JP2022036392A - Prescription reading program and prescription reader - Google Patents

Abstract

To provide a prescription reading program and a prescription reader, capable of appropriately reading prescriptions even though unclear characters exist therein.SOLUTION: A prescription reading program causes an information processing device to execute: reception processing for receiving prescription data; layout type determination processing for, based on the prescription data received at the reception processing, determining a layout type; text area determination processing for, based on the layout type determined at the layout type determination processing, determining a text area; text recognition processing for, on the text area determined at the text area determination processing, performing recognition processing of text data using a machine-learned character recognition algorithm; and extraction processing for, on the text data recognized at the text recognition processing, extracting words including a name and a dose of a medication using a machine-learned word extraction algorithm.SELECTED DRAWING: Figure 4

Description

The present invention relates to a prescription reading program and a prescription reading device that read information such as characters of a prescription.

Patent Document 1 describes a pharmacy business support system. In Patent Document 1, "The prescription information is, for example, image data obtained by photographing a prescription issued by a medical institution. In addition, the photographed image is character-recognized, the prescription content is converted into text data, and the image data and the text data are obtained. It may be used as prescription information. "

Japanese Unexamined Patent Publication No. 2020-086833

The pharmacy receives the prescription from the medical institution via paper or fax. When a pharmacy wants to convert a prescription received on paper into text data, it reads it with a scanner and converts it into text data by OCR (Optical Character Recognition). Alternatively, the pharmacy converts the image data received by FAX into text data by OCR (Optical Character Recognition). However, the scanned image data and the image data received by FAX may have unclear characters, and in many cases, the characters cannot be properly recognized by the conventional OCR.

Therefore, an object of the present invention is to provide a prescription reading program and a prescription reading device capable of appropriately reading a prescription even if the characters are unclear.

The prescription reading program according to the present invention is determined by a reception process for receiving prescription data, a layout type determination process for determining a layout type based on the prescription data received in the reception process, and a layout type determination process. A text area determination process for determining a text area based on the layout type and a text data recognition process for the text area determined by the text area determination process using a machine-learned character recognition algorithm. To have the information processing apparatus execute the recognition process and the extraction process of extracting the word including the name and quantity of the drug by using the machine-learned word extraction algorithm for the text data recognized by the text recognition process. It is characterized by.

As described above, the prescription reading program according to the present invention determines the layout type of the prescription, determines the text area, and uses a machine-learned character recognition algorithm such as a neural network and machine-learned word extraction for the determined text area. Extract words such as drug name and quantity using an algorithm. Therefore, the prescription reading program according to the present invention can perform character recognition and word recognition with extremely higher accuracy than the conventional OCR. Therefore, the prescription reading program according to the present invention can appropriately read the prescription even if the prescription data is unclear characters captured by FAX or a scanner.

The prescription reading program can further execute a correction process of correcting the error of the text data recognized by the text recognition process by using a machine-learned algorithm, and can further improve the reading accuracy of the prescription. In this case, the extraction process is performed on the text data modified by the modification process.

Further, the correction process is performed using a drug name dictionary and a pattern dictionary including a specific pattern included in the prescription. The pattern dictionary contains information such as the characters "general" used in generic drugs, the katakana "ro" and the kanji "mouth". As a result, the reading accuracy of the prescription is further improved.

The machine-learned word extraction algorithm is designed to understand the relationship between the first algorithm, which is learned to determine and tag the format of each part of the related text in the prescription, and the specific keywords in the prescription. Has a second algorithm, which is learned in. As a result, the prescription reading program can extract the unique description contents in the prescription with high accuracy and organize them in a format that can be appropriately registered in the database and can be used.

The layout type determination process is performed by comparing with the layout type stored in the database. Prescriptions may have new layout types due to revisions to the law, etc., and may not match the database. However, the prescription reading program accepts the operation of the text area from the user when there is no layout type that matches the database, and performs the registration process to register it in the database, so that it is appropriate even if a new layout type occurs. Can be read.

