Data collection method and system based on intelligent identification and filling of medical documents

Abstract

The present application discloses a data collection method and system based on intelligent recognition and filling of medical documents, which relates to the field of medical data technology. The method includes configuring key detection parameters of a target medical document based on an OCR tool; obtaining a medical document sample, analyzing and counting the document information category based on the medical document sample; dividing the document information category area corresponding to the position of the medical document sample based on the document information category, setting a regional identification mark according to the document information category area, and establishing regional logical associations based on the regional identification mark; establishing a medical document extraction template based on the regional identification mark and the regional logical association; using the key detection parameters and the OCR tool to identify the target medical document and obtain the medical document text information; extracting based on the medical document text information and the medical document extraction template, generating medical document key information, and filling the medical document key information into the system accordingly.

Description

Technical Field

[0001] The present application relates to the field of medical data technology, and in particular to a data collection method and system based on intelligent identification and filling of medical documents. Background Art

[0002] Traditionally, medical institutions have relied on manual data entry and paper archiving to manage medical documents, such as medical records and laboratory test reports. Medical staff spend a significant amount of time manually entering document information into electronic systems, which is not only inefficient but also prone to errors due to fatigue and negligence. Consequently, OCR (Optical Character Recognition) technology has been gradually introduced into the processing of medical documents.

[0003] However, existing medical document processing solutions based on OCR tools still have shortcomings. Common OCR recognition tools use universal recognition solutions for all types of images and lack specialized recognition solutions for medical documents. In addition, the formats and styles of medical documents vary greatly in different hospitals, making it difficult to accurately locate the text area of ​​medical documents, resulting in low recognition accuracy.

[0004] In view of this, we need a data collection method and system based on intelligent recognition and filling of medical documents to solve the above problems. Summary of the Invention

[0005] The purpose of this application is to solve the problem that common OCR tools have difficulty in accurately locating text areas, resulting in low recognition accuracy. In order to solve the above technical problems, a data collection method and system based on intelligent recognition and filling of medical documents is provided, which can accurately locate text areas and have higher recognition accuracy.

[0006] To achieve the above-mentioned purpose, the embodiment of the present application adopts the following technical solutions: a data collection method based on intelligent recognition and filling of medical documents, wherein the method is applied to a data collection system based on intelligent recognition and filling of medical documents, the method comprising: configuring key detection parameters of a target medical document based on an OCR tool, wherein the key detection parameters include detection limit side length, detection directory, detection image shape, and detection text length; obtaining a medical document sample, analyzing the medical document sample and counting the document information categories, wherein the document sample information categories include medical document date, patient name, drug name, and diagnosis information; Based on the document information category, the document information category area is divided according to the location of the medical document sample, and a regional identification mark is set according to the document information category area. The regional logical association is established based on the regional identification mark, wherein the document information category corresponds to the document information category area one by one; a medical document extraction template is established based on the regional identification mark and the regional logical association; the target medical document is identified using the detection key parameters and the OCR tool, and the medical document text information is obtained; the medical document text information and the medical document extraction template are used for extraction, and the medical document key information is generated. The medical document key information is marked and filled in the system accordingly.

[0007] Furthermore, according to an embodiment of the present application, the method also includes: based on the medical document sample, detecting and identifying the document table in the medical document sample, and segmenting the document table into document cells; separating the text information in the document cell according to the contrast difference between the document cell and the medical document sample, and generating a text binary image; inputting the text binary image into the OCR recognition module for recognition to extract the document information category.

[0008] Furthermore, according to an embodiment of the present application, the method further includes: obtaining a target medical document, and performing image preprocessing operations based on the target medical document; using a detection module to detect straight lines in the document image of the target medical document after the image preprocessing operation, and determining whether the straight lines in the document image meet a preset tilt angle, and if so, using a rotation matrix to perform rotation correction on the straight lines in the document image.

[0009] Further, according to an embodiment of the present application, obtaining a target medical document and performing an image preprocessing operation based on the target medical document specifically include the following steps: grayscale processing is performed on the target medical document to convert the collected document color image into a document grayscale image; filtering and denoising the document grayscale image, selecting the median value within the pixel range of the document grayscale image as the current pixel value to remove noise in the document grayscale image and generate a document denoised image; binarizing the document denoised image to obtain the target medical document after the image preprocessing operation.

