Medical data labeling method and device, electronic equipment and storage medium

Abstract

Embodiments of the present application disclose a medical data labeling method and device, electronic equipment and storage medium. The method comprises: obtaining to-be-labeled description information of to-be-labeled medical data; determining target description information matched with the to-be-labeled description information based on a pre-trained data matching model and a pre-stored medical standard data set; wherein the medical standard data set comprises standard description information of medical data and standard labeling information corresponding to the standard description information; determining the standard labeling information corresponding to the target description information as target labeling information of the to-be-labeled medical data, so as to label the to-be-labeled medical data according to the target labeling information. The technical scheme of the embodiments of the present application can improve the labeling efficiency and accuracy.

Description

TECHNICAL FIELD

[0001] Embodiments of the present application relate to the technical field of medical data labeling, and in particular to a medical data labeling method and device, electronic equipment and a storage medium. BACKGROUND

[0002] With the increasing number of medical institutions, the amount of data generated in the medical field is also increasing. Although the state has defined standards for data labeling of different types of medical data, in fact, each medical institution still labels medical data according to the labeling standards within the institution, resulting in different labeling information for the same data in different medical institutions, resulting in a large amount of medical data that needs to be stored and is difficult to maintain.

[0003] In the prior art, the acquired medical data is labeled according to the labeling standards in the form of manual labeling. However, in the process of implementing the present application, it is found that the prior art at least has the following technical problems: the manual labeling method has a large workload, is time-consuming and laborious, and is prone to errors in the manual labeling process, with low accuracy. SUMMARY

[0004] Embodiments of the present application provide a medical data labeling method, device, electronic equipment and storage medium to achieve the purpose of improving labeling efficiency and accuracy.

[0005] According to an aspect of the present application, a medical data labeling method is provided, comprising:

[0006] obtaining labeling description information of medical data to be labeled;

[0007] determining target description information matched with the labeling description information based on a pre-trained data matching model and a pre-stored medical standard data set;

[0008] wherein the medical standard data set includes standard description information of medical data and standard labeling information corresponding to the standard description information;

[0009] determining the standard labeling information corresponding to the target description information as target labeling information of the medical data to be labeled, and labeling the medical data to be labeled according to the target labeling information.

[0010] According to another aspect of the present application, a medical data labeling device is provided, comprising:

[0011] a labeling description information obtaining module for obtaining labeling description information of medical data to be labeled;

[0012] The target description information determination module is configured to determine target description information matched with the to-be-labeled description information based on a pre-trained data matching model and a pre-stored medical standard data set.

[0013] The medical standard data set includes standard description information of medical data and standard labeling information corresponding to the standard description information.

[0014] The target labeling information determination module is configured to determine the standard labeling information corresponding to the target description information as target labeling information of the to-be-labeled medical data, and label the to-be-labeled medical data according to the target labeling information.

[0015] According to another aspect of the present application, an electronic device is provided, which comprises:

[0016] at least one processor; and

[0017] a memory in communication with the at least one processor; wherein

[0018] 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 enable the at least one processor to perform the labeling method of medical data according to any one of the embodiments of the present application.

[0019] According to another aspect of the present application, a computer readable storage medium is provided, which stores computer instructions for enabling a processor to perform the labeling method of medical data according to any one of the embodiments of the present application when executed.

[0020] The technical solution of the embodiments of the present application comprises the following steps: obtaining to-be-labeled description information of to-be-labeled medical data, determining target description information matched with the to-be-labeled description information based on a pre-trained data matching model and a pre-stored medical standard data set, wherein the medical standard data set includes standard description information of medical data and standard labeling information corresponding to the standard description information, and determining the standard labeling information corresponding to the target description information as target labeling information of the to-be-labeled medical data, and labeling the to-be-labeled medical data according to the target labeling information. The entire labeling process does not require human intervention, solves the problems of time-consuming and laborious manual labeling and easy errors, and achieves the effects of improving labeling efficiency and accuracy.

