KR20230019270A - A system, a method and a program for managing medical documents - Google Patents

Abstract

The present invention relates to a medical document management system, a management method and a program therefor, and more specifically, to a medical document management system and a management method for digitizing and managing medical documents. The present invention provides a medical document management system and a management method using the same, the medical document management system comprising: a document analysis unit for analyzing the formats of medical documents and extracting data using a machine learning algorithm; a data storage unit for storing the data extracted from the document analysis unit; and a digital document generation unit for generating digital documents corresponding to the medical documents using the extracted data and the analyzed formats.

Description

Medical data management system, management method and program for it {A SYSTEM, A METHOD AND A PROGRAM FOR MANAGING MEDICAL DOCUMENTS}

The present invention relates to a medical data management system, management method, and program therefor, and more particularly, to a medical data management system, management method, and program for digitizing and managing medical data.

With the development of information and communication technology, the introduction of computerized Hospital Information System (HIS) using EMR (Electronic Medical System), PACS (Picture Archiving and Communication System), OCS (Order Communication System), etc. has become active. are losing Therefore, various medical records generated by medical institutions are managed and distributed in the form of digital data, and are widely used for various purposes in the smart healthcare industry.

Recently, as digital medical imaging devices, computer input devices, etc. are widely used, since most medical documents are produced as digital data, it is easy to computerize and store and manage them. However, conventional medical data are often made in the form of paper, such as X-ray films and medical charts written by hand. Therefore, in the process of computerizing conventional medical data, there is a problem in that a manager has to individually input each data, and reliability is deteriorated due to an error occurring in the input process.

The present invention is to solve the above problems, and converts conventional medical data into digital data and manages it, but a medical data management system and management method that can easily convert into digital data and improve the reliability of the converted data. And to provide a program for this.

In order to achieve the above object, the present invention, using a machine learning algorithm, a data analysis unit for analyzing the format of medical data and extracting data, a data storage unit for storing the data extracted from the data analysis unit, the A medical data management system including a digital data generation unit generating digital data corresponding to the medical data using the extracted data and the analyzed format is provided.

Here, the extracted data includes a plurality of text data, the data analysis unit analyzes the correlation of the plurality of extracted text data using the machine learning algorithm, and the data storage unit analyzes the plurality of data It can be stored based on the analyzed correlation.

For example, the data analysis unit extracts the plurality of text data, divides and extracts at least one item data and value data corresponding to the item data, and the data storage unit extracts the item data and the item data corresponding to the item data value data can be related and stored.

The data analysis unit includes a first analysis module for distinguishing and recognizing an image area or a text area of the medical data, a second analysis module for extracting and recognizing image data or text data classified by the first analysis module, and the second analysis module. A third analysis module for analyzing a correlation between image data or text data extracted through the analysis module may be included.

And, further comprising a template generating unit for generating a template corresponding to the format of the medical data analyzed by the data analyzing unit, wherein the digital data generating unit places the extracted data on a template corresponding to the format of the medical data. to create digital material. Here, the template includes information about the location and size of text data or image data included in the medical data.

Meanwhile, a display unit configured to display the digital data and the medical data generated by the digital data generation unit to a user may be further included. The display unit may display the medical data and the digital data so that the user can compare and view the medical data and the digital data.

The digital data storage unit may further include an input unit for allowing a user to correct errors in the digital data displayed through the display unit, and the data storage unit may store previously stored data of the medical data as data corrected by the user.

Furthermore, a retraining data generation unit for generating retraining data using data modified by a user through the input unit is further included, and the data analysis unit's machine learning algorithm uses the retraining data generated by the retraining data generation unit. can be retrained.

Further, a template storage unit for storing a plurality of templates having different formats may be included, and the digital data generation unit may generate digital data according to a format of a selected template among the plurality of templates according to a user's request. do.

