JP2006079189A - Receipt file creation system, medical chart file creation system and file creation system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To associate a plurality of coexisting dialects of the medical industry with words of the standard language. <P>SOLUTION: A receipt file creation system 3 includes a master table storage part 70 for storing one of a plurality of synonyms in a master table as one standard word, and a dictionary table storage part 72 for storing the other synonyms, i.e., dialects, in a dictionary table in association with the one standard word. A matching process part 52 searches the dictionary table for text data. When the text data is found to be stored in the dictionary table, a receipt file creating part 54 creates a receipt file in which standard words for the text data are associated with the text data. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a technique for generating a file.

In a medical institution such as a hospital, a doctor fills in a medical record with the name of a medical condition such as injury and illness, medication, injection, examination, and surgery (hereinafter referred to as “medical treatment content”). In recent years, many medical institutions have introduced a processing device called a “receipt computer” (also referred to as a “medical computer”), and a person in charge of the medical institution inputs the contents of medical treatment into the reception computer based on the medical record. A medical remuneration statement (hereinafter referred to as “receipt”) is created in a format stored in the computer. In addition to hospitals, reception computers are spreading in pharmacies. The medical institution submits the receipt together with a medical fee bill (hereinafter referred to as “bill”) to the social insurance medical fee payment fund of each prefecture. The payment fund checks whether there are any errors in the number of claims for medication, injection, surgery, etc., and the review committee reviews the bills and receipts checked by the payment fund. The amount of medical treatment fee is determined based on the receipt after such examination. Insurers, such as health insurance associations, conduct a secondary review of the reviews that have been reviewed, and pay medical fees to medical institutions through payment funds, etc., while those with doubts about medical treatment are reexamined with the review committee To charge.
Japanese Patent No. 3139485

  In recent years, there has been a budget deficit problem, and insurers have strengthened the secondary examination. Incorrect claims may be made due to mistakes in medical institutions, etc., so the aim is to avoid excessive payment of medical fees by finding them through a secondary review and submitting them to a reexamination. However, on the other hand, since medical expertise is required for the inspection of the receipt, a reexamination request may be made based on the misunderstanding of the insurer. Recently, as part of the deregulation, it became possible for the health insurance union to conduct the screening and payment of receipts directly to medical institutions. Since insurers and medical institutions that have conflicting interests will negotiate directly, there may be cases where unnecessary troubles occur due to any misunderstanding.

  In general, receipts are often delivered on a paper basis. The receipt expresses a medical practice performed on the patient, and the information on the patient's receipt over a plurality of months or even a plurality of years expresses the chronological injury history of the patient. If the insurer can summarize the injury history of individual patients, the injury history can be analyzed and used for health management of individual patients. By making paper-based receipts into text, it is possible to efficiently manage the patient's injury and illness history as data, and also to check whether the patient's ongoing medical practice is appropriate .

  Japan has created the “Disease, Injury and Death Cause Classification” in accordance with ICD-10 for the purpose of statistical surveys based on statistical methods and management of medical records as medical classifications at medical institutions. . The terms shown in the “Disease, Injury and Death Cause Classification” are preferably used in the medical industry in Japan for the names of wounds and diseases, and all medical records or receipts use the term “Disease, Injury and Death Cause Classification”. Thus, it is possible to efficiently perform originally intended statistical surveys, medical record management, and the like.

  However, in the medical industry, it is customary that a single injury or illness or medicine is expressed by a plurality of names. The same applies to the name of the test and the dosage unit of the drug. Regarding the name of injury and illness, when the term based on the above `` disease, injury and death classification '' is called `` standard language '' and other terms of consent are called `` dialect '', the use of the dialect is nothing special, Some doctors may not be aware that the terminology used is a dialect. Currently, there is an industry trend to standardize dialects and use standard words, but that consciousness should not permeate all doctors, and use familiar dialects for each doctor or for each reception computer. There is a current situation that continues. In addition, there are circumstances in which dialects are used depending on the type of medical institution. For example, in a small clinic or specialized hospital, using a standard language forcibly may make management work inefficient.

On the other hand, from the standpoint of reviewing the receipt, if the terms are not unified, the names of wounds and medicines cannot be easily understood, and efficient screening is hindered. For this reason, only the person who has the skill to understand all dialects can check the receipt by simply converting the contents of the paper receipt into text. In addition, when a patient visits multiple medical institutions, if the terminology is not uniform for each medical institution, it is not easy to extract information useful for individual patient health management from the text data of the receipt, It is not possible to perform statistical processing using all patient data.
Therefore, an object of the present invention is to provide a technique for associating a standard word with a synonym existing as a dialect and to generate a highly convenient data file.

  In order to solve the above problems, an aspect of the present invention relates to a system for generating a receipt file. The receipt file generation system of this aspect includes a master table storage unit that stores a master table that associates medical related terms that are standard words with their identification codes and attribute information, and has the same meaning as standard terms as medical related terms. A dictionary table storage unit that stores a dictionary table in which synonyms to be represented are associated with identification codes of the standard words, a text data processing unit that extracts text data included in a data file in which a receipt is converted into text, and When the text data is searched for in the dictionary table and the matching processing unit finds that the text data is stored as a medical term in the dictionary table, the text data is based on the identification code. That generates a receipt file that associates a master table with a standard word And a script file generation unit. The medical-related term is a term (name) related to medical treatment, and includes not only a term representing medical practice but also a drug name and the like, and particularly means a term used for notation of a receipt. Specifically, medical-related terms include terms such as injury and illness name, drug name, drug dosage unit, medical material name, and medical practice name. Synonyms representing the same meaning as standard words include synonyms (dialects) other than standard words. The synonym may include the standard word itself. In this case, the standard word is also included in the dictionary table, so the matching processing unit searches for the standard word by searching the dictionary table. Can do. When the standard word is not included in the dictionary table, the matching processor can search for the standard word by searching the master table.

  According to this receipt file generation system, it is possible to generate a receipt file in which synonyms other than standard words are associated with standard words. By using the receipt file generated by this system, various data processing such as statistical processing and personal health management can be performed using the standard word as a key.

  When the text data is not stored as medical terms in the dictionary table, the matching processing unit stores the text data as unknown data in a predetermined storage area. Once stored as unknown data, the data can be used later for expansion of the master table or dictionary table.