[0021] It should be understood that the content described in this part is not intended to identify key or important features of the embodiments of the present application, nor to limit the scope of the present application. Other features of the present application will become apparent from the following description. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to make the technical solutions in the embodiments of the present application clearer, the following will briefly introduce the drawings needed in the embodiments description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0023] Figure 1 is a flow chart of a medical data labeling method according to an embodiment of the present application;

[0024] Figure 2 is a flow chart of another medical data labeling method according to an embodiment of the present application;

[0025] Figure 3 is a flow chart of a surgical data labeling method according to an embodiment of the present application;

[0026] Figure 4 is a structural schematic diagram of a medical data labeling device according to an embodiment of the present application;

[0027] Figure 5 is a structural schematic diagram of an electronic device for implementing a medical data labeling method according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to make the technical solutions in the embodiments of the present application clearer, the following will briefly introduce the drawings needed in the embodiments description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0029] It should be noted that the terms "first", "second", and the like in the specification and claims of the present application and the above drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or a chronological sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "comprise" and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but can include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0030] Figure 1Fig. 1 is a flowchart of a medical data labeling method according to an embodiment of the present application. The embodiment can be applied to the case of labeling medical data obtained by each medical institution according to a unified standard. The method can be executed by a medical data labeling device, which can be implemented in the form of hardware and / or software.

[0031] As shown in Fig. 1, the method of the embodiment can specifically include the following steps. Figure 1

[0032] S110, obtaining to-be-labeled description information of to-be-labeled medical data.

[0033] In a specific implementation, the original medical data in the medical material library can be obtained as the to-be-labeled medical data. The original medical data in the medical material library can also be distinguished according to whether there is original labeling information, and divided into a first data set and a second data set. When it is necessary to label the medical data with original labeling information, the medical data in the first data set is obtained as the to-be-labeled medical data. The to-be-labeled description information can be the name, data classification, and other information of the to-be-labeled medical data. For example, the to-be-labeled medical data can be surgical data, medical examination data, medical test data, and the like.

[0034] S120, determining target description information matched with the to-be-labeled description information based on a pre-trained data matching model and a pre-stored medical standard data set.

[0035] The data matching model includes a SimCSE model. The medical standard data set includes standard description information of medical data and standard labeling information corresponding to the standard description information. The standard labeling information is a unified standard defined by the medical field when labeling the medical data, and the standard description information is a unified standard defined by the medical field when describing the medical data. The standard description information can be data standard name, data standard classification, and the like. The standard labeling information can be code, character, and the like. For example, when the to-be-labeled medical data is surgical data, the medical standard data set includes a surgical standard name and a surgical standard code.

[0036] In a specific implementation, based on the pre-trained data matching model and the pre-stored medical standard data set, the target description information matched with the to-be-labeled description information is determined, including: if the to-be-labeled medical data already has original labeling information, determining the standard description information with the greatest similarity to the to-be-labeled description information in the medical standard data set as the target description information based on the data matching model.

[0037] The original labeling information is the labeling information of the to-be-labeled medical data when it is stored in the medical institution. For example, the original labeling information can be data code.

[0038] ​Specifically, when the to-be-labeled medical data has original labeling information, the original labeling information can reflect the characteristics of the to-be-labeled medical data. The to-be-labeled description information corresponding to the to-be-labeled medical data can be input into the data matching model, and the standard description information with the highest similarity to the to-be-labeled description information, i.e., the description information reflecting the characteristics of the to-be-labeled medical data, can be output. The output standard description information can be used as the target description information. In this embodiment, the original labeling information is used to quickly determine the target description information.

[0039] In a specific implementation, based on the pre-trained data matching model and the pre-stored medical standard data set, the target description information matched with the to-be-labeled description information is determined, including: if the to-be-labeled medical data does not have original labeling information, the first preset number of first standard description information with the highest similarity to the to-be-labeled description information is determined in the medical standard data set based on the data matching model; the second preset number of second standard description information with the highest similarity to the to-be-labeled description information is determined in the medical standard data set based on the cosine similarity algorithm; and the second standard description information that is repeated with the first standard description information is determined as the target description information.

[0040] It should be noted that the first preset number and the second preset number can be set according to the number of standard description information by those skilled in the art. For example, the first preset number can be 5, and the second preset number can be 1.

[0041] In this embodiment, for the case where there is no original labeling information, the target description information can be directly determined by using the to-be-labeled description information. To improve the accuracy of determining the target description information, the first standard description information and the second standard description information can be compared. If there is repeated standard description information in the first standard description information and the second standard description information, it indicates that the repeated standard description information is the standard description information with the highest similarity to the to-be-labeled description information, regardless of whether the standard description information is obtained based on the data matching model or the pre-similarity algorithm. The repeated standard description information can be determined as the target description information, so that the target description information of the to-be-labeled medical data without original labeling information can be quickly and accurately determined.