The above object of the present invention is to analyze the format and data of medical data using a machine learning algorithm, store the data analyzed from the medical data, and use the analyzed format and extracted data to analyze the medical data. It can also be achieved by a medical data management method that includes generating digital data corresponding to the data, and can also be achieved by a computer program that is stored in a medium configured to execute it or downloadable.

According to the present invention, it is possible to easily convert existing legacy data into digital data using a machine learning algorithm, and thus, data that was previously difficult to utilize can be used as data for providing smart healthcare services There are advantages to being

In addition, in the case of the present invention, digitized data is converted and provided in various forms, so the usability of the data is high, and the learning ability is strengthened by retraining the machine learning algorithm using the conversion error information corrected by the user By doing so, the reliability of the automatic conversion function can be improved.

1 is a diagram showing a medical data management system and peripheral devices according to an embodiment of the present invention;
2 is a block diagram specifically showing the configuration of a medical data management system according to this embodiment;
Figure 3 is a block diagram showing the sub-configuration of the data analysis unit of Figure 2;
4 is a diagram showing medical data and a template corresponding thereto;
5 is a diagram showing the appearance of digital materials created using different templates;
6 is a view showing the appearance of medical data and digital data displayed on the display unit;
7 is a flowchart illustrating a method for managing medical data according to the present embodiment.

Hereinafter, a medical data management system and management method according to an embodiment of the present invention will be described in detail with reference to the drawings. In the following description, the positional relationship of each component is explained based on drawings in principle. In the drawings, the structure of the invention may be simplified or exaggerated if necessary for convenience of description. However, the present invention is not limited thereto, and it is of course possible to implement by adding, changing or omitting various devices other than this.

1 is a diagram illustrating a medical data management system and peripheral devices according to an embodiment of the present invention.

The medical data management system 1 is a system for storing and managing medical documents. Here, the medical data refers to various data generated in a medical institution or related to medical services, and may include various data such as medical examination data, medical records, prescription data, examination data, and receipt data. Such medical data may be in the form of data including text, images, or both.

Medical data can be interpreted to include both digitized and non-digitized data. Digital data refers to data in which text or image information included in the data is composed of digital data or is recognizable. With the spread of digital medical devices and computer input devices, it is common to create and manage recent medical data in the form of digital data. On the other hand, past medical data, such as medical charts produced by legacy devices or written by hand, are stored and managed in a non-digitized form. Of course, in some cases, conventional paper-type medical data is scanned and managed in the form of an image file, but this is also non-digitized data in that text information or image information included in the scanned file is not recognized as individual digital data. corresponds to

The medical data management system according to this embodiment (1) is configured to convert and manage these non-digitized medical data into digital data. According to this system, conventional non-digitized medical data can be stored and managed as digital data, and thus it is possible to utilize conventional medical data for various smart healthcare services.

However, in the following description of this system, in order to be able to explain the conventional medical data based on the converted characteristics, the 'medical data' mentioned in the future can be interpreted in a meaning limited to the above-mentioned non-digitized medical data. there is.

As shown in FIG. 1 , the system 1 may be connected to various external devices, external servers, and external databases through communication interfaces (not shown). As an example, the system 1 may receive medical data to be stored and managed from an external device 2 such as a scan device, a test device, an EMR system, and an external database. And, this system 1 can convert the provided data into digital data and transmit it to an external system 3 such as an external server or database.

2 is a block diagram showing the configuration of a medical data management system according to this embodiment in detail. Hereinafter, with reference to FIG. 2, the medical data management system 1 according to the present embodiment will be described in more detail.

As shown in FIG. 2 , the present system 1 includes a server 10 and a user terminal 20 . The server 10 is a component that performs various calculations for converting and generating medical data into digital data and stores related data and data. The server 10 may use an in-hospital server of a medical institution, and may also use a cloud server in which computation and storage are performed in the cloud. Alternatively, it is also possible that some of the components constituting the server are implemented in the cloud. The user terminal 20 is configured to display medical data and digital data to the user and allow the user to input various information and commands, and various electronic devices including a display module such as a PC, netbook, tablet PC, and smartphone can be used. there is.

