Medical data labeling method and system, terminal and storage medium

Abstract

The invention relates to the technical field of artificial intelligence, and particularly provides a medical data labeling method and system, a terminal and a storage medium, and the method comprises the steps: collecting historical diagnosis and treatment data which comprises medical records and medical images; recognizing medical record character data and medical image character data from the medical record and the medical image by utilizing a character recognition technology; extracting symptoms and diagnosis results from the medical record character data, and extracting symptoms from the medical image character data; based on the matching between the symptoms of the medical image and the symptoms of the case, establishing a corresponding relation between the medical image and the diagnosis result; and marking the corresponding diagnosis result as a label of the medical image by using a marking tool. According to the invention, the labeling efficiency of the medical image is improved.

Description

Technical Field

[0001] The present invention belongs to the field of artificial intelligence technology, and specifically relates to a medical data annotation method, system, terminal and storage medium. Background Art

[0002] With the rapid advancement of medical technology, artificial intelligence has penetrated deeper and deeper into the medical field, and its application scope has widely covered multiple key links such as accurate disease prediction, auxiliary diagnosis, and personalized treatment plan recommendation. Despite this, the current construction of medical AI models still faces significant challenges, especially the complexity and inefficiency of the data annotation process, which seriously restricts the acquisition of large-scale, high-quality sample data required for model training, and thus affects the full release of AI's application potential in the medical field. Therefore, optimizing the data annotation process and improving annotation efficiency have become key issues that need to be urgently addressed to promote the further development of medical AI technology. Summary of the invention

[0003] In view of the above-mentioned deficiencies in the prior art, the present invention provides a medical data annotation method, system, terminal and storage medium to solve the above-mentioned technical problems.

[0004] In a first aspect, the present invention provides a medical data labeling method, comprising: Collecting historical diagnosis and treatment data, including medical records and medical images; Using text recognition technology to identify medical record text data and medical image text data from medical records and medical images; Extract symptoms and diagnoses from medical record text data, and extract symptoms from medical imaging text data; Based on the matching of the symptoms of medical images and the symptoms of the cases, the corresponding relationship between medical images and diagnosis results is established; Use annotation tools to annotate the corresponding diagnosis results as labels of medical images.

[0005] In an optional implementation, using text recognition technology to recognize medical record text data and medical image text data from medical records and medical images includes: Obtain electronic medical records from case management systems; De-noising and enhancing electronic medical record files; Use deep learning algorithms to identify case text data from processed electronic medical records; The text area is extracted from the medical image based on the brightness difference, and the medical image text data is recognized from the text area using a deep learning algorithm.

[0006] In an optional embodiment, extracting symptoms and diagnosis results from medical record text data and extracting symptoms from medical image text data includes: Use named entity recognition to extract symptoms and diagnoses from medical record text data, and extract symptoms from medical imaging text data; Analyze the sentence structure in the case text data and extract the correspondence between symptoms and diagnosis results; The correspondence between symptoms and diagnosis results extracted from medical record text data is saved as structured data; The symptoms extracted from medical imaging text data are saved as structured data.

[0007] In an optional implementation, based on the matching of the symptoms of the medical image with the symptoms of the case, a correspondence between the medical image and the diagnosis result is established, including: Based on the symptoms of medical images and the correspondence between symptoms and diagnosis results, establish the correspondence between medical images and diagnosis results; The diagnosis result corresponding to the medical image is added to the file name of the medical image.

[0008] In a second aspect, the present invention provides a medical data annotation system, comprising: A data collection module, used to collect historical diagnosis and treatment data, wherein the historical diagnosis and treatment data includes medical records and medical images; A text recognition module, used to recognize medical record text data and medical image text data from medical records and medical images using text recognition technology; A text analysis module for extracting symptoms and diagnoses from medical record text data and symptoms from medical imaging text data; A file matching module is used to establish a correspondence between medical images and diagnosis results based on the matching of symptoms in medical images and symptoms in case records; The labeling execution module is used to use the labeling tool to label the corresponding diagnosis result as a label of the medical image.

[0009] In an optional embodiment, the data collection module includes: A file acquisition unit, used to acquire electronic case files from a case management system; A pre-processing unit, used for denoising and enhancing the electronic medical record files; A first recognition unit, used to recognize case text data of the processed electronic case file using a deep learning algorithm; The second recognition unit is used to capture a text area from the medical image based on brightness difference, and recognize medical image text data from the text area using a deep learning algorithm.