[0042] Further, if there is no repeated standard description information in the first standard description information and the second standard description information, the to-be-labeled medical data can be stored in the to-be-matched data set for staff to check.

[0043] S130, determining the standard labeling information corresponding to the target description information as the target labeling information of the to-be-labeled medical data, so as to label the to-be-labeled medical data according to the target labeling information.

[0044] Specifically, the target annotation information corresponding to the target description information can be determined in the medical standard data set, and the target annotation information is taken as the annotation information of the medical data to be annotated.

[0045] Further, for the case that the medical data to be annotated already has original annotation information, the standard annotation information corresponding to the target description information is determined as the target annotation information of the medical data to be annotated, including: determining whether the standard annotation information corresponding to the target description information is consistent with the original annotation information; if consistent, the standard annotation information corresponding to the target description information is determined as the target annotation information; if not consistent, the original annotation information is deleted, and the medical material library is updated based on the medical data to be annotated after the original annotation information is deleted.

[0046] To improve the accuracy of determining the target annotation information, for the case that the medical data to be annotated already has original annotation information, in order to fully consider the influence of the original annotation information on determining the target annotation information and improve the accuracy of determining the target description information, it can be determined whether the standard annotation information corresponding to the target description information is consistent with the original annotation information, if consistent, the standard annotation information corresponding to the target description information is directly determined as the target annotation information; if not consistent, to avoid annotation errors, the original annotation information can be deleted and the medical material library is updated, and the medical data to be annotated after the original annotation information is deleted is annotated according to the annotation mode of the medical data to be annotated without the original annotation information.

[0047] The technical scheme of the embodiment of the application, by obtaining the annotation description information of the medical data to be annotated, based on the pre-trained data matching model and the pre-stored medical standard data set, determines the target description information matched with the annotation description information; wherein the medical standard data set includes the standard description information of the medical data and the standard annotation information corresponding to the standard description information; the standard annotation information corresponding to the target description information is determined as the target annotation information of the medical data to be annotated, so as to annotate the medical data to be annotated according to the target annotation information. The whole annotation process does not need human participation, solves the problems of time-consuming and laborious manual annotation and easy errors, and realizes the effect of improving annotation efficiency and accuracy.

[0048] Figure 2 is a flowchart of another medical data annotation method provided by the embodiment of the application. Optionally, the annotation description information of the medical data to be annotated is obtained, including: performing a data cleaning operation on the original medical data stored in the medical material library; determining the medical data to be annotated in the original medical data after data cleaning, and obtaining the annotation description information of the medical data to be annotated; wherein the data cleaning operation includes at least one of data deduplication, data splitting and removing special identifiers in the original medical data. Wherein, the same or corresponding terms as in the above embodiments are not repeated here. For example, Figure 2As shown, the method comprises:

[0049] S210, performing a data cleaning operation on the original medical data stored in the medical material warehouse, determining the to-be-labeled medical data in the original medical data after data cleaning, and obtaining to-be-labeled description information of the to-be-labeled medical data.

[0050] The data cleaning operation includes at least one of data deduplication, data splitting, and removing special identifiers in the original medical data.

[0051] Specifically, when performing the data splitting operation, the original medical data can be split according to the punctuation marks contained in the original medical data; the punctuation marks can include at least one of a comma, a colon, a semicolon, and a colon. When removing the special identifiers in the original medical data, the numbers and parentheses before the name of the original medical data can be removed. Further, when performing the data cleaning operation, the original medical data can be compared with the standard medical data to remove redundant fields containing directional words and part words. According to the embodiment, the to-be-labeled description information of the to-be-labeled medical data is obtained after data cleaning, which facilitates the identification and matching of the to-be-labeled medical data by the data matching model, and is beneficial to improve the accuracy of determining the target labeling information of the to-be-labeled description information.

[0052] S220, determining the target description information matched with the to-be-labeled description information based on the pre-trained data matching model and the pre-stored medical standard data set.

[0053] S230, determining the standard labeling information corresponding to the target description information as the target labeling information of the to-be-labeled medical data, and labeling the to-be-labeled medical data according to the target labeling information.