However, in FIG. 2, the server and the user terminal are configured as separate devices, but are not limited thereto, and the server of FIG. 2 is configured with a processor and memory provided in the user terminal, so that the server and the user terminal are implemented as one device. It is also possible to become

As shown in FIG. 2, the server 10 includes a medical data storage unit 110, a data analysis unit 120, a data storage unit 130, a digital data generation unit 140, and a template generation unit 150. , a template storage unit 160 and a retraining data generation unit 170. Each component is configured using a memory and a processor provided in the server 10, and may be designed in various ways to perform each operation described below.

First of all, the medical data storage unit 110 is configured to store medical data and may include at least one memory. Here, the medical data is data to be digitally converted by this system, and may be data in the form of image files obtained by scanning or photographing conventional paper data. Such medical data is provided from external devices and external systems such as a separate scanning device, imaging device, EMR server, and the like, and stored in the medical data storage unit 110 .

The data analysis unit 120 is a component for analyzing data stored in the medical data storage unit 110 and is configured using a processor and memory. The data analysis unit 120 analyzes the format of non-digitized medical data and extracts and analyzes data included in the medical data.

This data analysis unit 120 is configured to automatically analyze medical data using a machine learning algorithm. Machine learning, a field of artificial intelligence, is a technique of learning an algorithm using a large amount of data and performing analysis and prediction on given data using the learned algorithm. The data analysis unit 120 of this embodiment includes at least one algorithm learned by machine learning technology, and extracts and analyzes the format and included data of the medical data by using the algorithm. These algorithms can be designed in various ways using various deep learning algorithm models such as deep neural networks, convolutional neural networks, and recurrent neural networks according to the purpose of analysis, and the prepared learning data It is used in an already learned state using .

Figure 3 is a block diagram showing the sub-configuration of the data analysis unit of Figure 2; As an example, the data analysis unit 120 may include a first analysis module 121 , a second analysis module 122 , and a third analysis module 123 as shown in FIG. 3 . Each analysis module includes a separate machine learning algorithm, and based on this algorithm, medical data is sequentially analyzed.

First of all, the first analysis module 121 is a module that analyzes the format of medical data. Medical data has various formats depending on the type of data and the subject of production. For example, it may be data composed of only tables and text, or data including images and text, and the arrangement and size of text and images may also differ depending on the data.

Accordingly, the first analysis module 121 analyzes the format of medical data to be analyzed using a pre-trained algorithm. Through this, the text area, image area, table area, etc. of the corresponding medical data are separated, and the location and size of each area are analyzed. Furthermore, information on the size, font, and arrangement pattern of text characters of medical data can be analyzed. Such an algorithm may be learned in various ways, such as learning a plurality of data provided from various types of devices as learning data or learning using training data for identifying a data area according to a change in contrast ratio.

And, the second analysis module 122 is a module that extracts and recognizes data of each region based on the format information analyzed by the first analysis module. For example, in a text area of medical data, text data of the corresponding area is extracted and recognized using a text recognition algorithm. Then, the image area is recognized by extracting image data of the corresponding area using an image recognition algorithm.

The aforementioned text recognition algorithm may be an algorithm learned to improve text recognition rate in extracting text data and recognizing it. Also, the image recognition algorithm may be an algorithm learned to increase resolution and remove noise so as to acquire image data of good quality. It is possible to use text/image recognition algorithms used in other fields for this algorithm, but in this embodiment, an algorithm learned using learning data specialized for medical data is used. For example, the text recognition algorithm may be trained using terms frequently used in medical data or frequently used in the medical field as learning data. In addition, the image recognition algorithm may be trained to process key information of the medical image, such as lesion information, so as not to be damaged, or to easily remove noise factors mainly appearing in medical devices.