According to the present invention, the prescription can be appropriately read even if the characters are not clear.

It is a block diagram which shows the structure of the prescription reading system 1. It is a block diagram which shows the structure of a server 11. It is a flowchart which shows the operation of the prescription reading system 1. It is a flowchart which shows the operation of the server 11 at the time of reading a prescription. It is a flowchart which shows the operation of the server 11 at the time of reading a prescription. It is a figure which shows an example of the prescription data.

Hereinafter, the prescription reading system 1 according to the embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing a configuration of a prescription reading system 1. The prescription reading system 1 includes a server 11, a reception terminal 12, a user terminal 13, and a receipt computer 14. The server 11, the reception terminal 12, the user terminal 13, and the receipt computer 14 are connected via a network.

The reception terminal 12 is realized by a general information processing device (personal computer). The reception terminal 12 is connected to a fax machine or a scanner (not shown) and receives prescription data obtained by converting a prescription into image data. The prescription will be sent by paper or fax from the medical institution. When the pharmacist at the pharmacy receives the prescription line on paper, he / she inputs the prescription data to the reception terminal 12 using a scanner (not shown). When the prescription is received by fax, the reception terminal 12 receives the prescription data from the fax machine. Alternatively, the pharmacist may print the prescription data with a fax machine, read the printed prescription with a scanner, and input it to the reception terminal 12. The reception terminal 12 transmits the prescription data to the server 11.

The server 11 is an example of the prescription reading device of the present invention. The server 11 may be installed in a pharmacy or may consist of a group of servers on the Internet. The server 11 receives the prescription data from the reception terminal 12 and performs an operation of reading the contents of the prescription.

The user terminal 13 is a terminal used by a person (user) who performs pharmacy business such as a pharmacist, and is a portable information processing device such as a smartphone or a tablet computer. The user confirms the contents of the prescription read by the server 11 using the user terminal 13, and inputs the contents of the prescription to the receipt computer 14. In addition, the user can confirm the drug history of the patient using the user terminal 13 and create a new drug history.

The receipt computer 14 is a computer that creates a receipt according to the contents of the prescription input from the user terminal 13. The receipt computer 14 is realized by an application program installed on a general personal computer.

FIG. 2 is a block diagram showing the configuration of the server 11. The server 11 includes a network I / F 51, a storage unit 52, a CPU 53, and a RAM 54.

The network I / F 51 is connected to a network such as the Internet. In the present embodiment, the server 11 is connected to the reception terminal 12 and the user terminal 13 via the network I / F 51.

The CPU 53 reads the program stored in the storage unit 52 into the RAM 54 and performs various operations. The storage unit 52 stores the basic program (OS) of the server 11 and the prescription reading program in the present embodiment. Further, the storage unit 52 includes various databases such as information indicating the layout type of the prescription data described later and personal information of the patient.

The CPU 53 functions as a prescription reading device by reading the prescription reading program from the storage unit 52 into the RAM 54.

FIG. 3 is a flowchart showing the operation of the prescription reading system 1. First, the prescription is sent from the medical institution by paper or fax (S11). The pharmacist at the pharmacy receives the prescription sent by paper or fax (S12).

Then, the pharmacist at the pharmacy scans the paper prescription using the reception terminal 12 and transmits it to the server 11 as prescription data, or transmits the prescription data received by FAX to the server 11 (S13). The pharmacist may send the prescription data to the server 11 by using the application program of the user terminal 13. Further, the reception terminal 12 or the user terminal 13 may automatically transmit the prescription data to the server 11 when the prescription data is received via FAX or received from the scanner. In particular, home-use pharmacies such as welfare facilities for the elderly receive a large number of prescriptions from medical institutions at the same time. Therefore, the prescription reading system 1 of the present embodiment automatically sends the prescription data to the server 11 when the prescription data is received, and performs the following reading operation, thereby significantly reducing the load on the pharmacy business. can.