[0054] Optionally, after determining the standard labeling information corresponding to the target description information as the target labeling information of the to-be-labeled medical data, the method further comprises: storing the labeled medical data into a labeled data set, extracting a preset proportion of the labeled medical data from the labeled data set as to-be-inspected medical data; sending to-be-inspected description information and to-be-inspected labeling information of the to-be-inspected medical data to an operation terminal; receiving an annotation accuracy value fed back by the operation terminal, and determining that the labeled medical data in the labeled data set is successfully labeled if the annotation accuracy value is greater than a preset threshold.

[0055] To ensure the accuracy and correctness of the labeled medical data, the labeling information of the labeled medical data can be sampled. Specifically, a preset proportion of the labeled medical data in the labeled data set is extracted as the to-be-inspected medical data. For example, the preset proportion can be 10%. The to-be-inspected description information and the to-be-inspected labeling information are sent to the operation terminal, so that the operation personnel manually inspect the to-be-inspected labeling information, and feed back the labeling accuracy degree value through the operation terminal; wherein the labeling accuracy degree value can be the number or proportion of the inspection passed labeling information. When the labeling accuracy degree value is greater than the preset threshold, it indicates that the accuracy of the labeling information of the labeled medical data is high, and it can be determined that the labeled medical data in the labeled data set is labeled successfully; when the labeling accuracy degree value is less than or equal to the preset threshold, it indicates that the accuracy of the labeling information of the labeled medical data is low, and the labeled medical data set is discarded, and the labeled medical data in the labeled medical data set is re-labeled. For example, the preset threshold can be 90%.

[0056] Optionally, it also includes receiving the labeling correction information fed back by the operation terminal, and correcting the error labeling information in the labeled data set based on the labeling correction information.

[0057] The labeling correction information includes the corrected correct labeling information and the error labeling information. The error to-be-inspected labeling information can be updated to the correct labeling information to complete the correction of the error labeling information. For example, for the labeled data set with a labeling accuracy degree value greater than the preset threshold, the error labeling information in the labeled data set can be corrected based on the labeling correction information received from the operation terminal, thereby further improving the accuracy of the labeling information in the labeled data set.

[0058] The above embodiments of the medical data labeling method are described in detail. In order to make the technical solution of the method further clear to those skilled in the art, taking the medical data as the operation data and the labeling information as the code as an example, the technical solution of the method is described.

[0059] Figure 3 is a flowchart of a surgical data labeling method according to an embodiment of the application; as Figure 3 shown, the specific process of labeling the surgical data includes:

[0060] 1. Surgical data can be collected from various hospital institutions to form a raw material library. The surgical data in the raw material library is subjected to data cleaning operation. The surgical standard data set is used as sample data of the model to train the SimCSE model, and the SimCSE model trained is used to label the surgical data.

[0061] 2, extract the operation data in the raw material warehouse, and the operation data has an original name and an original code. The original data can be divided into two parts based on whether the original code is empty.

[0062] 3, for the original code not empty operation data to be labeled, the SimCSE model is identified, the standard name of the standard operation data in the standard operation data set with the highest similarity to the original name of the operation data to be labeled is determined, and it is determined whether the standard code corresponding to the standard name is the same as the original code. If the same, the standard code is used as the code of the operation data to be labeled; If not, delete the original code of the operation data to be labeled, and process it as the operation data to be labeled with empty original code.

[0063] 4, for the operation data with empty original code, the original name is identified by SimCSE model and cosine similarity algorithm respectively; 5 first standard names with the highest similarity to the original name in the standard operation data set are determined by SimCSE model, and 1 second standard name with the highest similarity to the original name in the standard operation data set is determined by cosine similarity algorithm; If there is a standard code in the first standard code corresponding to each first standard name which is the same as the second standard code corresponding to the second standard name, the operation data to be labeled is labeled based on the repeated standard code; If not, discard the operation data to be labeled.

[0064] 5, the operation data with completed labeling is subjected to data sampling inspection, and if the correct rate after sampling inspection is more than 90%, it is judged that the labeled data set is successfully labeled, and if it is less than 90%, it is judged that the labeled data set is not successfully labeled, and the operation data in the labeled data set is discarded. When data sampling inspection, the operation data code can be de-duplicated to avoid extracting the same operation data code; The operation data after de-duplication can be randomly extracted.

[0065] 6, the operation data with completed labeling is stored in the unified database as the initialization data set, and the operation data to be labeled is compared with the operation data in the initialization data set before labeling. If there is repetition, the operation data to be labeled is discarded to filter out the repeated data.