  The receipt file generation system further includes a temporary dictionary table storage unit that stores a temporary dictionary table in which character strings that are not stored in the dictionary table and are not normally used are associated with identification codes of standard words. The generation unit generates a receipt file in which character string text data is associated with a standard word in the master table based on the temporary dictionary table. Examples of character strings that are not stored in the dictionary table and are not normally used include unusual expressions that are not generally used, such as expressions that are entered incorrectly or slang expressions that are used only by specific medical institutions. Applicable. Even if it is such a character string, while maintaining the content of the original receipt by generating a receipt file associated with the standard word stored in the master table while saving the character string itself, A highly useful receipt file can be generated.

  The receipt file generation unit generates a receipt file by using the correspondence stored in the temporary dictionary table for text data of character strings that are not stored in the dictionary table and are not normally used for a predetermined period. Also good. By using it only for a predetermined period, it is possible to maintain the usefulness of the receipt file even when the same medical terms as the character string appear later.

  Each of the master table storage unit and the dictionary table storage unit stores the master table and the dictionary table for each type of medical related term. For example, by creating a master table and a dictionary table for each type of medical-related terms such as names of wounds and medicines, it is possible to easily update, manage, and maintain each table.

  The receipt file generation system further includes a distribution table storage unit that stores a distribution table in which a character string is associated with a type of dictionary table to be referred to, and the matching processing unit extracts the character string by referring to the distribution table. A dictionary table to be referenced with respect to the text data thus set may be specified. The matching processing unit can shorten the time required to specify the dictionary table by referring to the distribution table.

  The receipt file generation system further includes a decomposition table storage unit that stores a character string and a decomposition table associated with a plurality of medical terms that are decomposed by dividing the character string, and the text data processing unit refers to the decomposition table. Then, text data obtained by decomposing the text data of the character string for each medical term may be extracted. By using the disassembly table, the text data processing unit can efficiently extract medical terms from a character string.

  The receipt file generation system further includes a combined table storage unit that stores a combined table that lists medical related terms generated by combining a plurality of character strings, and the text data processing unit refers to the combined table, Text data obtained by combining text data of a plurality of character strings may be extracted. By using a combination table, for example, one medical related term with a blank inserted in between, or one medical related term inserted into a new line and described across multiple levels It can be efficiently extracted as medical terms.

  The matching processing unit determines the matching between the extracted text data and the medical-related terms stored in the master table or the dictionary table, and when the unmatched portion is a quantity expression, the quantity expression is Alternatively, it may be determined that it should be associated with the quantity expression included in the medical terms stored in the dictionary table. Even if there is no expression that the text data shows in the dictionary table, the matching processing unit autonomously links the text data with the registered medical terms to reduce the burden of human processing be able to.

  Another aspect of the present invention relates to a system for generating a medical record file. The medical record file generation system according to this aspect includes a master table storage unit that stores a master table that associates medical related terms that are standard words with their identification codes and attribute information, and has the same meaning as standard words as medical related terms. A dictionary table storage unit that stores a dictionary table in which synonyms to represent are associated with identification codes of the standard words, a text data processing unit that extracts text data included in a data file of the electronic medical record, and extracted text data If the matching processing unit that searches the dictionary table and the matching processing unit find that the text data is stored as medical terms in the dictionary table, the text data is mastered based on the identification code. A file generator that generates a chart file associated with the standard words of the table Obtain. The medical-related term is a term (name) related to medical treatment, and includes not only a term representing medical practice but also a pharmaceutical name, and means a term particularly used for notation of medical records. Specifically, medical-related terms include terms such as injury and illness name, drug name, drug dosage unit, medical material name, and medical practice name. Synonyms representing the same meaning as standard words include synonyms (dialects) other than standard words. The synonym may include the standard word itself. In this case, the standard word is also included in the dictionary table, so the matching processing unit searches for the standard word by searching the dictionary table. Can do. When the standard word is not included in the dictionary table, the matching processor can search for the standard word by searching the master table.

  According to this chart file generation system, a chart file in which synonyms other than standard words are associated with standard words can be generated. By generating a medical record file associated with a standard word, it is possible to substantially unify medical related terms in medical records of a plurality of hospitals.

  Still another aspect of the present invention provides a master table storage unit that stores a master table that associates a term that becomes a standard word with its identification code and attribute information, a standard word, and a synonym that represents the same meaning as the standard word. A dictionary table storage unit for storing a dictionary table in which words are associated with identification codes of the standard words, a text data processing unit for extracting text data included in the data file, and the extracted text data in the dictionary table When the matching processing unit that searches for and the matching processing unit finds that the text data is stored in the dictionary table, the text data is associated with the standard word of the master table based on the identification code. A file generation unit for generating a data file.

  According to this file generation system, a file in which a synonym (dialect) other than a standard word is associated with a standard word can be generated. Therefore, various data processing using the file is performed using the standard word as a key. It becomes possible.

  It should be noted that any combination of the above-described constituent elements and a conversion of the expression of the present invention between a method, an apparatus, a system, a recording medium, a computer program, etc. are also effective as an aspect of the present invention.

  According to the present invention, a file in which a standard word and a synonym other than the standard word are associated can be generated.

  FIG. 1 shows a reception processing flow in an embodiment of the present invention. This receipt process flow may be executed by one main body or may be executed in cooperation by a plurality of main bodies. Usually, it is considered that there are many cases in which a single role processing system is realized as a whole by performing a clear division of roles among a plurality of subjects. In the receipt processing flow shown in FIG. 1, a paper receipt provided from a medical institution or a payment fund is converted into data, and a procedure for efficiently generating a receipt file suitable for data processing such as statistical processing is expressed. Each step shown below may involve human intervention, but most are executed by computer processing by the system.

  The receipt processing flow of the present embodiment is as follows: sorting operation (S10), paper receipt imaging (S12), OCR processing (S14), punching processing (S16), logic check processing (S18), correspondence check processing (S20), Text data decomposition process (S22), text data combination process (S23), decomposition table, update of combined table (S24), reading of dictionary table (S26), matching process (S28), update of master table and dictionary table (S30) ), Temporary dictionary table generation (S32), receipt file generation (S34), and data processing (S36).

(1) Sorting work In the sorting work (S10), each form printed on a paper medium such as an in-hospital receipt, an out-of-hospital receipt, or a dispensing receipt is sorted. This sorting operation is generally performed manually, but may be automatically executed by computer processing at the time of subsequent imaging in S12.