[0010] In an optional implementation, the text analysis module includes: An entity recognition unit, for extracting symptoms and diagnosis results from medical record text data and symptoms from medical image text data using a named entity recognition method; The relation extraction unit is used to analyze the sentence structure in the case text data and extract the corresponding relationship between symptoms and diagnosis results; A first conversion unit, used for saving the correspondence between the symptoms and the diagnosis results extracted from the medical record text data as structured data; The second conversion unit is used to save the symptoms extracted from the medical image text data as structured data.

[0011] In an optional implementation, the file matching module includes: A symptom matching unit, used for establishing a correspondence between the medical image and the diagnosis result based on the symptoms of the medical image and the correspondence between the symptoms and the diagnosis result; The naming update unit is used to add the diagnosis result corresponding to the medical image to the file name of the medical image.

[0012] In a third aspect, a terminal is provided, including: processor, memory, wherein: The memory is used to store computer programs. The processor is used to call and run the computer program from the memory, so that the terminal executes the above-mentioned terminal method.

[0013] In a fourth aspect, a computer storage medium is provided, wherein the computer-readable storage medium stores instructions, and when the computer-readable storage medium is run on a computer, the computer executes the methods described in the above aspects.

[0014] The beneficial effect of the present invention is that the medical data annotation method, system, terminal and storage medium provided by the present invention combine text recognition technology and text analysis technology to establish a correspondence between medical images and diagnostic results, and automatically annotate medical images based on the correspondence, thereby improving the annotation efficiency of medical images.

[0015] In addition, the invention has a reliable design principle, a simple structure and a very broad application prospect. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0017] Figure 1 is a schematic flow chart of a method according to an embodiment of the present invention.

[0018] Figure 2 is a schematic block diagram of a system according to an embodiment of the present invention.

[0019] Figure 3 A schematic diagram of the structure of a terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0022] The medical data annotation method provided in the embodiment of the present invention is executed by a computer device, and accordingly, the medical data annotation system runs in the computer device.

[0023] Figure 1 is a schematic flow chart of a method according to an embodiment of the present invention. Figure 1 The execution subject may be a medical data annotation system. According to different requirements, the order of the steps in the flowchart may be changed, and some may be omitted.

[0024] like Figure 1 As shown, the method includes: Step 110, collecting historical diagnosis and treatment data, wherein the historical diagnosis and treatment data includes medical records and medical images; Step 120, using text recognition technology to identify medical record text data and medical image text data from medical records and medical images; Step 130, extracting symptoms and diagnosis results from the medical record text data, and extracting symptoms from the medical image text data; Step 140, establishing a correspondence between the medical image and the diagnosis result based on the matching of the symptoms of the medical image and the symptoms of the case; Step 150: Use a labeling tool to label the corresponding diagnosis result as a label of the medical image.

[0025] To facilitate understanding of the present invention, the medical data labeling method provided by the present invention is further described below based on the principle of the medical data labeling method of the present invention and in combination with the process of labeling medical data in the embodiment.

[0026] Specifically, the medical data annotation methods include: S1. Collect historical diagnosis and treatment data, which includes medical records and medical images.

[0027] Collect various historical diagnosis and treatment data from the hospital's internal database and other hospitals' shared databases, including but not limited to detailed paper and electronic medical records, high-resolution medical imaging materials (such as X-rays, CT scans, MRI images, etc.). These data are the basis for subsequent analysis and model training, and are crucial to improving the accuracy of medical AI.

[0028] S2. Use text recognition technology to identify medical record text data and medical image text data from medical records and medical images.

[0029] S201. Obtain electronic medical record files from a case management system; perform denoising and enhancement processing on the electronic medical record files; and use a deep learning algorithm to identify case text data in the processed electronic medical record files.

[0030] Preprocessing steps: Format conversion: Convert electronic medical record files into a format suitable for subsequent processing (such as PDF to TXT or image format, depending on the file type and content). De-noising: Use image processing technology (for scanned copies) or text cleaning technology (for plain text files) to remove irrelevant elements in the file, such as headers and footers, advertisements, watermarks, garbled characters, etc. Image enhancement: For scanned medical record images, use image enhancement algorithms (such as contrast adjustment, brightness balance, sharpening, etc.) to improve the readability and recognition rate of text. Customized processing: Depending on the specific characteristics of the electronic medical record, customized denoising and enhancement processing may also be required. For example, for handwritten medical records, more complex handwritten text recognition preprocessing technology may be required.