The second analysis module 122 may be provided with a single text recognition algorithm and an image recognition algorithm, respectively, but is configured to include a plurality of algorithms and to perform analysis by selecting an appropriate algorithm according to the type of medical data to be analyzed. It could be. For example, determining the category of medical data (eg, treatment subject, data type, device type from which the data was produced, etc.) based on the result of format analysis of the medical data, data size, data name, included image, etc. It is also possible to proceed with the analysis by selecting an algorithm suitable for the category.

And, the third analysis module 123 is a module that analyzes a correlation between a plurality of extracted text data, between a plurality of image data, or between text data and image data. For example, some of the plurality of texts extracted through the second analysis module are item data and other parts are value data for corresponding items, and a correlation exists between the extracted data. Alternatively, some of the extracted text data means a photographed region of the extracted image data, a photographic parameter, etc., and thus, there is a correlation between the extracted data.

Therefore, the third analysis module 123 uses an algorithm learned using various information, such as a positional relationship between data and a value data display method for each item, and analyzes the correlation between data extracted through this. Due to this, it is possible to grasp the comprehensive meaning of information included in medical data, rather than individually recognizing and managing the extracted data.

Meanwhile, although not shown in FIG. 3 , the data analysis unit 120 may additionally include a fourth analysis module. The fourth analysis module is a module that performs additional analysis not included in the medical data to be analyzed based on the data analyzed through the first to third analysis modules. For example, the fourth analysis module may be an analysis module that detects the location of the abnormal tissue through an automatic diagnosis algorithm based on the extracted image data. Alternatively, it may be an analysis module that analyzes and categorizes patient condition information based on the extracted examination information (eg, high blood pressure risk group). This fourth analysis module can be configured in various ways according to the purpose, and can perform analysis to derive necessary information by selecting a desired algorithm according to a user's selection and request among various algorithms provided in the system. As such, in the case of using the fourth analysis module, in the process of simply converting existing data into digital data, there is an advantage in that additional analysis desired by the user can be performed using the corresponding data.

The above-described data analysis unit 120 obtains information such as data format, text data, image data, and their correlations from medical data using various machine learning algorithms. However, in the above, although the data analysis unit including the first to fourth analysis modules has been described as an example, this is for convenience of description and is not limited thereto. In the above-described example, the object to be analyzed in each analysis module is not analyzed independently by a separate algorithm, but the algorithm may be configured in such a way that each analysis proceeds in parallel or is linked to each other to transform or repeat the order, , it is also possible to be configured to be analyzed through one integrated algorithm.

Again, referring to FIG. 2 , the data storage unit 130 is a component that stores various types of data obtained through the data analysis unit 120 described above. Specifically, the data stored in the data storage unit 130 includes text data and image data included in medical data to be analyzed. At this time, text data and image data may be stored including location information displayed in medical data, and may be stored in a data structure in which correlation between data is reflected (eg, item data among a plurality of text data and related information thereof). Value data is stored in relation to each other using data structures, etc.). In addition, information obtained through the analysis process of the data analysis unit (eg, the expected success rate of recognition of recognized text data, etc.) may also be stored in association with each data, and the analysis additionally obtained through the fourth analysis module, etc. Information may also be stored in the data store.

The data storage unit 130 includes a plurality of memories, and various types of information included in the analyzed medical data are stored in the form of digital data. Accordingly, it is possible to manage and utilize the information included in the medical data as digital data, and create corresponding digital data using this.

On the other hand, the template generation unit 150 is a component that generates a template corresponding to the format of the medical data to be analyzed. A template constitutes a format of digital material described below, and includes information such as the location and size of text data, image data, etc. displayed on the digital material. Hereinafter, with reference to FIG. 4, the template generated by the template generation unit will be described in detail.

Figure 4a shows an example of the medical data to be analyzed, and Figure 4b shows a template corresponding to the format of the medical data. When the above-described data analysis unit 120 analyzes the format of the medical data shown in a of FIG. 4, the analyzed information is provided to the template generation unit 150. Then, the template generation unit 150 creates a template of the same format as the medical data as shown in b of FIG. 4 based on this.