Next, the server 11 executes a prescription reading program to perform a prescription reading operation (S14).

4 and 5 are flowcharts showing the operation of the server 11 when reading the prescription. First, the server 11 extracts the feature amount of the image from the prescription data (S101). This process is performed by an algorithm for extracting features such as SIFT (Scale Invariant Feature Transform).

As mentioned above, the pharmacy receives the prescription data by fax or reads it with a prescription scanner. Therefore, in the prescription data, the size and inclination of the image are not constant. The server 11 can extract the feature amount of the prescription image that does not depend on the image size and the inclination by using the feature amount extraction algorithm of SIFT.

Next, the server 11 determines the corresponding layout type from the database based on the extracted features (S102). The database stores multiple prescription layout types. The server 11 uses an estimation algorithm such as RANSAC (Random Sample Consensus) to refer to a database and determine a layout type that matches the layout type of the received prescription data.

The server 11 determines whether or not a matching layout type exists (S103). When the server 11 determines that a matching layout type exists, the server 11 performs a determination process for the text area (S104). The text area includes at least an area in which patient information such as the patient's name is described, an area in which a date is described, an area in which drug information is described, and the like. The layout type registered in the database contains position information indicating which text area exists in which position in the prescription data. The server 11 determines the text area by reading the position information of the text area from the layout type registered in the database.

On the other hand, when the server 11 determines that there is no matching layout type, the server 11 accepts the text area from the user (S105). A user such as a pharmacist manually designates an area in which the text is described in the prescription data by using the user I / F of the reception terminal 12 or the user terminal 13. The server 11 registers the designated text area as a new layout type in the database (S106), and performs the determination process of the text area (S104).

Prescriptions may have new layout types due to revisions to the law, etc., and may not match the layout types stored in the database. However, when the server 11 does not have a layout type that matches the database, the server 11 accepts a text area designation operation from the user and performs a registration process for registering in the database, so that even if a new layout type occurs, the next time. The layout type can be automatically determined from.

After that, the server 11 performs a text recognition process on the text area in which the patient information is described (S107). The patient information includes information such as the patient's name and birthday. The text recognition process is performed by an arbitrary character recognition algorithm (for example, Tesseract-OCR equipped with an algorithm such as a neural network).

The server 11 compares the patient information recognized by the text recognition process with the database, and determines whether or not the corresponding patient information exists (S108). When the server 11 determines that the corresponding patient exists, the server 11 continues the processing of S201 to S207 shown in FIG. When the server 11 determines that the corresponding patient does not exist, the server 11 registers the recognized patient information in the database (S109) and proceeds to the process of S201 to S207.

Next, as shown in FIG. 5, the server 11 performs segment processing in the text area in which the drug information is described (S201). FIG. 6 is a diagram showing an example of prescription data. The segment processing is a processing for dividing a text area in which drug information is described into segments for each line, as shown in the hatched portion of FIG.

After that, the server 11 performs text recognition processing in each segment (S202). The text recognition process is performed by an arbitrary character recognition algorithm (for example, Tesseract-OCR equipped with an algorithm such as a neural network). This process preferably uses a machine-learned character recognition algorithm machine-learned for prescriptions. Of course, the server 11 may perform text recognition processing using a general machine-learned algorithm.

Next, the server 11 performs spell correction processing on the recognized text data (S203). The spelling correction process uses an arbitrary algorithm such as SymSpell. The SymSpel algorithm uses a government-provided drug name dictionary to correct text recognition errors related to drug names. As a result, the server 11 can appropriately correct the error generated in the text recognition process of S202.

Further, the server 11 performs a correction process of a specific pattern in the prescription (S204). An arbitrary algorithm such as n-gram is used for the correction process of the specific pattern. The n-gram algorithm corrects a text recognition error that could not be corrected by the processing of S203 by using a pattern dictionary containing a specific pattern in the prescription. Certain patterns in prescriptions include, for example, the letters "general" used in generic drugs. The n-gram algorithm also corrects the difference between the katakana "ro" character and the kanji "mouth".