[0031] Model selection and training: Select a suitable OCR (optical character recognition) model for training.

[0032] Training data preparation: Collect and annotate a large number of high-quality electronic medical record samples as training data to ensure that the model can learn the key features of the medical records.

[0033] Model training and optimization: Use training data to train the model and improve recognition accuracy by adjusting model parameters, optimizing algorithms, etc.

[0034] Text recognition execution: File input: Input the denoised and enhanced electronic medical record file into the trained deep learning model. Text recognition: The model automatically recognizes the text in the file and outputs the recognition results. This usually includes key content such as patient information, condition description, diagnosis results, treatment plan, etc. in the medical record.

[0035] S202, extracting a text area from the medical image based on brightness difference, and using a deep learning algorithm to recognize medical image text data from the text area.

[0036] Image standardization: First, medical images are standardized, including adjusting image size, resolution, color space, etc., to ensure consistency and accuracy of subsequent processing.

[0037] Contrast enhancement: To address the problems of uneven brightness and low contrast that may exist in medical images, appropriate image enhancement technology is used to improve the contrast and clarity of the image, making the text area more prominent.

[0038] Brightness analysis: Analyze the brightness distribution of medical images and identify areas with significantly different brightness from surrounding areas as potential text areas. This usually involves statistical analysis of pixel values ​​and threshold setting.

[0039] Region segmentation: Use image processing techniques (such as edge detection, morphological operations, etc.) to accurately segment potential text areas and separate text areas from complex backgrounds.

[0040] Denoising and smoothing: Denoise and smooth the segmented text area to eliminate image noise and jagged edges and improve the clarity of the text area.

[0041] Text recognition execution: The preprocessed and segmented text area of ​​the medical image is input into the trained OCR model. The model automatically extracts features, performs sequence modeling and character recognition on the text area, and finally outputs the recognition result.

[0042] S3. Extract symptoms and diagnosis results from medical record text data, and extract symptoms from medical image text data.

[0043] Use named entity recognition methods to extract symptoms and diagnosis results from medical record text data, and extract symptoms from medical image text data; analyze the sentence structure in the case text data and extract the correspondence between symptoms and diagnosis results; save the correspondence between symptoms and diagnosis results extracted from medical record text data as structured data; save the symptoms extracted from medical image text data as structured data.

[0044] NER of medical record text data: Select or train a NER model for the medical field that can identify entities such as symptoms, diseases, drugs, and examination items in medical record text. Preprocess the medical record text data, including word segmentation and part-of-speech tagging, to facilitate the input of the NER model. Use the NER model to automatically extract key information such as symptoms and diagnosis results from the medical record text.

[0045] NER of medical imaging text data: Similarly, for the text descriptions in medical imaging reports, the NER model in the medical field is used to identify symptom entities. Considering the particularity of medical imaging text data (such as many professional terms and complex sentence structures), the NER model needs to be adjusted or optimized to adapt to these characteristics.

[0046] Sentence structure analysis: Using the syntax analysis technology in natural language processing (NLP), the grammatical structure and dependency relationship of the sentences in the medical record text data are analyzed. Through syntax analysis, the logical relationship and semantic connection between the various components in the sentence are understood.

[0047] Correspondence extraction: Based on the results of syntactic analysis, design rules or algorithms to extract the correspondence between symptoms and diagnostic results. This may involve identifying the subject-verb-object structure, causal relationships, etc. in the sentence. Use pattern matching, statistical learning, or deep learning to further improve the accuracy and coverage of correspondence extraction.

[0048] Structured storage of medical record text data: symptoms, diagnosis results and their corresponding relationships extracted from medical record text data are saved in a structured data format (such as JSON, XML or database table). Design a reasonable data model to ensure data consistency and queryability.

[0049] Structured storage of medical imaging text data: The symptom entities extracted from the medical imaging text data are also saved in a structured data format. Consider combining the information of the medical image itself (such as examination type, patient information, etc.) with the symptom entities in the text description to form a more comprehensive medical record.