[0010] Furthermore, according to an embodiment of the present application, the method also includes: determining whether the confidence of the key information of the medical document is lower than a preset confidence, and generating a confidence judgment result; comparing the recognition result of the key information of the medical document with the preset document key information, and generating a key information comparison result; summarizing the confidence judgment result and the key information comparison result and feeding them back to the system.

[0011] Further, according to an embodiment of the present application, the method specifically includes the following steps: when the confidence of the key information of the medical document is lower than the preset confidence, the confidence judgment result is uncertain content, otherwise the confidence judgment result is certain content; when the recognition result of the key information of the medical document is missing compared with the preset document key information, the key information comparison result is that the key information is missing, otherwise the key information comparison result is that the key information is complete.

[0012] To achieve the above-mentioned purpose, the embodiment of the present application also adopts the following technical solutions: a data acquisition system based on intelligent recognition and filling of medical documents, wherein the system includes: a parameter configuration module, the parameter configuration module is used to configure the key detection parameters of the target medical document based on the OCR tool, wherein the key detection parameters include the detection limit side length, the detection directory, the detection image shape and the detection text length; an information category statistics module, the information category statistics module is used to obtain medical document samples, analyze and count the document information categories based on the medical document samples, and the document sample information categories include the medical document date, patient name, drug name and diagnosis information; an identification area recognition module, the identification area recognition module is used to divide the position of the corresponding medical document sample based on the document information category Document information category area, an identification recognition area is set according to the document information category area, and a logical association of the identification area is established according to the identification recognition area, wherein the document information category corresponds to the document information category area one by one; an extraction template establishment module, an extraction template establishment module is used to establish a medical document extraction template based on the identification recognition area and the logical association of the identification area; a document text information acquisition module, a document text information acquisition module is used to use key detection parameters and OCR tools to identify the target medical document and obtain the medical document text information; a document key information filling module, a document key information filling module is used to extract based on the medical document text information and the medical document extraction template, and generate medical document key information, mark the medical document key information and fill it in accordingly into the system.

[0013] Furthermore, according to an embodiment of the present application, the system also includes: a cell segmentation module, which is used to detect and identify the document form in the medical document sample based on the medical document sample, and segment the document form into document cells; a text binary image generation module, which is used to separate the text information in the document cell according to the contrast difference between the document cell and the medical document sample, and generate a text binary image; a document information category extraction module, which is used to input the text binary image into the OCR recognition module for recognition to extract the document information category.

[0014] To achieve the above-mentioned purpose, an embodiment of the present application also discloses an electronic device, which includes a processor; a memory, in which computer program instructions are stored. When the computer program instructions are executed by the processor, the processor executes the data collection method based on intelligent identification and filling of medical documents as described above.

[0015] To achieve the above-mentioned purpose, an embodiment of the present application further discloses a computer-readable storage medium on which computer program instructions are stored. When the computer program instructions are executed by a processor, the processor executes the data collection method based on intelligent identification and filling of medical documents as described above. Beneficial effects

[0016] This application uses an OCR tool to configure key detection parameters of a target medical document; obtains a medical document sample, analyzes the medical document sample, and counts the document information category; based on the document information category, divides the document information category area corresponding to the position of the medical document sample, sets a region identification mark according to the document information category area, and establishes a region logical association based on the region identification mark; establishes a medical document extraction template based on the region identification mark and the region logical association; uses key detection parameters and an OCR tool to identify the target medical document and obtain the medical document text information; extracts the medical document based on the text information and the medical document extraction template, generates the medical document key information, and fills the medical document key information into the system accordingly, thereby achieving the technical effect of accurately locating the text area and accurately extracting the text content, and solving the technical problem that conventional OCR tools are difficult to accurately locate the text area, resulting in incorrect text content extraction. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The present application is further described below with reference to the accompanying drawings and examples.

[0018] Figure 1 This is a flowchart of the data collection method based on intelligent recognition and filling of medical documents in this application.

[0019] Figure 2This is a structural diagram of the data collection system based on intelligent recognition and filling of medical documents in this application.