(2) Imaging of paper receipt In imaging of paper receipt (S12), the sorted paper receipt is read by a scanner and converted into image data. Here, while continuously inputting paper receipts into the scanner, the image being input is displayed in real time on the display, and the operator confirms the scanning status. When a paper receipt is input on the back side, scanning is automatically stopped by the image recognition process, and the operator is notified accordingly. Also, when the paper direction is upside down, the operator is notified of this. As described above, the image data of the paper receipt having the same orientation is generated. By adding not only the image recognition function but also the receipt sorting function to the scanner CPU, the paper receipt is imaged and the sorting process is performed at the same time, and the image data is stored in a folder for each type of receipt. Store it.

(3) OCR process In the OCR process (S14), characters are read from the image data of the receipt by optical character recognition. Personal information such as the insured symbol number and patient's name is listed at the top of the receipt, and below it is the name of the wound, medical information such as medications and injections, the drugs used and their usage. The medical details to be shown are described. In the present specification, information other than personal information of the receipt is referred to as drug history information. The medical history information of a receipt should be protected secretly by being associated with a specific individual by the personal information of the receipt, and handling thereof requires sufficient care. Therefore, as a pre-stage of OCR processing, first, the personal information image and the medicine history image are separated.

  First, the personal information image of the receipt is converted into text by OCR processing and converted into a unique code by encryption technology. Subsequently, this unique code is combined with the image data of the medical history image of the receipt. As a result, the operator cannot identify an individual even when viewing the combined data, and can associate the encrypted personal information with the medical treatment details information. Next, the medicine history image is converted into text by OCR processing. The text data is classified into items (fields) in a predetermined format and is filed. In this OCR process, it is preferable to realize a high character recognition rate.

(4) Punching process In the punching process (S16), the puncher inputs or corrects a part that cannot be converted to text by the OCR process or a part that has been erroneously converted to text. As the accuracy of the OCR process increases, the work amount of the puncher decreases, so that it is possible to reduce the total cost for converting the paper receipt into data.

  Steps S10 to S16 are executed by the text data generation system. The text data of the paper receipt is collected as a file, and this data file is delivered to a receipt file generation system that executes the processing after S18.

(5) Logic Check Process In the logic check process (S18), the logic check of the text data is performed. The computer performs a logic check and detects a logical error during reception. A logical error is an error in which the patient's birthday is a future date, for example.

(6) Correspondence Check Processing In the correspondence check processing (S20), the correspondence between text data in the data file is checked. Here, for example, check for errors such as the reverse relationship between the medical treatment start date and the second and subsequent medical treatment dates, or the fact that the drug name is present in the summary column of the drug history but the usage amount does not exist . Also, check for errors such as characters being entered in the item where the number should be entered. The computer retains the correspondence relationship between the data in advance, and determines whether or not the correspondence relationship can be matched, thereby detecting a correspondence relationship error. The detected error is notified to the operator. The operator looks at the error content and corrects it to the correct content. The computer may check not only the correspondence between the data but also the correspondence between the receipts. For example, if there is a dispensing receipt but no medical receipt, the computer notifies the operator to that effect.

(7) Text Data Decomposition Processing In the text data decomposition processing (S22), text strings converted into text are classified and classified. For example, if the medical receipt summary column data is classified into pharmaceuticals, medical materials, and medical practices, and multiple items of pharmaceuticals, medical materials, and medical practices exist as a single line of text data, they are classified by item. Disassembled into Specifically, commas, blanks (blanks), line breaks, and the like included in the character string are detected, and the character string is divided into medical terms. Further, the character string may be decomposed into medical terms by referring to a continuous character string and a decomposition table in which the character strings are associated with a plurality of medical terms. For example, when the injury and illness name and the medicine name are converted into text as a continuous character string, the decomposition table stores the character string in association with the injury and illness name and the medicine name.

(8) Text data combining process In the text data combining process (S23), character strings that have not been extracted as medical terms are combined among the character strings decomposed by blanks or line breaks in the text data decomposition process (S22). Then, one medical related term is extracted. At this time, a combined table in which medical related terms generated by combining a plurality of character strings are listed is referred to. For example, if a single drug name is made into text with a space in between, the combination table holds the drug name and stores it, and combines a plurality of decomposed character strings. If the drug names stored in the combination table match, the plurality of character strings are combined to extract one drug name.

(9) Update of disassembly table and join table In the update process of disassembly table and join table (S24), the text data that could not be decomposed in the text data decomposition process of S22 is divided into medical terms, Expansion is performed, and a plurality of text data that could not be combined in the text data combination processing of S23 is combined as one medical related term to expand the combination table. This work is performed manually.

  In S22, the operator disassembles the character string that cannot be classified into a plurality of medical terms, and replenishes the content of the disassembly table as appropriate. In particular, a receipt computer used in a large-scale medical institution may be customized with unique specifications. For this reason, depending on the receipt, a plurality of drug names may be entered in succession, or a wound name and a drug name may be entered in succession. When the operator finds such a character string, the operator categorizes each corresponding medical term and increases the registration contents of the disassembly table. As a result, the reliability of the text data decomposition process to be executed next time can be reliably increased as compared to the previous time, and the processing time can be shortened.

  Further, the operator appropriately registers medical related terms that could not be combined in S23 in the combination table. The number of characters in medical terms varies, but especially for medical terms that are long strings, a space is often inserted between them or a line break is entered in the summary column. Basically, the text data is decomposed at the blank portion and the line feed portion at S22, but in this decomposition processing, although it is originally one medical related term, it is unnecessarily decomposed to one medical related term. It may lead to results that cannot be specified as terms. As a countermeasure for such a case, the operator registers one medical-related term to be decomposed in the combination table, and increases the accuracy of the combination process in S23. By appropriately updating the join table, the reliability of the text data extraction process can be improved compared to the previous time, and the processing time can be shortened.

(10) Reading Dictionary Table In the dictionary table reading process (S26), the dictionary table stored in the storage unit is read. There may be a plurality of expressions representing the same meaning in medical-related terms. For example, the names of appendices “appendicitis”, “cecum”, and “appe” all mean the same injury. Terminology inconsistency becomes a hindrance to subsequent statistical processing, etc., so in the Receipt File Generation System, one of multiple synonyms is designated as “Standard Word” in order to effectively use the receipt data. ”And synonyms other than the standard language are set as“ dialect ”and handled. The standard word is stored in the master table in association with the identification code and attribute information. The relationship between the master table and the dictionary table is shown below.