The created template may include information such as the location and size of a text area or image area, data items located in each area, size of text characters, font, display method, and the like. For example, in the case of b of FIG. 4 , an image data area (a) is located in the center of the template, text data areas (b1, b2) for patient information and shooting date are located above the image area, and below the image area contains information on the location of the text data areas (b3, b4, b5) for image information, shooting conditions, and material index information. Furthermore, information on the display method and arrangement of data in each area may be included. Therefore, when creating digital data, it is possible to create digital data in the same form as the medical data to be analyzed by displaying the data stored in the digital storage unit on the template according to the conditions included in the template.

The template generation unit 150 selectively performs a template generation step when a template corresponding to the medical data to be analyzed is not stored in the system, and the template storage unit 160 to be described later stores the corresponding template. In this case, the template creation step is omitted.

Again, referring to FIG. 2 , the template storage unit 160 is configured to store a plurality of templates having different formats and includes a memory. The stored template may be a template generated by the above-described template generation unit provided and stored cumulatively, or a template provided externally or produced by a user. The template storage unit 160 provides a template corresponding to the analyzed medical data or a template selected by the user to the digital data generator.

The digital data generation unit 140 is a component that generates digital data using the data extracted through the above steps and the analyzed format, and may be configured using a processor and a memory. The digital data generator 140 receives a template of the corresponding format from the template storage unit 160 and receives various data obtained from the corresponding medical data from the data storage unit 130 . Then, the digital material is created by arranging and displaying the corresponding data on the provided template (according to the policy included in the template).

With this configuration, digital data corresponding to the medical data to be analyzed are generated, and stored and managed as digital data. As described above, the information contained in such digital data can be recognized as each digital data. Therefore, unlike medical data before conversion, it is possible to modify some data or selectively extract and process only some data, and furthermore, it is possible to utilize the included data in various smart healthcare service industries.

Such digital data may be generated in a form having the same format as medical data and including the same information. Alternatively, additional information (eg, expected success rate of text recognition, auto-diagnostic findings of image data, etc.) obtained in the process of analyzing medical data may be further included. However, in addition to this, the digital data generation unit may also generate digital data in various formats different from the format of the medical data.

5 is a diagram showing the appearance of digital data created using different templates. As shown in FIG. 5 , when a user selects a desired one from among various types of templates stored in the template storage unit, the digital material generating unit 140 generates digital material using the selected template. In this case, digital data including the same contents as the medical data to be analyzed but having a different format from the medical data is generated (compare with a of FIG. 4). When using this, the same kind of medical data prepared in various formats (for example, examination records taken from examination equipment of different manufacturers, treatment charts prepared at different medical institutions, etc.) can be managed by unifying the form as digital data in the same format. There are advantages to being able to

Again, referring to FIG. 2, the user terminal 20 of this system includes a display unit 210 for displaying medical data or digital data and an input unit 220 for the user to input commands and information into the system. do. The user terminal 20 is connected to the server 10, receives various data stored in the server, and transmits various information and instructions input from the user to the server 10. The user can view the stored data through the user terminal 20, process or modify the data, or instruct various tasks related to management and distribution of the data.

As described above, the display unit 210 includes at least one display device, and the user can view digitally converted digital data through the display unit 210 . Here, the displayed digital data is data generated by the digital data generation unit, and as described above, may include images with improved resolution compared to non-digitized medical data, and additional information obtained in the data analysis process (eg For example, the location of an abnormal lesion in an image, a patient categorization result, an expected text recognition rate, etc.) may be included. In addition, it is also possible to view the information included in the analysis target in a format desired by the user (refer to the description of FIG. 5).