[0020] Figure 3 It is a structural diagram of the electronic device of this application. DETAILED DESCRIPTION

[0021] In order to clearly and completely describe the objectives and technical solutions of the present invention and make the advantages more clearly understood, the embodiments of the present invention are further described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are part of the embodiments of the present invention, not all of them, and are only used to explain the embodiments of the present invention, not to limit the embodiments of the present invention. All other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0022] In the description of the present invention, it should be noted that the terms "center," "middle," "upper," "lower," "left," "right," "inner," "outer," "top," "bottom," "side," "vertical," "horizontal," and the like, indicating positions or location relationships, are based on the positions or location relationships shown in the accompanying drawings and are intended only to facilitate the description of the present invention and simplify the description. They are not intended to indicate or imply that the devices or components referred to must have a specific orientation, be constructed, or operate in a specific orientation. Therefore, they should not be construed as limiting the present invention. Furthermore, the terms "one," "first," "second," "third," "fourth," "fifth," and "sixth" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0023] In the description of the present invention, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they may refer to fixed, detachable, or integral connections; mechanical or electrical connections; direct or indirect connections through an intermediate medium; and internal communication between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on the specific circumstances.

[0024] For the purpose of simplicity and illustration, the principles of the embodiments are described primarily with reference to examples. In the following description, many specific details are provided to provide a thorough understanding of the embodiments. However, it will be apparent to those skilled in the art that these embodiments may not be limited to these specific details in practice. In some instances, well-known methods and structures are not described in detail to avoid unnecessarily obscuring the understanding of these embodiments. In addition, all embodiments may be used in combination with each other. Example 1

[0025] like Figure 1As shown, this embodiment provides a data collection method based on intelligent recognition and filling of medical documents, wherein the method is applied to a data collection system based on intelligent recognition and filling of medical documents, and the method includes:

[0026] S10. Configuring key detection parameters of the target medical document based on the OCR tool, wherein the key detection parameters include detection limit side length, detection directory, detection image shape, and detection text length;

[0027] It should be noted that the Optical Character Recognition (OCR) tool is an electronic device that detects characters on medical documents by detecting dark and light patterns. That is, it uses optical methods to recognize text on medical documents. Before recognition, the key detection parameters of the medical document must be configured to make the OCR tool more suitable for medical document recognition. The following are some key detection parameters and their configurations:

[0028] Detection limit side length: Set the maximum side length of the image during detection to adapt to the size of the medical document;

[0029] Detection directory: specifies the path of the detection model. You can select or train a specific detection model based on the characteristics of the medical document.

[0030] Detect image shape: Set the shape of the image during recognition to adapt to the size and proportion of the text on the medical document;

[0031] Detection text length: Set the maximum text length considered during recognition to accommodate long text that may appear on medical documents.

[0032] S20. Obtain a medical document sample, analyze the medical document sample, and count the document information categories, where the document sample information categories include the medical document date, patient name, drug name, and diagnosis information;

[0033] S30. Based on the document information categories, the document information category regions are divided according to the locations of the medical document samples. Region identification identifiers are set according to the document information category regions. A region logical association is established based on the region identification identifiers, wherein the document information categories correspond to the document information category regions in a one-to-one manner.

[0034] Specifically, we collected a large number of medical document samples, analyzed them, and counted the frequently appearing and key information categories in each document. In addition to the date, patient name, and drug name, we also included information such as the name of the medical institution, diagnosis and treatment items, and fee details, without limitation.

[0035] To determine the fields to be extracted, we delineate corresponding areas on the template based on the format of the medical document sample and the actual location of each information category in the medical document sample. This is known as the document information category area. For example, the date is usually located in the upper right or upper left corner of the document, the patient's name follows the specific identification area, and the drug name is often in the drug list.

[0036] Secondly, a unique area identification identifier is set for each document information category area so that the identified text can be accurately matched with the corresponding field when the text information is extracted later. This setting method can use image recognition or manual labeling, and at the same time establish regional logical associations, that is, the logical associations between each document information category area, such as the corresponding billing relationship between the fee details area and the medical treatment item area, and the patient name area is associated with the identity verification area to ensure completeness and accuracy.