  The master table is generated by associating medical-related terms such as injury and illness names and drug names adopted as standard words in the system with identification codes of the medical-related terms and attribute information thereof. For example, as for the names of wounds and sicknesses, the names of wounds and sicknesses classified as “classification of disease, injury and death” may be set as standard words. The identification code is used for associating the master table with the dictionary table, and an identification code uniquely set in the receipt file generation system may be used. Also, identification codes in various code systems existing in the medical industry may be diverted as identification codes for linking the master table and the dictionary table.

  Specifically, when the wound name “appendicitis” is set as a standard word, the master table stores “appendicitis” in association with its identification code and its attribute information. The attribute information includes information such as ICD classification of appendicitis. Similarly, the standard word, the identification code, and the attribute information are associated with the master table of the drug name and other categories. The drug name attribute information includes a drug price (insurance score). The master table is created for a plurality of sections, such as a sickness and disease master table that standardizes wound names and a pharmaceutical product master table that standardizes medicines.

  The dictionary table is generated by associating a synonym representing the same meaning as a standard word as a medical-related term with an identification code of the standard word. The synonym includes a synonym (dialect) other than the standard word, and may include the standard word itself. Specifically, the dictionary table stores “appendicitis”, “cecum”, and “upe” in association with the identification code of “appendicitis”. Here, appendicitis is a standard language, and the cecum and appe are dialects. The dictionary table is created for a plurality of sections such as a disease / disease dictionary table in which the names of wounds and diseases are converted into a dictionary and a drug dictionary table in which drugs are converted into a dictionary corresponding to the master table.

(11) Matching process In the matching process (S28), matching is performed between the text data decomposed for each item and the data of the read dictionary table. When the text data matches the data registered in the dictionary table, the identification code associated with the data is read and transferred to the subsequent receipt file generation process.

  If another name other than the cecum and appe that expresses appendicitis is described as text data, the new name is not registered in the dictionary table, and the computer cannot recognize the name as a dialectitis of appendicitis. If there is no name corresponding to the dictionary table, the name is stored as unknown data in a predetermined storage area, and a message to that effect is output.

(12) Master Table / Dictionary Table Update In updating the master table / dictionary table (S30), first, the operator confirms the name stored in the predetermined storage area as unknown data. This confirmation operation may be performed each time unknown data is generated, or may be performed when a plurality of unknown data are collected. When the operator determines that the unknown data is a new name for appendicitis, the operator adds the name as an appendicitis dialect and updates the dictionary table. In addition, when a new medicine comes out or when a new injury occurs, the operator sets an identification code for a new medical term and updates the master table and the dictionary table.

(13) Temporary Dictionary Table Generation In the temporary dictionary table generation process (S32), a temporary dictionary table is generated. When the text data is erroneously written or is a special expression used only in a certain medical institution, the operator generates a temporary dictionary table that associates the character string indicated by the text data with the standard word identification code. . The temporary dictionary table is used only for the current month, for example.

(14) Receipt file generation In the receipt file generation process (S34), a receipt file in which dialects and special expressions in the text data are linked to standard words is generated. Specifically, in the receipt file, an identification code of a standard word used in the master table is linked to the expression in the text data. This receipt file is made up of the contents entered in the original paper receipt as it is, and the dialects and special expressions contained therein are configured in correspondence with standard words by referring to the dictionary table.

(15) Data processing In the data processing process (S36), various processes according to the user's needs such as statistical processing and predictive medicine are executed based on the generated receipt file. This is possible because medical terms in the receipt file are linked to standard words, and various data processing is possible using the standard word identification code as a key. In addition, since each receipt file is associated with an individual patient, it is possible to grasp the patient's injury history and predict medical care. Steps S18 to S36 are executed by the receipt file generation system.

  FIG. 2 shows a receipt processing system 1 according to an embodiment of the present invention. The receipt processing system 1 includes a text data generation system 2 that reads text from image data of a paper receipt and generates text data, and a receipt that associates dialects and the like with standard words from the text data generated in the text data generation system 2. A receipt file generation system 3 for generating a file is provided. The text data generation system 2 executes processing steps from the paper receipt imaging process (S12) to the punching process (S16) in FIG. The receipt file generation system 3 executes processing steps from the logic check process (S18) to the data processing process (S36). The text data generation system 2 and the receipt file generation system 3 may be managed and operated by the same entity, or may be jointly operated by another entity in cooperation.

  The functions of the receipt processing system 1 in this embodiment are realized by the CPU, the memory, a program loaded in the memory, and the like in the text data generation system 2 and the receipt file generation system 3. The program may be built in the text data generation system 2 and the receipt file generation system 3, or may be supplied from the outside in a form stored in a recording medium. Accordingly, those skilled in the art will understand that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof.

  FIG. 3 shows a configuration of the text data generation system 2. The text data generation system 2 includes an image data generation unit 10, an OCR unit 12, an input unit 16, a display 18, a drive device 20, and a storage unit 22. The input unit 16 is an input interface of the text data generation system 2 and includes, for example, a keyboard and a pointing device such as a mouse. The input unit 16 may be configured as a touch panel provided on the display 18, for example. The drive device 20 is a device that performs data writing and / or data reading from a storage medium 30 such as a DVD or a CD.

  The image data generation unit 10 is a scanner, and receives paper receipt and generates image data (receipt image) of the paper receipt. The image data is divided into a personal information image in which personal information is entered and a medicine history image obtained by removing the personal information image from all images. The personal information image and the medicine history image may be divided by the operator specifying the range of the receipt image using the pointing device of the input unit 16 while looking at the display 18.

  FIG. 4 shows image data of a medical fee remuneration statement (receipt). This receipt is an example of a medical receipt. The receipt image 32 is divided into a personal information image 34 above the one-dot chain line 33 and a medicine history image 36 below. The medicine history image 36 includes a summary column 38 in which medicines, medical materials, medical treatments, and the like are described.

  In the summary column 38, the number at the left end indicates the medical treatment division code, and the asterisk indicates simultaneous prescription. A group of pharmaceuticals, medical materials, and medical practices that are blocked by one asterisk means that they are simultaneously prescribed pharmaceuticals, medical materials, and medical practices. In terms of insurance points, simply add each of them. Rather than become a subject of reduction. The number on the right end indicates the number of times, and the number on the left side of “x” on the left side indicates the score. The insurance score of the block is determined by “number of points × number of times”. Among the character strings on the left side of the score, the character strings that are not included in the simultaneous prescription column or the medical treatment classification code indicate the drug name, the medical material name, and the medical practice name.