Furthermore, the display unit 210 of this embodiment is displayed so that it is possible to compare and view medical data and digital data corresponding thereto. Since digital data is automatically recognized and generated by the machine learning algorithm of the data analysis unit 120, recognition errors or error information may be included depending on the recognition accuracy of the algorithm. Therefore, when reading digital data, the user can compare and confirm the original medical data, thereby recognizing accurate information, and further correct errors or information errors included in the digital data.

6 is a diagram illustrating medical data and digital data displayed on a display unit. For example, the display unit divides one screen to display non-digitized medical data (scanned image) on one side (left screen in FIG. 6), and displays digital data generated by the digital data generator on the other side. (Right screen in Fig. 6). Accordingly, the user can view medical data and digital data while checking them at the same time. However, this is just an example, and various modifications such as displaying medical data and digital data alternately on the display unit by a screen change request using a toggle key or the like can be implemented. Alternatively, medical data (scanned images) and digital data form separate layers and are displayed overlapping each other on the display unit, and the user adjusts the transparency of each layer to compare and view the two overlapping data. It is also possible to do

Meanwhile, the user terminal 20 further includes an input unit 220 for inputting information or commands desired by the user. The input unit 220 may be composed of various input interface devices such as a keyboard and a mouse. As described above, the user can input evaluation information on the digital data through the input unit 220 while comparing and confirming the medical data and the digital data through the display unit 210 . In this case, the digital data may display information about an expected success rate of data recognition when analyzing the data, and data with a low expected success rate are highlighted on the digital data, so that the user can easily identify and evaluate the data.

For example, as a result of comparison and confirmation, when it is determined that the information of the digital data is correct, the user may input a confirmation opinion that the data information converted for the corresponding digital data is accurate through the input unit 220 . In this case, the confirmation findings on the corresponding digital material may be stored together in the data storage unit 130 and managed.

As another example, if it is confirmed that the digital material contains an error or error information as a result of comparison and confirmation, the fact that the digital material contains an error or error information can be input through the input unit 220 . In this case, the system 1 may instruct the data analysis unit to reanalyze the corresponding medical data, and newly create digital data based on the reanalyzed data.

Alternatively, if it is confirmed that the digital data contains an error or error information as a result of comparison and confirmation, the user directly inputs the correct data confirmed through the medical data onto the digital data displayed on the display unit 210 to perform correction and supplementation. . In this case, the corresponding digital material is modified by the data input by the user, and the modified data is transferred to the data storage unit so that the data for the corresponding material can be changed/updated to the modified data and stored/managed.

Furthermore, the information corrected by the user for the digital data in this way can be used as retraining data for retraining the machine learning algorithm of the present system. The presence of corrections in the converted digital data means that there is room for additional learning of the machine learning algorithm used by the data analysis unit 120 . Therefore, the retraining data generation unit 170 (see FIG. 2) of the server generates retraining data using data modified by the user through the input unit, and each analysis module of the data analysis unit 120 performs the added retraining data. The algorithm (eg, the algorithm of the second analysis module) may be retrained using the training data to improve recognition success rate and accuracy during digital conversion.

7 is a flowchart illustrating a method for managing medical data according to the present embodiment. Hereinafter, referring to FIG. 7, a management method using the medical data management system according to the present embodiment will be described in detail.

First, medical data to be analyzed are provided to the medical data storage unit of the system 1 . Medical data provided may be in the form of scanned files that are not digitized. Such medical data may be provided by an external device, an external server, or an external storage device.

In addition, when the medical data to be analyzed is determined by the user's selection, etc., the automatic analysis step for the medical data is performed using the machine learning algorithm of the present system.

First, a step of analyzing the format of medical data is performed (S10). This can be performed by the first analysis module 121 of the data analysis unit 120, and through this, various information related to the format of the data, such as the data area of the corresponding medical data and the data characteristics of the corresponding area, is analyzed.