[0037] S40, establishing a medical document extraction template based on the regional identification identifier and regional logical association;

[0038] S50, using key detection parameters and OCR tools to identify the target medical document and obtain the medical document text information;

[0039] S60: Extract based on the medical document text information and the medical document extraction template, generate key information of the medical document, mark the key information of the medical document and fill it into the system accordingly.

[0040] Specifically, the key information of the medical document is extracted according to the medical document extraction template through the identified text, that is, the text information of the medical document, and the extracted key information of the medical document such as the date, patient name, and drug name is mapped to the corresponding fields in the template to complete the identification and extraction of the target medical document. The extracted key information of the medical document is then marked and filled in the system accordingly.

[0041] Furthermore, the method further comprises:

[0042] Identify key information in medical documents and sensitive information databases, and analyze key information in medical documents with sensitive keywords in the sensitive information database to find out whether there is a match;

[0043] If there is a match, the sensitive keywords corresponding to the key information of the medical document will be fuzzy processed.

[0044] By using the medical document key information and sensitive information database for identification in the above method, it is possible to automatically identify whether the medical document key information contains patient sensitive information. If relevant sensitive information exists, the sensitive information is automatically coded. It should be noted that the sensitive information database and sensitive keywords are pre-set; secondly, considering special scenarios, users can also be supported to manually adjust the coded content, which can greatly reduce the risk of patient information leakage and the manual workload of patient information desensitization.

[0045] Furthermore, the method also includes: based on the medical document sample, detecting and identifying the document table in the medical document sample, and segmenting the document table into document cells; separating the text information in the document cell according to the contrast difference between the document cell and the medical document sample, and generating a text binary image; inputting the text binary image into the OCR recognition module for recognition to extract the document information category.

[0046] Specifically, the text binary image is input into the OCR recognition module for recognition. For example, the recognition process can be combined with the medical field vocabulary library to recognize the text in the text binary image to ensure that the text content can be accurately extracted, and the document information category is obtained after recognition.

[0047] Furthermore, the method further comprises:

[0048] Obtaining target medical documents and performing image preprocessing operations according to the target medical documents;

[0049] Among them, the preprocessing operation includes the following specific steps: grayscale processing is performed on the target medical document to convert the collected document color image into a document grayscale image; the document grayscale image is filtered and denoised, and the median value within the pixel range of the document grayscale image is selected as the current pixel value to remove the noise in the document grayscale image and generate a document denoised image; the document denoised image is binarized to obtain the target medical document after the image preprocessing operation.

[0050] Specifically, the target medical document is subjected to image preprocessing to ensure image quality optimization. First, grayscale processing is used to convert the collected document color image into a document grayscale image, that is, converting the color image into a grayscale image to reduce the interference of color information on subsequent processing. Secondly, the document grayscale image is filtered and denoised, and the median value within the pixel range of the document grayscale image is selected as the current pixel value to remove the noise in the document grayscale image. The noise can be salt and pepper noise and Gaussian noise, etc., so that the edges of the document grayscale image are clearer, which is more suitable for subsequent recognition, and a document denoised image is generated. Finally, the document denoised image is binarized to obtain the target medical document after image preprocessing, that is, the target medical document is converted into only black and white, enhancing the contrast between the text and the background, making the text part more obvious, and being more suitable for identifying and extracting text information.

[0051] After the image preprocessing operation, the target medical document uses the detection module to detect the straight line of the document image and determine whether the straight line of the document image meets the preset tilt angle. If so, the rotation matrix is ​​used to perform rotation correction on the straight line of the document image.

[0052] Specifically, the target medical document after image preprocessing is detected by a detection module. The inclination angle of the document image straight line in the target medical document is determined by the detection module, and an analysis is made as to whether the document image straight line meets the preset inclination angle. If so, a rotation matrix is ​​used to perform rotation correction on the document image straight line, so that the document image straight line, that is, the text line, in the target medical document remains horizontal, facilitating the subsequent extraction of text information.

[0053] Furthermore, the method also includes: determining whether the confidence level of the key information of the medical document is lower than a preset confidence level, and generating a confidence level determination result; comparing the recognition result of the key information of the medical document with the preset key information of the document, and generating a key information comparison result; summarizing the confidence level determination result and the key information comparison result and feeding them back to the system.