  Returning to FIG. 3, the OCR unit 12 first reads characters from the personal information image 34 by optical character recognition. The reading result of the personal information image 34 is encrypted and associated with the medicine history image 36.

  Next, the OCR unit 12 reads characters from the medicine history image 36 by optical character recognition. In the medicine history image 36, the summary column 38 indicates insurance points such as used medicines and medical treatments, and is directly related to the payment amount. Therefore, the OCR unit 12 needs to be able to read the summary column 38 with high accuracy.

  The text data of the medicine history image 36 read by the OCR unit 12 is displayed on the display 18, and the operator (puncher) corrects the displayed text data. The higher the character recognition rate of the OCR unit 12, the lighter the operator's workload, and the lower the cost of making paper receipt text.

  FIG. 5 shows a display example of a data file generated by recognizing the summary column 38 and classifying it for each item. The puncher corrects the data file by operating the input unit 16 while comparing the data file displayed on the display 18 with the image in the original summary field 38. FIG. 5 shows an example of successful character recognition. However, if there is a misrecognition of characters, the puncher corrects it appropriately.

  The data file that has been corrected by the puncher is stored in the storage unit 22. The drive device 20 records the receipt data file in the storage medium 30. The storage medium 30 is delivered to the receipt file generation system 3.

  FIG. 6 shows the configuration of the receipt file generation system 3. The receipt file generation system 3 includes a text data processing unit 50, a matching processing unit 52, a receipt file generation unit 54, a logic check processing unit 55, a correspondence check processing unit 56, a data processing unit 58, an input unit 60, a display 62, and a reference table. A storage unit 64, a data storage unit 66, and a drive device 68 are provided. The input unit 60 is an input interface of the receipt file generation system 3 and includes, for example, a keyboard and a pointing device such as a mouse. The input unit 60 may be configured as a touch panel provided on the display 18, for example. The drive device 68 is a device that performs data writing and / or data reading of the storage medium 30 such as a DVD or a CD. When the text data generation system 2 and the receipt file generation system 3 are configured as an integrated reception processing system 1, the configurations of an input unit, a display, a drive device, a storage unit, and the like may be shared. The reference table storage unit 64 includes a master table storage unit 70, a dictionary table storage unit 72, a temporary dictionary table storage unit 74, a decomposition table storage unit 76, a combined table storage unit 77, and a distribution table storage unit 78. .

  The storage medium 30 in which the receipt data file is recorded is inserted into the drive device 68, and the data file is written into the data storage unit 66. The logic check processing unit 55 reads the data file from the data storage unit 66 and performs a logic check on the text data. For example, when the patient's birthday is a future date, the logic check processing unit 55 detects the error and notifies the operator. The operator receives the notification and corrects the error part.

  The correspondence check processing unit 56 checks the correspondence between the text data. Referring to FIG. 5, the medical material name “image recording film (1) one half cut”, the score (2238), and the number of times (1) are data that should be associated with each other. It exists on a different line in the file. Therefore, the correspondence check processing unit 56 corrects them so that they are on the same line. The same applies to “ultrasound examination (tomography) / chest and abdomen”, its score (550), and number of times (1).

  In order to execute such a check, the correspondence check processing unit 56 holds a correspondence relationship between data items in advance. This correspondence relationship may be stored in the data storage unit 66. For example, in the data in the summary column shown in FIG. 5, the data items in the “same” column, “medicine name etc.” column, “score” column, and “number of times” column must be associated with each other. If the correspondence relationship is not satisfied, the correspondence check processing unit 56 detects a data error. The detected error is notified to the operator through the display 62. The operator looks at the error content and corrects the data to the correct correspondence. Furthermore, as described above, the correspondence check processing unit 56 has a reverse relationship between the medical treatment start date and the second and subsequent medical treatment dates, or characters are entered in items to which numbers are to be input. Check for errors. By performing the above logic check and correspondence check, errors contained in the data file can be removed.

  The text data processing unit 50 classifies and classifies the text data included in the data file, and extracts the text data that has been decomposed up to the data items of the receipt file to be finally generated. For example, if the medical receipt summary column data is classified into pharmaceuticals, medical materials, and medical practices, and multiple items of pharmaceuticals, medical materials, and medical practices exist as a single line of text data, they are Disassemble each item. In addition, when the contents written as a medicine indicate the name and amount of the medicine, a process for decomposing them is performed. For example, “Luplac tablets 4 mg, 1 tablet” shown in FIG. 5 is broken down into Luplac tablets 4 mg (pharmaceutical name) and 1 tablet (amount). First, the text data processing unit 50 detects a comma or a blank (blank) included in the character string, and classifies the character string for each medical term. At this time, the character string may be divided based on the medical treatment division code.

  In this embodiment, the text data processing unit 50 may decompose the text data of the character string by referring to the decomposition table, and extract the text data of the medical terms from the character string. The disassembly table is stored in the disassembly table storage unit 76, and a continuous character string is associated with a plurality of medical terms that are decomposed by dividing the character string. This association is created in advance by the operator as described with reference to S24 in FIG. 1, and when a new character string in which a plurality of medical terms are connected appears, the operator selects the character string. Is decomposed into medical terms and a disassembly table is created sequentially. As a result, when the same character string appears at a later timing, the text data processing unit 50 can decompose the character string into appropriate medical terms by referring to the decomposition table. The decomposed data is recorded in a predetermined file.

  This file is recorded without damaging the relationship of one block specified by the asterisk. As described above, since the asterisk indicates the simultaneous prescription to be reduced, the correspondence with the insurance score can be maintained by maintaining the relationship of one block. In FIG. 5, an inspection method “scintigram (whole body) (bone)” and a medical material “image recording film (1) one piece” are associated with one asterisk. An insurance score of 2338 × 1 is calculated for the material, and therefore, the decomposed data needs to be recorded in a file while maintaining these relationships.

  In order to determine whether or not the character string extracted by the text data processing unit 50 is a registered medical related term, the matching processing unit 52 refers to the medical related term registered in the distribution table storage unit 78. To do. The distribution table is configured by associating all medical terms stored in all types of dictionary tables with the types of dictionary tables to be referred to. Therefore, the matching processing unit 52 can determine whether or not the character string extracted by the text data processing unit 50 is a medical related term by referring to the distribution table. In some cases, it is possible to specify a dictionary table to be referred to. This shortens the matching processing time by specifying the dictionary based on the distribution table rather than searching the extracted character strings for all types of dictionary tables when performing the matching processing described later. It becomes possible to do. Details of the matching process will be described later. If the extracted character string is not a medical term, the text data processing unit 50 performs a combining process on the character string.