In addition, a step of extracting and analyzing data of medical data is performed (S20). This step may be performed by the second analysis module 122 and the third analysis module 123 of the data analysis unit 120 . Specifically, the second analysis module 122 extracts and recognizes data of each region based on the analyzed format information, and the third analysis module 123 analyzes a correlation between extracted/recognized data. .

Through the above-mentioned steps, the present system can perform an overall analysis of the format of the medical data to be analyzed, the included information, and the correlation between the information. Here, each of the first to third analysis modules may include at least one machine learning algorithm, and specifically, one analysis module includes a plurality of usable algorithms and uses an algorithm selected in consideration of characteristics of data to analyze the data. analysis can be performed. However, the above steps S10 and S20 are not limited to sequentially proceeding, and may be performed in parallel or in a form in which each analysis step is organically related to each other.

On the other hand, when data is extracted through analysis of medical data, a step of storing the data on the corresponding data in the data storage unit 130 proceeds (S30). Through this step, various information included in non-digitized medical data can be stored and managed in the form of digital data.

Thereafter, a step of generating digital data corresponding to the analyzed medical data is performed (S40). This step may be performed by placing the data of the corresponding medical data stored in the data storage unit 130 on the same template as the format of the medical data.

In this step, a template for generating digital data may be selected from templates stored in the template storage unit 160 . However, if the template of the corresponding format is not stored in the template storage unit, the template generation unit 150 creates a corresponding template based on the format information of the data analyzed by the data analysis unit, and uses it to generate digital data. can do.

However, the digital data generated in this step may be generated to have a different format using a template desired by the user in addition to the format corresponding to the target of analysis.

The generated digital data is provided to the display unit of the terminal 20 and displayed to the user through the display unit 210 (S50). Therefore, the user can view the corresponding Mediker data in the form of digital data, and can process and edit it in various ways. At this time, the original medical data and the converted digital data may be displayed simultaneously or alternately on the display unit so that comparative reading is possible.

Then, in the process of checking medical data and digital data through the display unit, when an error or error information is found in the digital data, the user may proceed with a step of correcting some data of the digital data through the input unit 220 ( S60). The modified data is updated in the data storage unit, and the accuracy of the managed data can be improved.

Furthermore, the retraining data generation unit 170 may perform a step of generating retraining data using the data information modified by the user in the above step and retraining the algorithm of the analysis module using the same (S70). ). Through these steps, the reliability of the system can be improved by reducing the possibility of errors occurring during digital conversion through this system.

According to the medical data management system and management method according to the present embodiment described above, it is possible to automatically digitally convert non-digitized conventional medical data using a machine learning algorithm. In this case, it is possible to easily convert without individual input work, and the accuracy of digital data conversion can be improved by continuously re-learning the algorithm using the user's correction information to strengthen the analysis performance. In addition, by digitizing, storing, and managing conventional medical data, it is possible to process, edit, and use it in various ways according to user needs and purposes, and use such data for various medical services including medical treatment activities.

The medical data management method according to the present embodiment described above can be implemented in such a way that at least one processor included in the present system controls various components of the system to perform each step.

In addition, the medical data management method according to the present embodiment described above may be implemented as a program (including an application) to be combined/installed in an electronic device and executed. That is, it is possible to implement the aforementioned medical data management system and management method by installing and executing the program in an electronic device. Accordingly, the present invention can be implemented by such a program (a program stored in a medium or downloadable) or a medium storing such a program.

Such a program may include codes coded in a program language readable by a computing device of an electronic device so that the electronic device reads the program and executes the methods implemented as a program. These codes may include functional codes related to functions defining necessary functions for executing the methods, and include control codes related to execution procedures necessary for the terminal's computing device to execute the functions according to a predetermined procedure. can do.

In addition, a medium in which such a program is stored is not a medium that stores data for a short moment, such as a register, cache, or memory, but a medium that stores data semi-permanently and is readable by a device. Specifically, examples of the storage medium include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc., but are not limited thereto. For example, the program may be stored in various recording media on various web servers accessible by electronic devices. In addition, it should be noted that the medium may be distributed in computer systems connected by a network, and computer readable codes may be stored in a distributed manner.