[0054] When the confidence level of the key information of a medical document is lower than the preset confidence level, the confidence judgment result is uncertain content, otherwise the confidence judgment result is certain content; when there is a missing in the recognition result of the key information of a medical document compared with the preset document key information, the key information comparison result is missing key information, otherwise the key information comparison result is complete key information.

[0055] Specifically, it is determined whether the confidence level of the key information of the medical document is lower than a preset confidence level, and a confidence level determination result is generated. When the confidence level of the key information of the medical document is lower than the preset confidence level, the confidence level determination result is uncertain; otherwise, the confidence level determination result is certain. For example, when the confidence level of the OCR recognition result is lower than a preset confidence level, assuming it is lower than 70%, the confidence level determination result is uncertain.

[0056] Compare the recognition result of the medical document key information with the preset document key information and generate a key information comparison result. If there is a missing in the recognition result of the medical document key information compared with the preset document key information, the key information comparison result is marked as key information missing. Otherwise, the key information comparison result is key information complete. For example, if the key fields "drug name" and "diagnosis information" do not appear in the recognition result, the key information comparison result is marked as key information missing.

[0057] The marked confidence judgment results and key information comparison results are summarized and fed back to the system. The summary can be sorted according to urgency and importance, with more urgent and important data placed at the front. During manual review, they are verified and corrected in order of priority. After the correction is completed, it is fed back to the system to update the corresponding data to ensure the accuracy of the identified and extracted content.

[0058] This application uses an OCR tool to configure key detection parameters of a target medical document; obtains a medical document sample, analyzes the medical document sample, and counts the document information category; based on the document information category, divides the document information category area corresponding to the position of the medical document sample, sets a region identification mark according to the document information category area, and establishes a region logical association based on the region identification mark; establishes a medical document extraction template based on the region identification mark and the region logical association; uses key detection parameters and an OCR tool to identify the target medical document and obtain the medical document text information; extracts the medical document based on the text information and the medical document extraction template, generates the medical document key information, and fills the medical document key information into the system accordingly, thereby achieving the technical effect of accurately locating the text area and accurately extracting the text content, and solving the technical problem that conventional OCR tools are difficult to accurately locate the text area, resulting in incorrect text content extraction. Example 2

[0059] like Figure 2 As shown, based on the same inventive concept as the data collection method based on intelligent recognition and filling of medical documents in the aforementioned embodiment, the present invention also provides a data collection system based on intelligent recognition and filling of medical documents, the system comprising:

[0060] A parameter configuration module is used to configure key detection parameters of the target medical document based on the OCR tool, where the key detection parameters include detection limit side length, detection directory, detection image shape, and detection text length;

[0061] An information category statistics module is used to obtain medical document samples, analyze them, and count the document information categories. The document sample information categories include medical document date, patient name, drug name, and diagnosis information.

[0062] The identification area recognition module is used to divide the document information category area based on the document information category and the location of the corresponding medical document sample, set the identification area according to the document information category area, and establish a logical association between the identification areas according to the identification area, wherein the document information category corresponds to the document information category area one by one;

[0063] An extraction template establishment module, which is used to establish a medical document extraction template based on the identification area and the logical association of the identification area;

[0064] The document text information acquisition module is used to identify the target medical document using key detection parameters and OCR tools, and obtain the medical document text information;

[0065] The document key information filling module is used to extract and generate medical document key information based on medical document text information and medical document extraction templates, mark the medical document key information and fill it into the system accordingly.

[0066] Furthermore, the system also includes:

[0067] Sensitive information analysis module: The sensitive information analysis module is used to identify key information of medical documents based on the sensitive information database, analyze the key information of medical documents with sensitive keywords in the sensitive information database, and find out whether there is a match;

[0068] The sensitive keyword fuzzy filling module is used to fuzzy process the sensitive keywords corresponding to the key information of the medical document if there is a match.

[0069] Furthermore, the system also includes: a cell segmentation module, which is used to detect and identify the document form in the medical document sample based on the medical document sample, and segment the document form into document cells; a text binary image generation module, which is used to separate the text information in the document cell according to the contrast difference between the document cell and the medical document sample, and generate a text binary image; a document information category extraction module, which is used to input the text binary image into the OCR recognition module for recognition to extract the document information category.