  The text data processing unit 50 performs processing for combining the text data that could not be extracted as medical related terms after the text data extracted using the distribution table is determined as medical related terms. In the reception, blanks are often inserted for each medical term for readability. However, as described above, a plurality of medical terms may be described in succession. On the other hand, a space may be inserted between one long medical term and the like. In such a case, the text data processing unit 50 once extracts one medical-related term as a plurality of text data, but these are combined and extracted as one medical-related term. Is preferred.

  For this purpose, the combined table storage unit 77 holds a combined table that lists medical related terms generated by combining a plurality of character strings. As described with reference to S24 in FIG. 1, this list has been executed by an operator in advance, and medical terms that are not registered in the distribution table and divided into a plurality of spaces or line breaks have appeared. In this case, the operator adds the medical related terms to the list and sequentially creates a connection table.

  The text data processing unit 50 combines continuously extracted text data from a plurality of text data not registered in the distribution table. When one medical-related term is extracted by being divided by a blank or a line feed, these are extracted as continuous text data. Therefore, the text data processing unit 50 tries to combine the extracted text data once, and investigates whether or not the combined text data is a medical related term registered in the combination table. At this time, the number of text data to be combined may be two or more, and the number is not limited. When the extracted text data is joined and the joined table is examined, if it is determined that the term is a medical related term registered in the joined table, the text data processing unit 50 displays the joined text data. Are extracted as medical terms. The matching processing unit 52 refers to the distribution table and specifies a dictionary table to be referred to for the medical related term.

  Note that the matching processing unit 52 may determine whether the combined text data is a medical term by referring to the sorting table. On the other hand, when using a combined table as described above, since the amount of data registered in the combined table is not larger than the distribution table, the text data combined using the distribution table is a medical term. Compared with the case of determining whether or not, the processing time can be shortened.

  The master table storage unit 70 stores a master table in which standard words of names such as sickness names and medicines employed in the receipt file generation system 3 are associated with their identification codes and attribute information. The identification code is the only code for specifying the standard word, and may be a unique code in the receipt file generation system 3. An arbitrary identification code may be assigned to an injury or illness, but since a drug price may change over time for a pharmaceutical product, a new identification code is assigned each time the attribute information changes. Also good. In addition, when fixing an identification code, it is also possible to cope with a change in drug price by performing version management of the master table.

  There are various types of attribute information depending on the type of medical terms, and the attribute information may include identification codes in various code systems in the medical industry. For example, the master table stores the injury / illness name “appendicitis” in association with an ICD-10 code for identifying appendicitis or an ICD classification. This ICD classification is “digestive diseases”. The same applies to the drug name. In the case of a drug, it has insurance points as attribute information. The master table is created for a plurality of categories, such as an injury and disease master table that standardizes wound names and a pharmaceutical product master table that standardizes drug names.

  The dictionary table storage unit 72 stores a dictionary table in which a standard word and its dialect are associated with the identification code of the standard word as a synonym of the standard word. In this example, “cecum” and “appe” are dialects for the standard word of the disease “appendicitis”, but “appendicitis”, “appe”, and “cecum” A dictionary table is stored in association with the identification code. The dictionary table is created for a plurality of sections such as a disease / disease dictionary table in which the names of wounds and diseases are converted into a dictionary and a drug dictionary table in which drugs are converted into a dictionary corresponding to the master table.

  FIG. 7 shows registration contents of a dictionary table and a master table generated for one medicine. In the dictionary table, the version, drug name, and drug code of the master table are registered in association with each other. As shown in the figure, there are a plurality of terms representing the same medicine, and each medicine name is associated with the same medicine code (identification code). Various attribute information is registered in the master table in association with the drug name and drug code. Here, the attribute information includes drug price (insurance score), total amount, concentration, unit, unit amount, manufacturer, and the like.

  The matching processing unit 52 searches the text data extracted by the text data processing unit 50 from the dictionary table specified by the distribution table, and performs matching processing. Specifically, the matching processing unit 52 searches the dictionary table to determine whether or not medical related terms of the text data are stored. In the present embodiment, since the extracted text data is determined to be a medical term by referring to the distribution table, the text data exists in the dictionary table. . For example, when the text data is “ephedrine hydrochloride injection solution (4% 1 ml)”, the matching processing unit 52 determines that the same medical related term exists in the drug dictionary table, reads the drug code, and , Notify the receipt file generation unit 54. Upon receipt of this notification, the receipt file generation unit 54 associates the text data “ephedrine hydrochloride injection solution (4% 1 ml)” with the standard word “ephedrine hydrochloride injection solution 4% 1 ml” based on the drug code. Generate a file. For example, the receipt file generation unit 54 may generate a receipt file that associates the text data with the drug code (100000014741), or a receipt file that associates the data address of the ephedrine hydrochloride injection solution in the master table. May be generated. In any case, the receipt file generation unit 54 generates a receipt file by associating the text data with the standard word of the master table based on the medicine code.

  When the text data is another dialect expressing “ephedrine hydrochloride injection solution 4% 1 ml” not included in the dictionary table, the matching processing unit 52 uses the name “ephedrine hydrochloride injection solution 4% 1 ml”. It cannot be recognized as a dialect. When there is no name corresponding to the dictionary table, the matching processing unit 52 stores the text data as unknown data in a predetermined storage area of the data storage unit 66.

  When the operator looks at the name stored in the predetermined storage area and determines that this is another name for the ephedrine hydrochloride injection solution 4% 1 ml, the operator updates the dictionary table, and the ephedrine hydrochloride injection solution 4% 1 ml. Add as a dialect. At this time, as shown in FIG. 7, a new dialect and a medicine code (100000014741) are additionally registered in association with each other. If a new injury or illness occurs or if a new drug appears, the master table should be updated before the matching process, but if the table update is not in time, The operator updates the master table and the dictionary table as appropriate.

  In the case of pharmaceuticals, in particular, quantitative expressions are relatively free on the receipt in many cases. For the text data “Ephedrine hydrochloride injection solution 4% 1 ml 1A”, the text data processing unit 50 changes “Ephedrine hydrochloride injection solution 4% 1 ml 1A” to “Ephedrine hydrochloride injection solution 4% 1 ml” (pharmaceutical name), “1”. (Quantity) and “A” (Unit). At this time, if the operator determines that the expression “A” (unit) indicates “ampoule”, “ephedrine hydrochloride injection solution 4% 1 ml 1A” is added to the dictionary table. The dictionary table of FIG. 7 shows a state including 1 ml 1A of ephedrine hydrochloride injection solution 4% added by the operator in this way.