Above, one embodiment of the present invention has been described in detail, but the present invention is not limited to the above embodiment. It is revealed that those skilled in the art to which the present invention pertains can carry out various modifications or changes to the present invention without departing from the scope of the technical features of the present invention defined in the appended claims. put

1: Medical data management system, 10: Server
20: user terminal 120: data analysis unit
140: digital data generation unit 150: template generation unit
160: template storage unit 210: display unit

Claims (17)

A data analysis unit that analyzes the format of medical data and extracts data using a machine learning algorithm;
a data storage unit for storing the data extracted from the data analysis unit;
A medical data management system comprising: a digital data generating unit generating digital data corresponding to the medical data using the extracted data and the analyzed format. According to claim 1,
The extracted data includes a plurality of text data,
The data analysis unit analyzes the correlation of the extracted plurality of text data using the machine learning algorithm,
The data storage unit stores the plurality of data based on the analyzed correlation. According to claim 1,
The data analysis unit extracts a plurality of text data, divides and extracts at least one item data and value data corresponding to the item data,
The data storage unit is a medical data management system that associates and stores the item data and value data corresponding to the item data. The method of claim 1, wherein the data analysis unit,
A first analysis module for distinguishing and recognizing an image area or a text area among the medical data;
a second analysis module for extracting and recognizing image data or text data classified by the first analysis module;
A medical data management system comprising a third analysis module for analyzing a correlation between image data or text data extracted through the second analysis module. According to claim 1,
Further comprising a template generation unit for generating a template corresponding to the format of the medical data analyzed by the data analysis unit,
The digital data generating unit generates digital data by arranging the extracted data on a template corresponding to the format of the medical data. According to claim 5,
The template includes information on the location and size of text data or image data included in the medical data. According to claim 1,
Further comprising a display unit configured to display the digital data and the medical data generated by the digital data generation unit to a user,
The display unit displays the medical data and the digital data so that the user can compare and view the medical data and the digital data. According to claim 7,
Further comprising an input unit for a user to correct errors in the digital data displayed through the display unit;
The data storage unit stores the previously stored data of the medical data as data modified by the user. According to claim 8,
Further comprising a retraining data generation unit for generating retraining data using data modified by the user through the input unit;
A medical data management system in which the machine learning algorithm of the data analysis unit is retrained using the retraining data generated by the retraining data generation unit. According to claim 1,
Further comprising a template storage unit for storing a plurality of templates having different formats,
The digital data generation unit generates digital data according to a format of a selected template among the plurality of templates according to a user's request. Analyzing the format and data of medical data using a machine learning algorithm;
Storing data analyzed from the medical data;
A medical data management method comprising the step of generating digital data corresponding to the medical data using the analyzed format and the extracted data. According to claim 11,
Analyzing the data extracts a plurality of text data included in the medical data and analyzes the correlation between the extracted text data,
The step of storing the data is a medical data management method of storing the extracted plurality of text data based on the analyzed correlation.

The method of claim 12, wherein analyzing the data comprises:
distinguishing and recognizing an image data area or a text data area of the medical data;
extracting and recognizing image data or text data of the divided data area; and
A method for managing medical data comprising the step of analyzing a correlation between the extracted image data or text data. According to claim 11,
Further comprising generating a template corresponding to the format of the analyzed medical data,
The step of generating the digital data is a medical data management method of generating digital data by arranging data of the analyzed medical data on the generated template. According to claim 11,
Further comprising displaying the medical data and the generated digital data to a user through a display unit,
The display unit is configured to enable the user to compare and view the medical data and the digital data. According to claim 15,
receiving correction data for errors in the digital data from a user through an input unit; and
The method of managing medical data further comprising retraining the machine learning algorithm by using the input correction data as retraining data. A computer program stored in a medium or downloadable to execute the method of claim 11.

Images