[0066] 7, for the operation data with labeling error after sampling inspection, it can be sent to the operation terminal to make the operator correct manually, and the correct code of the operation data after correction is fed back through the operation terminal, and the operation data is labeled according to the correct code to improve the labeling accuracy.

[0067] In this embodiment, the SimCSE model and the cosine similarity algorithm are used to label the operation data in the raw material warehouse, and the operation data after labeling is subjected to sampling inspection to improve the accuracy of the coding labeling operation, improve the labeling efficiency, and reduce the waste of unnecessary human resources.

[0068] Figure 4 Fig. 1 is a structural schematic diagram of a medical data labeling device according to an embodiment of the present application. The device is used to execute the medical data labeling method provided by any of the above embodiments. The device and the medical data labeling method of each of the above embodiments belong to the same inventive concept. Details not described in the embodiment of the medical data labeling device can be referred to the embodiment of the medical data labeling method. As shown in Fig. 1, the device comprises: Figure 4

[0069] a to-be-labeled description information acquisition module 10, configured to acquire to-be-labeled description information of to-be-labeled medical data;

[0070] a target description information determination module 11, configured to determine target description information matched with the to-be-labeled description information based on a pre-trained data matching model and a pre-stored medical standard data set;

[0071] wherein the medical standard data set comprises standard description information of medical data and standard labeling information corresponding to the standard description information;

[0072] a target labeling information determination module 12, configured to determine the standard labeling information corresponding to the target description information as target labeling information of the to-be-labeled medical data, so as to label the to-be-labeled medical data according to the target labeling information.

[0073] In the embodiment of any optional technical solution of the present application, optionally, the data matching model comprises a SimCSE model; and the target description information determination module 11 comprises:

[0074] a first target description information determination unit, configured to, if the to-be-labeled medical data already has original labeling information, determine, based on the data matching model, the standard description information with the greatest similarity to the to-be-labeled description information in the medical standard data set as the target description information.

[0075] In the embodiment of any optional technical solution of the present application, optionally, the target labeling information determination module 12 comprises:

[0076] an original labeling information unit deletion, configured to determine whether the standard labeling information corresponding to the target description information is consistent with the original labeling information; if yes, the standard labeling information corresponding to the target description information is determined as the target labeling information; if not, the original labeling information is deleted, and the medical material warehouse is updated based on the to-be-labeled medical data after the original labeling information is deleted.

[0077] In the embodiment of any optional technical solution of the present application, optionally, the target description information determination module 11 comprises: ​

[0078] The first standard description information determination unit is configured to, if the to-be-labeled medical data does not have original label information, determine, based on a data matching model, a first preset number of first standard description information in the medical standard data set that has the greatest similarity with the to-be-labeled description information.

[0079] The second standard description information determination unit is configured to determine, based on a cosine similarity algorithm, a second preset number of second standard description information in the medical standard data set that has the greatest similarity with the to-be-labeled description information.

[0080] The second target description information is configured to determine, as target description information, the second standard description information that is repeated with the first standard description information.

[0081] In the embodiment of the present application, the to-be-labeled description information acquisition module 10 comprises, optionally:

[0082] The data cleaning unit is configured to perform a data cleaning operation on the original medical data stored in the medical raw material library.

[0083] The to-be-labeled description information acquisition unit is configured to determine, from the original medical data after data cleaning, to-be-labeled medical data, and acquire to-be-labeled description information of the to-be-labeled medical data.

[0084] The data cleaning operation comprises at least one of data deduplication, data splitting, and removal of special identifiers in the original medical data.

[0085] In the embodiment of the present application, the to-be-labeled description information acquisition module 10 comprises, optionally:

[0086] The labeled medical data extraction module is configured to, after determining, as target label information, the standard label information corresponding to the target description information, store the labeled medical data into a labeled data set, and extract a preset proportion of the labeled medical data from the labeled data set as to-be-inspected medical data.

[0087] The to-be-inspected label information sending module is configured to send the to-be-inspected description information and the to-be-inspected label information of the to-be-inspected medical data to the operation terminal.

[0088] The label accuracy value receiving module is configured to receive a label accuracy value fed back by the operation terminal, and determine that the labeled medical data in the labeled data set is labeled successfully if the label accuracy value is greater than a preset threshold.