[0070] The various variations and specific examples of the data collection method based on intelligent recognition and filling of medical documents in the aforementioned embodiment 1 are also applicable to the data collection system based on intelligent recognition and filling of medical documents in this embodiment. Through the aforementioned detailed description of the data collection method based on intelligent recognition and filling of medical documents, those skilled in the art can clearly understand the implementation method of the data collection system based on intelligent recognition and filling of medical documents in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. Example 3

[0071] like Figure 3 As shown, the electronic device includes one or more processors and memory.

[0072] The processor may be a central processing unit (CPU) or other form of processing unit with data processing capability and / or instruction execution capability, and may control other components in the electronic device to perform desired functions.

[0073] The memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. The volatile memory may include, for example, random access memory (RAM) and / or cache memory. The non-volatile memory may include, for example, read-only memory (ROM), a hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and the processor may execute the program instructions to implement the data collection method for intelligent identification and filling of medical documents based on various embodiments of the present application described above and / or other desired functions. Various content, such as key detection parameters, may also be stored in the computer-readable storage medium, and the volatile memory may be capable of storing functions related to the present application.

[0074] Secondly, the electronic device also includes an input device and an output device, and these components are interconnected via a bus system and / or other forms of connection mechanisms (not shown).

[0075] For example, when the electronic device is a stand-alone device, the input device may be a communication network connector. In addition, the input device may also include, for example, a keyboard, a mouse, and the like.

[0076] The output device can output various information to the outside, including the generated area identification mark, etc. The output device can include, for example, a display, a speaker, a printer, a communication network and a remote output device connected thereto, etc.

[0077] Of course, for simplicity, the figure only shows some components of the electronic device related to the present application, omitting components such as buses, input / output interfaces, etc. In addition, the electronic device may further include any other appropriate components depending on the specific application.

[0078] In addition, an embodiment of the present application may also be a computer-readable storage medium having computer program instructions stored thereon, which, when executed by a processor, enables the processor to execute the steps of the data collection method based on intelligent identification and filling of medical documents according to various embodiments of the present application described in the above "Exemplary Method" section of this specification.

[0079] The computer readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can, for example, include but is not limited to a system, device or component of electricity, magnetism, light, electromagnetic, infrared, or semiconductor, or any combination thereof. More specific examples (non-exhaustive list) of readable storage media include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0080] Although the above describes the illustrative specific implementation methods of the present application so that those skilled in the art can understand the present application, the present application is not limited to the scope of the specific implementation methods. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present application defined and determined by the attached claims, all application creations based on the concept of the present application are protected.

Claims

1. A data collection method based on intelligent identification and filling of medical documents, wherein: The method is applied to a data collection system based on intelligent recognition and filling of medical documents, and the method comprises: Configuring key detection parameters of the target medical document based on the OCR tool, wherein the key detection parameters include detection limit side length, detection directory, detection image shape, and detection text length; Obtaining a medical document sample, analyzing the medical document sample and counting document information categories, wherein the document sample information categories include the medical document date, patient name, drug name, and diagnosis information; Based on the document information category, document information category areas are divided corresponding to the positions of the medical document samples, area identification marks are set according to the document information category areas, and area logical associations are established according to the area identification marks, wherein the document information categories correspond to the document information category areas in a one-to-one manner; Establishing a medical document extraction template based on the regional identification identifier and the regional logical association; Using the key detection parameters and the OCR tool to identify the target medical document and obtain medical document text information; Extracting the medical document text information and the medical document extraction template, generating key information of the medical document, marking the key information of the medical document and filling it into the system accordingly; Identify the medical document key information based on the medical document key information and the sensitive information database, analyze the medical document key information with sensitive keywords in the sensitive information database, and find out whether there is a match; If there is a matching item, the sensitive keyword corresponding to the key information of the medical document is fuzzy processed.

2. The data collection method based on intelligent identification and filling of medical documents according to claim 1 is characterized in that: The method further comprises: Based on the medical document sample, detecting and identifying a document form in the medical document sample, and segmenting the document form into document cells; Separating text information in the document cells according to a contrast difference between the document cells and the medical document sample, and generating a text binary image; The text binary image is input into an OCR recognition module for recognition to extract the document information category.