  On the other hand, when the dictionary table for “ephedrine hydrochloride injection solution 4% 1 ml 1A” is not created, the matching processing unit 52 may autonomously determine this correspondence. The matching processing unit 52 determines the consistency between the extracted text data and the medical-related terms stored in the master table or the dictionary table, and when the unmatched portion is a quantity expression, the quantity expression is Or it judges that it should be matched with the quantity expression contained in the medical related term memorize | stored in the dictionary table.

  Specifically, the matching processing unit 52 searches the master table for medical related terms including “ephedrine hydrochloride injection solution 4% 1 ml”. At this time, it may be possible to refer to the master table after first searching the dictionary table to identify the drug code. The matching processing unit 52 refers to the data content of the ephedrine hydrochloride injection solution 4% 1 ml contained in the master table and recognizes that the unit of the quantity expression of the ephedrine hydrochloride injection solution 4% 1 ml is “1 tube”. The matching processing unit 52 recognizes that the second “1” of “ephedrine hydrochloride injection 4% 1 ml 1A” is a number because it is a number, and “A” has an unknown classification. However, it is a unit expression because it is a character following a number. Based on this guess, the matching processing unit 52 recognizes that “A” represents “tube”, and determines that “1A” corresponds to “1 tube”. As described above, the matching processing unit 52 can associate the dialect of the quantity expression with the standard word even if the expression indicated by the text data is not registered in the dictionary table, and the receipt file generating unit 54 can select “Ephedrine hydrochloride injection 4 It is possible to generate a receipt file in which the drug code (100000014741) is associated with “% 1ml1A”.

  When the name stored in the predetermined storage area is a special expression, the operator creates a dictionary table (temporary dictionary table) that is used only for a predetermined period. The special expression means a character string that is not stored in the dictionary table as described above and is not normally used. For example, it cannot be classified as another common name for 4% 1 ml of ephedrine hydrochloride injection solution, and it is an expression used only in a certain medical institution, or is simply a typographical error in medical terms The operator classifies the character string as a special expression and creates a temporary dictionary table.

  The temporary dictionary table is a dictionary that is used only for a predetermined period, for example, the month. When the temporary dictionary table is created, the temporary dictionary table is stored in the temporary dictionary table storage unit 74. If the text name of the drug name “ephedrine hydrochloride injection 4% 1 ml” is erroneously written as “ephedrine hydrochloride injection 4% 1 ml”, there is no other option, clearly “ephedrine hydrochloride injection 4 If it is determined that the error is “% 1 ml”, “the ephedrine hydrochloride injection solution 4% 1 ml” is linked to the official name “ephedrine hydrochloride injection solution 4% 1 ml” as the dictionary for the month. The same applies to special expressions that are used only in certain medical institutions, and are linked to standard words by the operator. In this way, the temporary dictionary table is created by associating a special expression with the identification code of the standard word stored in the master table. The receipt file generation system 3 does not correct a special expression such as an error in a receipt entry, but associates a standard language identification code with the special expression to save the receipt entry in a unified manner. It will be handled as data that can be processed.

  In the case of medicines, naming is basically free, and in particular, there are many combinations of katakana, so even if there are no medicines with similar names in this month, “Ephadrine Hydrochloride Injection 4” There is a possibility that a drug matching “% 1 ml” will appear as a new drug. Therefore, names that are recognized as special expressions such as typographical errors are not registered in the dictionary table as dialects, but are registered in the temporary dictionary table for use as a temporary dictionary and stored in the temporary dictionary table storage unit 74. Store. The receipt file generation unit 54 refers to the temporary dictionary table and generates a receipt file in which text data indicating a special expression is associated with a standard word of the master table.

  As described above, the receipt file generation unit 54 generates a receipt file in which dialects and special expressions in text data are linked to standard words. This receipt file leaves the contents entered in the original paper receipt as they are, and the dialects and special expressions contained therein are associated with standard words by referring to the dictionary table or temporary dictionary table.

  There are various ways to use the receipt file. For example, it can be used for data processing useful for researchers and pharmaceutical manufacturers, such as statistical processing and trends in pharmaceuticals used. Furthermore, since the receipt file is created in association with the encrypted personal information, it is possible to grasp the personal time-series medical history and check whether appropriate medical practice is being performed. Predictive medicine is also possible and can be used as data for maintaining the health of patients. These processes are executed by the data processing unit 58.

  In the above-described embodiment, an example in which a dictionary table is selected using a sorting table has been described. As an alternative method, the matching processing unit 52 uses all sorts of dictionary tables without using the sorting table. The corresponding medical term may be searched with reference to the reference.

  Further, in order to further reduce the search time than using the above-described distribution table, the distribution table storage unit 78 associates the character strings characteristic of the standard word and its dialect with the type of dictionary table to be referred to. A sorting table may be stored. Regarding the ephedrine hydrochloride injection solution 4% 1 ml, this sorting table associates the character string “ephedrine injection solution” common to all dialects with a dictionary table to be referred to, that is, a pharmaceutical dictionary table. The matching processing unit 52 can specify the medicine dictionary table as the dictionary table to be referred to by referring to the distribution table. By reducing the amount of data in the distribution table, the search time for the distribution table can be shortened, and as a result, the matching processing time can be shortened.

FIG. 8 shows a processing flow for associating with a standard word. In this flow, the drug name is taken as an example.
First, the text data is decomposed and the drug name is extracted (S102). The matching processing unit 52 searches the drug name in the dictionary table (S104), and if it is found that the drug name is stored in the dictionary table (Y in S104), the receipt file generating unit 54 reads the text. A receipt file in which the data is associated with the drug code and the text data is associated with the standard word of the master table is generated (S106).

  When not registered in the dictionary table (N in S104), the matching processing unit 52 stores the medicine name as unknown data in a predetermined storage area of the data storage unit 66. If the operator confirms the drug name and determines that it is a new drug (Y in S108), the operator adds the drug name to the master table and dictionary table (S110), and associates the standard word with the drug name. Generate a receipt file.

  If it is not a new drug (N in S108) and is a new dialect for a certain standard word (Y in S112), the operator adds this drug name to the dictionary table, and the receipt file generation unit 54 A receipt file in which the corresponding standard word is associated with the drug name is generated (S114).