[0089] In the embodiment of the present application, the to-be-labeled description information acquisition module 10 comprises, optionally:

[0090] The error labeling information correction module is configured to receive the labeling correction information fed back by the operation terminal, and correct the error labeling information in the labeled data set based on the labeling correction information.

[0091] The technical scheme of the embodiment of the application comprises the following steps: obtaining to-be-labeled description information of to-be-labeled medical data, determining target description information matched with the to-be-labeled description information based on a pre-trained data matching model and a pre-stored medical standard data set, wherein the medical standard data set comprises standard description information of medical data and standard labeling information corresponding to the standard description information, and determining the standard labeling information corresponding to the target description information as target labeling information of the to-be-labeled medical data, so as to label the to-be-labeled medical data according to the target labeling information. The entire labeling process does not need human participation, solves the problems of time-consuming and laborious manual labeling and easy errors, and achieves the effects of improving labeling efficiency and accuracy.

[0092] It is worth noting that the units and modules included in the above embodiment of the medical data labeling device are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of the functional units are only for easy mutual differentiation, and are not used to limit the protection scope of the application.

[0093] Figure 5 FIG. 1 is a structural schematic diagram of an electronic device for implementing the medical data labeling method according to the embodiment of the application. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframes, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown in this figure, their connections, and their functions, are merely examples and are not intended to limit the implementation of the application described and / or claimed herein.

[0094] As Figure 5As shown, the electronic device 20 includes at least one processor 21, and a memory, such as a read-only memory (ROM) 22, a random access memory (RAM) 23, etc., connected to the at least one processor 21 in communication. The memory stores computer programs executable by the at least one processor 21, and the processor 21 can perform various appropriate actions and processes according to the computer programs stored in the read-only memory (ROM) 22 or loaded from the storage unit 28 into the random access memory (RAM) 23. In the RAM 23, various programs and data required for the operation of the electronic device 20 can also be stored. The processor 21, the ROM 22, and the RAM 23 are connected to each other through a bus 24. An input / output (I / O) interface 25 is also connected to the bus 24.

[0095] Various components in the electronic device 20 are connected to the I / O interface 25, including an input unit 26, such as a keyboard, a mouse, etc., an output unit 27, such as various types of displays, a speaker, etc., a storage unit 28, such as a magnetic disk, an optical disk, etc., and a communication unit 29, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 29 allows the electronic device 20 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks.

[0096] The processor 21 can be various general and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 21 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 21 performs various methods and processes described above, such as the labeling method of medical data.

[0097] In some embodiments, the labeling method of medical data can be implemented as a computer program tangibly embodied in a computer readable storage medium, such as the storage unit 28. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 20 via the ROM 22 and / or the communication unit 29. When the computer program is loaded into the RAM 23 and executed by the processor 21, one or more steps of the labeling method of medical data described above can be performed. Alternatively, in other embodiments, the processor 21 can be configured to perform the labeling method of medical data by any other appropriate means, such as by means of firmware.

[0098] The various embodiments of the systems and techniques described above can be implemented in digital electronic circuitry, integrated circuitry, a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on a chip (SOC), a load programmable logic device (CPLD), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.

[0099] Computer programs used to implement the processes of the application can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer program

[0100] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store computer programs for use by or in connection with an instruction execution system, apparatus, or device. Computer-readable storage media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), optical fibers, portable compact disc read-only memories (CD-ROMs), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0101] To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.

[0102] The systems and techniques described here can be implemented in a computing system that includes a back end component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front end component (e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0103] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. A server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, to solve the defects of large management difficulty and weak business scalability in traditional physical host and VPS service.

[0104] It should be understood that the various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the scope of the present disclosure. For example, the steps recited in the present disclosure can be executed in parallel, executed in sequence, or executed in a different order, as long as the desired results of the present disclosure are achieved, and the present disclosure is not limited herein.

[0105] The specific embodiments described above are not intended to be limiting, and persons skilled in the art will appreciate that various modifications, combinations, sub-combinations and alternatives can be made to the specific embodiments without departing from the spirit and principles of the disclosure. Accordingly, the disclosure is not limited to the specific embodiments described above, but only by the scope of the appended claims.