[0050] Throughout the process, strict data verification and quality control measures are implemented to ensure that the extracted information is accurate. The extracted symptoms, diagnostic results and their corresponding relationships are reviewed and corrected using methods such as manual review, cross-validation or automatic verification algorithms.

[0051] S4. Based on the matching of the symptoms of medical images with those of case, a corresponding relationship between medical images and diagnosis results is established.

[0052] Based on the symptoms of the medical image and the correspondence between the symptoms and the diagnosis results, a correspondence between the medical image and the diagnosis results is established; and the diagnosis results corresponding to the medical image are added to the file name of the medical image.

[0053] Specifically, design a reasonable file name naming convention to ensure that the file name contains necessary metadata (such as patient ID, examination date, etc.) and clearly reflects the diagnosis results of the medical image. For example, the file name can be in the format of "patient ID_examination date_diagnosis result brief description.extension". If there are multiple possible diagnosis results, the most likely diagnosis result can be selected or additional annotations can be added to indicate uncertainty.

[0054] S5. Use annotation tools to annotate the corresponding diagnosis results as labels of medical images.

[0055] Read the file name of the medical image and extract the diagnosis result from it. Use the diagnosis result as the annotation content and call the annotation tool to annotate the medical image.

[0056] The labeling tool uses LabelImg, and an automatic execution script is pre-compiled. The labeling tool is called through the automatic execution script, and the diagnosis result in the file name of the medical image is imported into the labeling tool.

[0057] Annotation tool optimization and automation process In the process of medical image annotation, we chose LabelImg, an open source and powerful image annotation tool, which supports multiple image formats and allows users to easily add labels to images through a graphical interface. In order to further improve the annotation efficiency, we pre-compiled an automatic execution script that can seamlessly integrate the LabelImg annotation tool to realize the automatic import and annotation of diagnostic results in medical image file names.

[0058] Design and implementation of automatic execution scripts: Script development environment configuration: First, configure the necessary libraries and dependencies in a suitable development environment (such as Python environment) to ensure that the script can be executed smoothly and call LabelImg.

[0059] LabelImg interface docking: Study the command line interface of LabelImg (if available) or write a wrapper layer so that the script can call the functions of LabelImg through specific commands or functions. If LabelImg does not have a direct command line interface, you may need to consider using a graphical interface automation tool (such as PyAutoGUI) to simulate user operations.

[0060] File name parsing module: Develop a module to read the file name of a medical image and parse out the part containing the diagnosis results. This usually involves string processing operations such as segmentation and searching.

[0061] Automatic annotation logic: The script starts LabelImg and loads the medical image file to be labeled.

[0062] The parsed diagnostic results are input into LabelImg as label content through an interface or graphical interface automation.

[0063] If needed, the script can also automatically set the label's properties such as position (such as the center of the image), size or color.

[0064] Repeat the above steps until all medical images to be labeled have been processed.

[0065] Execution of automated annotation process: The user places the medical image files to be annotated into a designated folder and ensures that the file name contains valid diagnosis result information.

[0066] Run the automatic execution script. The script will automatically traverse all medical image files in the specified folder, read the diagnosis results in the file name, and call LabelImg for annotation.

[0067] After the labeling is completed, the user can view and save the labeling results in LabelImg. If necessary, the script can also export the labeling results to a file in a specific format (such as XML, JSON, etc.) for subsequent analysis and use.

[0068] In some embodiments, the medical data annotation system may include multiple functional modules composed of computer program segments. The computer programs of each program segment in the medical data annotation system may be stored in a memory of a computer device and executed by at least one processor to perform (see Figure 1 Description) Function of medical data annotation.

[0069] In this embodiment, the medical data annotation system can be divided into multiple functional modules according to the functions it performs, such as Figure 2 As shown. The functional modules of the system 200 may include: a data collection module 210, a text recognition module 220, a text analysis module 230, a file matching module 240 and a annotation execution module 250. The module referred to in the present invention refers to a series of computer program segments that can be executed by at least one processor and can complete fixed functions, which are stored in a memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.