  If it is not a new dialect (N in S112) and is a special expression (Y in S116), the operator adds this medicine name to the temporary dictionary table, and the receipt file generation unit 54 selects the corresponding standard word. Then, a receipt file associated with the specially expressed drug name is generated (S118). If it is not a special expression (N in S116), an error message is notified to the operator.

  In the above, this invention was demonstrated based on the Example. These embodiments are exemplifications, and it is understood by those skilled in the art that various modifications can be made to the combination of each component and each processing process, and such modifications are within the scope of the present invention. is there.

  In the present embodiment, the description has been given on the assumption that the receipt file generation system 3 generates a reception file in which standard words of medical related terms are associated with dialects. For example, this technology can be used not only as a receipt, but also as a medical record file generation system that generates a medical record file that associates standard medical terms and dialects with texts such as electronic medical records. Is possible. In the medical record file generation system, it is possible to cope with the problem by replacing the target in this embodiment with a medical record, and it is possible to greatly contribute to the unification of medical related terms among a plurality of hospitals. Moreover, this system can be suitably applied in a field where there are many synonyms as well as medical-related terms and functions usefully. For example, in order to uniformly manage documents such as contracts created by lawyers in the legal industry, legal terms used to represent the same meaning are classified into standard words and dialects, and dialects correspond to standard words. By creating the attached data file, it is possible to easily search for documents.

It is a figure which shows the receipt processing flow in an Example. It is a figure which shows the receipt processing system in an Example. It is a figure which shows the structure of a text data generation system. It is a figure which shows the image data of a medical remuneration statement (receipt). It is a figure which shows the example of a display of the data file which recognized the summary column and classified and produced | generated for every item. It is a figure which shows the structure of a receipt file production | generation system. It is a figure which shows the registration content of the dictionary table produced | generated with respect to one pharmaceutical, and a master table. It is a figure which shows the processing flow which matches with a standard word.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Receptacle processing system, 2 ... Text data generation system, 3 ... Receipt file generation system, 10 ... Image data generation part, 12 ... OCR part, 16 ... Input part, 18 ... Display, 20 ... Drive device, 22 ... Storage unit, 30 ... Storage medium, 32 ... Receive image, 34 ... Personal information image, 36 ... Medical history image, 38 ... Summary column, 50 ... Text data processing unit, 52 ... Matching processing unit, 54 ... Receipt file generation unit, 55 ... Logic check processing unit, 56 ... Corresponding check processing unit, 58 ... Data processing unit, 60 ... Input unit, 62 ... Display, 64 ... Reference table storage unit, 66 ... Data storage unit, 68 ... Drive device, 70 ... Master table Storage unit, 72 ... dictionary table storage unit, 74 ... one o'clock dictionary table storage unit, 76 ... separation table storage unit, 77 ... coupling table storage unit, 78 ... sorting table storing unit.

Claims (11)

A master table storage unit that stores a master table that associates medical-related terms serving as standard words with their identification codes and attribute information;
A dictionary table storage unit for storing a dictionary table in which synonyms representing the same meaning as a standard word as medical related terms are associated with an identification code of the standard word;
A text data processing unit for extracting text data contained in a data file in which a receipt is converted into text;
A matching processing unit that searches the extracted text data in a dictionary table;
When the matching processing unit finds that the text data is stored as a medical term in the dictionary table, a receipt file that associates the text data with the standard word of the master table based on the identification code A receipt file generator for generating
A receipt file generation system comprising:   The said matching process part memorize | stores the text data in a predetermined | prescribed storage area as unknown data, when the said text data is not memorize | stored as a medical-related term in the dictionary table. Receipt file generation system. A temporary dictionary table storage unit that stores a temporary dictionary table in which character strings that are not stored in the dictionary table and are not normally used are associated with identification codes of standard words;
The said receipt file production | generation part produces | generates the receipt file which matched the text data of the said character string with the standard word of a master table based on the temporary dictionary table. Receipt file generation system.   The said receipt file production | generation part produces | generates a receipt file only using the correspondence memorize | stored in the temporary dictionary table about the text data of the said character string only for a predetermined period, The receipt of Claim 3 characterized by the above-mentioned. File generation system.   5. The receipt file generation according to claim 1, wherein each of the master table storage unit and the dictionary table storage unit stores the master table and the dictionary table for each type of medical terms. system. A distribution table storage unit that stores a distribution table in which character strings are associated with types of dictionary tables to be referenced;
6. The receipt file generation system according to claim 5, wherein the matching processing unit specifies a dictionary table to be referred to the extracted text data with reference to the distribution table. A decomposition table storage unit for storing a character string and a decomposition table associated with a plurality of medical terms that are decomposed by dividing the character string;
7. The receipt file generation according to claim 1, wherein the text data processing unit extracts text data obtained by disassembling text data of a character string for each medical term with reference to a disassembly table. system. A combined table storage unit that stores a combined table that lists medical-related terms generated by combining a plurality of character strings;
8. The receipt file generation system according to claim 7, wherein the text data processing unit extracts text data obtained by combining text data of a plurality of character strings with reference to a combination table.   The matching processing unit determines the matching between the extracted text data and medical related terms stored in the master table or the dictionary table, and when the unmatched portion is a quantity expression, the quantity expression is The receipt file generation system according to any one of claims 1 to 8, wherein it is determined that it should be associated with a quantity expression included in a medical-related term stored in a master table or a dictionary table. A master table storage unit that stores a master table that associates medical-related terms serving as standard words with their identification codes and attribute information;
A dictionary table storage unit for storing a dictionary table in which synonyms representing the same meaning as a standard word as medical related terms are associated with an identification code of the standard word;
A text data processing unit for extracting text data included in the data file of the electronic medical record;
A matching processing unit that searches the extracted text data in a dictionary table;
When the matching processing unit finds that the text data is stored as a medical term in the dictionary table, based on the identification code, the medical record file that associates the text data with the standard word of the master table A file generator for generating
A medical record file generation system comprising: A master table storage unit that stores a master table that associates a term that is a standard word with its identification code and attribute information;
A dictionary table storage unit that stores a dictionary table in which a standard word and a synonym representing the same meaning as the standard word are associated with an identification code of the standard word;
A text data processing unit for extracting text data contained in the data file;
A matching processing unit that searches the extracted text data in a dictionary table;
A file for generating a data file in which the text data is associated with the standard word of the master table based on the identification code when the matching processing unit finds that the text data is stored in the dictionary table A generator,
A file generation system comprising:

Images