Claims

1. A method for labeling medical data, characterized by, The method comprises: obtaining to-be-labeled description information of to-be-labeled medical data; determining target description information matched with the to-be-labeled description information based on a pre-trained data matching model and a pre-stored medical standard data set; wherein the medical standard data set comprises standard description information of medical data and standard labeling information corresponding to the standard description information; determining the standard labeling information corresponding to the target description information as target labeling information of the to-be-labeled medical data, so as to label the to-be-labeled medical data according to the target labeling information; wherein the data matching model comprises a SimCSE model; determining the target description information matched with the to-be-labeled description information based on the pre-trained data matching model and the pre-stored medical standard data set comprises: if the to-be-labeled medical data already has original labeling information, determining the standard description information with the greatest similarity to the to-be-labeled description information in the medical standard data set as the target description information based on the data matching model; determining the target description information matched with the to-be-labeled description information based on the pre-trained data matching model and the pre-stored medical standard data set comprises: if the to-be-labeled medical data does not have original labeling information, determining a first preset number of first standard description information with the greatest similarity to the to-be-labeled description information in the medical standard data set based on the data matching model; determining a second preset number of second standard description information with the greatest similarity to the to-be-labeled description information in the medical standard data set based on a cosine similarity algorithm; determining the second standard description information repeated with the first standard description information as the target description information; wherein the standard description information is a unified standard defined by the medical field when describing medical data, and the standard description information at least comprises data standard name and data standard classification.

2. The method of claim 1, wherein, determining the target labeling information of the to-be-labeled medical data based on the standard labeling information corresponding to the target description information comprises: determining whether the standard labeling information corresponding to the target description information is consistent with the original labeling information; if consistent, determining the standard labeling information corresponding to the target description information as the target labeling information; if inconsistent, deleting the original labeling information and updating a medical material library based on the to-be-labeled medical data after deleting the original labeling information.

3. The method of claim 1, wherein, obtaining to-be-labeled description information of to-be-labeled medical data comprises: performing a data cleaning operation on original medical data stored in a medical material library; determining the to-be-labeled medical data in the original medical data after data cleaning, and obtaining the to-be-labeled description information of the to-be-labeled medical data; wherein the data cleaning operation comprises at least one of data deduplication, data splitting, and removing special identifiers in the original medical data.

4. The method of claim 1, wherein, after determining the target labeling information of the to-be-labeled medical data based on the standard labeling information corresponding to the target description information, the method further comprises: Store the labeled medical data into a labeled data set, and extract a preset proportion of the labeled medical data in the labeled data set as to-be-inspected medical data; Send the to-be-inspected description information and to-be-inspected labeling information of the to-be-inspected medical data to an operation terminal; Receive a labeling accuracy degree value fed back by the operation terminal, and if the labeling accuracy degree value is greater than a preset threshold, determine that the labeled medical data in the labeled data set is labeled successfully.

5. The method of claim 4, wherein, Also includes: Receive labeling correction information fed back by the operation terminal, and correct the error labeling information in the labeled data set based on the labeling correction information. 6.A medical data labeling apparatus, characterized by comprising: Includes: A to-be-labeled description information acquisition module for acquiring to-be-labeled description information of to-be-labeled medical data; A target description information determination module for determining target description information matched with the to-be-labeled description information based on a pre-trained data matching model and a pre-stored medical standard data set; The medical standard data set includes standard description information of medical data and standard labeling information corresponding to the standard description information; A target labeling information determination module for determining the standard labeling information corresponding to the target description information as target labeling information of the to-be-labeled medical data, and labeling the to-be-labeled medical data according to the target labeling information; The data matching model includes a SimCSE model; The target description information determination module includes: If the to-be-labeled medical data already has original labeling information, the target description information most similar to the to-be-labeled description information in the medical standard data set is determined based on the data matching model; The target description information determination module includes: If the to-be-labeled medical data does not have original labeling information, the first preset number of first standard description information most similar to the to-be-labeled description information in the medical standard data set is determined based on the data matching model; The second preset number of second standard description information most similar to the to-be-labeled description information in the medical standard data set is determined based on a cosine similarity algorithm; The second standard description information repeated with the first standard description information is determined as the target description information; The standard description information is a unified standard defined by the medical field when describing medical data, and at least includes data standard name and data standard classification.

7. An electronic device, comprising: The electronic device includes: At least one processor; and The memory is in communication connection with the at least one processor; wherein 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 enable the at least one processor to execute the medical data labeling method in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for causing the processor to implement the medical data labeling method in any one of claims 1-5 when executed.

Images