-11 2021-10-08 中国医学科学院阜外医院 一种基于神经网络的医疗科室推荐方法及装置
CN113782189A (zh) * 2021-09-16 2021-12-10 平安国际智慧城市科技股份有限公司 基于地域性疾病图谱的智能辅助诊疗方法及装置
CN114596958B (zh) * 2022-03-15 2024-03-19 平安科技(深圳)有限公司 基于级联分类的病理数据分类方法、装置、设备及介质
CN115274086B (zh) * 2022-09-27 2023-01-13 无码科技(杭州)有限公司 一种智能导诊方法及系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180137250A1 (en) * 2016-11-15 2018-05-17 Hefei University Of Technology Mobile health intelligent medical guide system and method thereof
CN110085307A (zh) * 2019-04-04 2019-08-02 华东理工大学 一种基于多源知识图谱融合的智能导诊方法和系统
CN111816301A (zh) * 2020-07-07 2020-10-23 平安科技(深圳)有限公司 医疗问诊辅助方法、装置、电子设备及介质
CN111897967A (zh) * 2020-07-06 2020-11-06 北京大学 一种基于知识图谱和社交媒体的医疗问诊推荐方法
CN113111162A (zh) * 2021-04-21 2021-07-13 康键信息技术(深圳)有限公司 科室推荐方法、装置、电子设备及存储介质

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109036545B (zh) * 2018-05-31 2023-04-18 平安医疗科技有限公司 医疗信息处理方法、装置、计算机设备和存储介质
CN110648754A (zh) * 2018-06-27 2020-01-03 北京百度网讯科技有限公司 科室推荐方法、装置及设备
CN109817312A (zh) * 2018-12-13 2019-05-28 平安科技(深圳)有限公司 一种就诊引导方法及计算机设备
CN111403011B (zh) * 2020-03-12 2023-05-02 腾讯科技(深圳)有限公司 挂号科室推送方法、装置、系统、电子设备及存储介质
CN111785385A (zh) * 2020-06-29 2020-10-16 微医云(杭州)控股有限公司 疾病分类方法、装置、设备和存储介质
CN112001177A (zh) * 2020-08-24 2020-11-27 浪潮云信息技术股份公司 融合深度学习与规则的电子病历命名实体识别方法及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20180137250A1 (en) * 2016-11-15 2018-05-17 Hefei University Of Technology Mobile health intelligent medical guide system and method thereof
CN110085307A (zh) * 2019-04-04 2019-08-02 华东理工大学 一种基于多源知识图谱融合的智能导诊方法和系统
CN111897967A (zh) * 2020-07-06 2020-11-06 北京大学 一种基于知识图谱和社交媒体的医疗问诊推荐方法
CN111816301A (zh) * 2020-07-07 2020-10-23 平安科技(深圳)有限公司 医疗问诊辅助方法、装置、电子设备及介质
CN113111162A (zh) * 2021-04-21 2021-07-13 康键信息技术(深圳)有限公司 科室推荐方法、装置、电子设备及存储介质

Also Published As

Publication number Publication date
CN113111162A (zh) 2021-07-13

Similar Documents

Publication Publication Date Title
WO2022222943A1 (zh) 科室推荐方法、装置、电子设备及存储介质
US11232365B2 (en) Digital assistant platform
US10235633B2 (en) Method and system for linking heterogeneous data sources
CN113707297B (zh) 医疗数据的处理方法、装置、设备及存储介质
CN109906449B (zh) 一种查找方法及装置
WO2022121171A1 (zh) 相似文本匹配方法、装置、电子设备及计算机存储介质
WO2023029512A1 (zh) 基于知识图谱的医疗问题解答方法、装置、设备及介质
WO2023024411A1 (zh) 基于机器学习对关联规则进行评估的方法及装置
WO2022088632A1 (zh) 用户数据监控分析方法、装置、设备及介质
WO2022222942A1 (zh) 问答记录生成方法、装置、电子设备及存储介质
WO2022194062A1 (zh) 疾病标签检测方法、装置、电子设备及存储介质
WO2022100032A1 (zh) 系统分析可视化方法、装置、电子设备及计算机可读存储介质
WO2022062449A1 (zh) 用户分群方法、装置、电子设备及存储介质
WO2023178978A1 (zh) 基于人工智能的处方审核方法、装置、设备及介质
WO2022227171A1 (zh) 关键信息提取方法、装置、电子设备及介质
CN116578704A (zh) 文本情感分类方法、装置、设备及计算机可读介质
US20210202111A1 (en) Method of classifying medical records
CN116720525A (zh) 基于问诊数据的疾病辅助分析方法、装置、设备及介质
CN116844711A (zh) 基于深度学习的疾病辅助识别方法及装置
WO2022141860A1 (zh) 文本去重方法、装置、电子设备及计算机可读存储介质
US20140278527A1 (en) Large scale identification and analysis of population health risks
CN111444159B (zh) 精算数据处理方法、装置、电子设备及存储介质
US11270155B2 (en) Duplicate image detection based on image content
CN114927180A (zh) 病历结构化方法、装置及存储介质
CN115544218A (zh) 一种信息搜索方法、装置及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22791050

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 060224)