3. The data collection method based on intelligent identification and filling of medical documents according to claim 1 is characterized in that: The method further comprises: Acquiring the target medical document and performing image preprocessing operations according to the target medical document; After the image preprocessing operation, the target medical document uses a detection module to detect the straight line of the document image, and determines whether the straight line of the document image meets the preset tilt angle. If so, the rotation matrix is ​​used to perform rotation correction on the straight line of the document image.

4. The data collection method based on intelligent identification and filling of medical documents according to claim 3 is characterized in that: The obtaining of the target medical document and performing image preprocessing according to the target medical document specifically includes the following steps: Grayscale processing is performed on the target medical document to convert the collected document color image into a document grayscale image; Filtering and denoising the document grayscale image, selecting the median value within the pixel range of the document grayscale image as the current pixel value to remove noise in the document grayscale image and generate a document denoised image; The denoised document image is binarized to obtain the target medical document after image preprocessing.

5. The data collection method based on intelligent identification and filling of medical documents according to claim 1 is characterized in that: The method further comprises: Determining whether the confidence level of the key information of the medical document is lower than a preset confidence level, and generating a confidence level determination result; Comparing the recognition result of the medical document key information with the preset document key information, and generating a key information comparison result; The confidence judgment results and the key information comparison results are summarized and fed back to the system.

6. The data collection method based on intelligent identification and filling of medical documents according to claim 5 is characterized in that: The method specifically comprises the following steps: When the confidence level of the key information of the medical document is lower than the preset confidence level, the confidence level judgment result is uncertain content; otherwise, the confidence level judgment result is certain content; When there is a missing in the comparison between the recognition result of the medical document key information and the preset document key information, the key information comparison result is that the key information is missing; otherwise, the key information comparison result is that the key information is complete.

7. A data collection system based on intelligent identification and filling of medical documents, characterized by: The system comprises: A parameter configuration module, configured to configure key detection parameters of a target medical document based on an OCR tool, wherein the key detection parameters include detection limit side length, detection directory, detection image shape, and detection text length; An information category statistics module, which is used to obtain medical document samples, analyze and count document information categories based on the medical document samples, and the document information categories of the document samples include medical document date, patient name, drug name, and diagnosis information; an identification region identification module, the identification region identification module being configured to divide the document information category regions based on the document information category and corresponding to the location of the medical document sample, set region identification identifiers according to the document information category regions, and establish region logical associations based on the region identification identifiers, wherein the document information categories correspond to the document information category regions in a one-to-one manner; An extraction template establishment module, wherein the extraction template establishment module is used to establish a medical document extraction template based on the area identification mark and the area logical association; a document text information acquisition module, configured to identify the target medical document using the key detection parameters and the OCR tool, and acquire medical document text information; A document key information filling module is used to extract and generate medical document key information based on the medical document text information and the medical document extraction template, mark the medical document key information and fill it into the system accordingly; A sensitive information analysis module, configured to identify the key information of the medical document based on the key information of the medical document and the sensitive information database, analyze the key information of the medical document with sensitive keywords in the sensitive information database, and search for matches; A sensitive keyword fuzzy filling module is used to fuzzy process the sensitive keyword corresponding to the key information of the medical document if the matching item exists.

8. The data collection system based on intelligent identification and filling of medical documents according to claim 7 is characterized in that: The system further comprises: a cell segmentation module, the cell segmentation module being configured to detect and identify a document form in the medical document sample based on the medical document sample, and segment the document form into document cells; a text binary image generation module, the text binary image generation module being used to separate the text information in the document cells according to the contrast difference between the document cells and the medical document sample, and to generate a text binary image; The document information category extraction module is used to input the text binary image into the OCR recognition module for recognition to extract the document information category.

9. An electronic device comprising: processor; A memory, wherein computer program instructions are stored in the memory, and when the computer program instructions are executed by the processor, the processor executes the data collection method based on intelligent recognition and filling of medical documents as described in any one of claims 1 to 6.

10. A computer-readable storage medium having computer program instructions stored thereon, wherein when the computer program instructions are executed by a processor, the processor is caused to execute the data collection method based on intelligent recognition and filling of medical documents as claimed in any one of claims 1 to 6.

Images