[0070] A data collection module, used to collect historical diagnosis and treatment data, wherein the historical diagnosis and treatment data includes medical records and medical images; A text recognition module, used to recognize medical record text data and medical image text data from medical records and medical images using text recognition technology; A text analysis module for extracting symptoms and diagnoses from medical record text data and symptoms from medical imaging text data; A file matching module is used to establish a correspondence between medical images and diagnosis results based on the matching of symptoms in medical images and symptoms in case records; The labeling execution module is used to use the labeling tool to label the corresponding diagnosis result as a label of the medical image.

[0071] Optionally, as an embodiment of the present invention, the data collection module includes: A file acquisition unit, used to acquire electronic case files from a case management system; A pre-processing unit, used for denoising and enhancing the electronic medical record files; A first recognition unit, used to recognize case text data of the processed electronic case file using a deep learning algorithm; The second recognition unit is used to capture a text area from the medical image based on brightness difference, and recognize medical image text data from the text area using a deep learning algorithm.

[0072] Optionally, as an embodiment of the present invention, the text analysis module includes: An entity recognition unit, for extracting symptoms and diagnosis results from medical record text data and symptoms from medical image text data using a named entity recognition method; The relation extraction unit is used to analyze the sentence structure in the case text data and extract the corresponding relationship between symptoms and diagnosis results; A first conversion unit, used for saving the correspondence between the symptoms and the diagnosis results extracted from the medical record text data as structured data; The second conversion unit is used to save the symptoms extracted from the medical image text data as structured data.

[0073] Optionally, as an embodiment of the present invention, the file matching module includes: A symptom matching unit, used for establishing a correspondence between the medical image and the diagnosis result based on the symptoms of the medical image and the correspondence between the symptoms and the diagnosis result; The naming update unit is used to add the diagnosis result corresponding to the medical image to the file name of the medical image.

[0074] Figure 3This is a schematic diagram of the structure of a terminal 300 provided in an embodiment of the present invention. The terminal 300 can be used to execute the medical data labeling method provided in an embodiment of the present invention.

[0075] The terminal 300 may include: a processor 310, a memory 320 and a communication unit 330. These components communicate via one or more buses. Those skilled in the art will appreciate that the server structure shown in the figure does not limit the present invention, and it may be a bus structure or a star structure, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0076] The memory 320 can be used to store the execution instructions of the processor 310, and the memory 320 can be implemented by any type of volatile or non-volatile storage terminal or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. When the execution instructions in the memory 320 are executed by the processor 310, the terminal 300 can perform some or all of the steps in the following method embodiments.

[0077] The processor 310 is the control center of the storage terminal, and uses various interfaces and lines to connect various parts of the entire electronic terminal. It runs or executes software programs and / or modules stored in the memory 320, and calls data stored in the memory to perform various functions of the electronic terminal and / or process data. The processor can be composed of an integrated circuit (IC), for example, it can be composed of a single packaged IC, or it can be composed of multiple packaged ICs with the same or different functions. For example, the processor 310 can only include a central processing unit (CPU). In the embodiment of the present invention, the CPU can be a single computing core or multiple computing cores.

[0078] The communication unit 330 is used to establish a communication channel so that the storage terminal can communicate with other terminals, receive user data sent by other terminals or send user data to other terminals.

[0079] The present invention also provides a computer storage medium, wherein the computer storage medium may store a program, and when the program is executed, the program may include some or all of the steps in each embodiment provided by the present invention. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM).

[0080] Therefore, the present invention combines text recognition technology and text analysis technology to establish a correspondence between medical images and diagnostic results, and automatically annotates medical images based on the correspondence, thereby improving the annotation efficiency of medical images. The technical effects that can be achieved by this embodiment can be found in the description above and will not be repeated here.

[0081] Those skilled in the art can clearly understand that the technology in the embodiments of the present invention can be implemented by means of software plus a necessary general hardware platform. Based on this understanding, the technical solution in the embodiments of the present invention, in essence or in other words, the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, and other media that can store program codes, including several instructions for enabling a computer terminal (which can be a personal computer, a server, or a second terminal, a network terminal, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention.

[0082] In this specification, the same or similar parts between the various embodiments can be referred to each other. In particular, for the terminal embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the description in the method embodiment.

[0083] In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of systems or modules, which can be electrical, mechanical or other forms.

[0084] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0085] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0086] Although the present invention has been described in detail with reference to the accompanying drawings and in combination with the preferred embodiments, the present invention is not limited thereto. Without departing from the spirit and essence of the present invention, a person of ordinary skill in the art may make various equivalent modifications or substitutions to the embodiments of the present invention, and these modifications or substitutions shall be within the scope of the present invention. Any person of ordinary skill in the art may easily think of changes or substitutions within the technical scope disclosed by the present invention, and these shall be within the scope of protection of the present invention.

Claims

1. A medical data annotation method, characterized in that: include: Collecting historical diagnosis and treatment data, including medical records and medical images; Using text recognition technology to identify medical record text data and medical image text data from medical records and medical images; Extract symptoms and diagnoses from medical record text data, and extract symptoms from medical imaging text data; Based on the matching of the symptoms of medical images and the symptoms of the cases, the corresponding relationship between medical images and diagnosis results is established; Use annotation tools to annotate the corresponding diagnosis results as labels of medical images.

2. The method according to claim 1, characterized in that Use text recognition technology to identify medical record text data and medical image text data from medical records and medical images, including: Obtain electronic medical records from case management systems; De-noising and enhancing electronic medical record files; Use deep learning algorithms to identify case text data from processed electronic medical records; The text area is extracted from the medical image based on the brightness difference, and the medical image text data is recognized from the text area using a deep learning algorithm.

3. The method according to claim 1, characterized in that: Extract symptoms and diagnoses from medical record text data, and extract symptoms from medical imaging text data, including: Use named entity recognition to extract symptoms and diagnoses from medical record text data, and extract symptoms from medical imaging text data; Analyze the sentence structure in the case text data and extract the correspondence between symptoms and diagnosis results; The correspondence between symptoms and diagnosis results extracted from medical record text data is saved as structured data; The symptoms extracted from medical imaging text data are saved as structured data.

4. The method according to claim 3, characterized in that Based on the matching of the symptoms of medical images and the symptoms of the cases, the corresponding relationship between medical images and diagnostic results is established, including: Based on the symptoms of medical images and the correspondence between symptoms and diagnosis results, establish the correspondence between medical images and diagnosis results; The diagnosis result corresponding to the medical image is added to the file name of the medical image.

5. A medical data annotation system, characterized in that: include: A data collection module, used to collect historical diagnosis and treatment data, wherein the historical diagnosis and treatment data includes medical records and medical images; A text recognition module, used to recognize medical record text data and medical image text data from medical records and medical images using text recognition technology; A text analysis module for extracting symptoms and diagnoses from medical record text data and symptoms from medical imaging text data; A file matching module is used to establish a correspondence between medical images and diagnosis results based on the matching of symptoms in medical images and symptoms in case records; The labeling execution module is used to use the labeling tool to label the corresponding diagnosis result as a label of the medical image.

6. The system according to claim 5, characterized in that The data collection module includes: A file acquisition unit, used to acquire electronic case files from a case management system; A pre-processing unit, used for denoising and enhancing the electronic medical record files; A first recognition unit, used to recognize case text data of the processed electronic case file using a deep learning algorithm; The second recognition unit is used to capture a text area from the medical image based on brightness difference, and recognize medical image text data from the text area using a deep learning algorithm.

7. The system according to claim 5, characterized in that The text analysis module includes: An entity recognition unit, for extracting symptoms and diagnosis results from medical record text data and symptoms from medical image text data using a named entity recognition method; The relation extraction unit is used to analyze the sentence structure in the case text data and extract the corresponding relationship between symptoms and diagnosis results; A first conversion unit, used for saving the correspondence between the symptoms and the diagnosis results extracted from the medical record text data as structured data; The second conversion unit is used to save the symptoms extracted from the medical image text data as structured data.

8. The system according to claim 7, characterized in that The file matching module includes: A symptom matching unit, used for establishing a correspondence between the medical image and the diagnosis result based on the symptoms of the medical image and the correspondence between the symptoms and the diagnosis result; The naming update unit is used to add the diagnosis result corresponding to the medical image to the file name of the medical image.

9. A terminal, characterized in that: include: A memory, used for storing a medical data annotation program; A processor, used to implement the steps of the medical data labeling method as described in any one of claims 1 to 4 when executing the medical data labeling program.

10. A computer-readable storage medium storing a computer program, characterized in that: The readable storage medium stores a medical data labeling program, and when the medical data labeling program is executed by the processor, the steps of the medical data labeling method according to any one of claims 1 to 4 are implemented.

Images