The server 11 performs the first word extraction process (S205) and the second word extraction process (S206) using the modified text data. Both the first word extraction process and the second word extraction process are performed using a machine-learned word extraction algorithm such as a neural network. The machine-learned word extraction algorithm is, for example, an NLP (Natural Language Processing) algorithm. The first NLP algorithm used in the first word extraction process of the present embodiment is learned to determine and tag the format of each part of the relevant text in the prescription. Each tag includes, for example, the name of the drug, dosage, duration of dosing, timing of ingestion, and the like.

The second NLP algorithm used in the second word extraction process is learned to understand the relationships (syntax and order) between specific keywords in a prescription. For example, in the prescription data shown in FIG. 6, the second NLP algorithm is 1. Generic name Prescription drug name "○○○○ tablets" 12 mg, 2. Dosing "3 tablets" 3. Dosing period "14 days", 4. The intake timing "before going to bed once a day" is related in this order. Similarly, the second NLP algorithm is 1. Generic name Prescription drug name "○ △△ tablets" 7.5 mg, 2. With a dosage of "1 tablet", 3. Dosing period "14 days", 4. The intake timing "before going to bed once a day" is related in this order.

By using the above-mentioned first NLP algorithm and second NLP algorithm, the server 11 can extract the specific description contents in the prescription with high accuracy and organize them in a format that can be appropriately registered in the database and can be used.

The server 11 registers the words extracted as described above in the database (S207). Then, returning to the flowchart of FIG. 3, the server 11 converts the reading result into a QR code (registered trademark) of a standard (for example, JAHIS) that can be read by the receipt computer 14 (S15). Further, the server 11 registers the read prescription data, the position information of the word registered in S207 (information indicating the position of each word in the prescription data), and the QR code (registered trademark) generated in S15 in the database. (S16).

On the other hand, the server 11 transmits the information registered in the database to the user terminal 13 and updates the information of the user terminal 13 (S17). The pharmacist can confirm the contents of each prescription by using the application program of the user terminal 13.

After that, the QR code (registered trademark) generated by the server 11 is input to the receipt computer 14 via the application program of the user terminal 13 (S18).

As described above, the prescription reading program of the present embodiment determines the layout type of the prescription, determines the text area, and uses a machine-learned character recognition algorithm such as a neural network and a machine-learned word for the determined text area. Words such as the name and quantity of the drug are extracted using the extraction algorithm. Therefore, the prescription reading program of the present embodiment can perform character recognition and word recognition with extremely higher accuracy than the conventional OCR. Therefore, the prescription reading program according to the present embodiment can appropriately read the prescription even if the prescription data is unclear characters captured by FAX or a scanner. In particular, a home-use pharmacy receives a large number of prescriptions from a medical institution at the same time, but the prescription reading program of the present embodiment can significantly reduce the load of prescription reading and input work to a receipt computer.

The server 11 stores the reading result information in the database for each patient. The pharmacist can refer to all the past reading results for each patient by using the application program of the user terminal 13. When the information of the user terminal 13 is updated in the process of S17 of FIG. 3, the pharmacist confirms the contents of the prescription, explains the prescribed medicine to the patient via a telephone or the like, and the patient's taking status. Can be confirmed. The pharmacist records these contents as a drug administration history management record (hereinafter referred to as “drug history”) using the user terminal 13. The drug history includes subjective information (contents spoken by the patient, subjective symptoms) (S: Subjective), objective information (prescription contents, changes, pharmacist's observations, test data) (O: Objective), evaluation and guidance. (Evaluation / judgment as a pharmacist from S and O and guidance content based on it) (A: AAssessment), and plan (points to ask next time and points to note) (P: Plan). The pharmacist can easily create a new drug history by referring to the past information in addition to the current prescription after the information of the user terminal 13 is updated in the process of S17. As a result, the prescription reading system 1 of the present embodiment also functions as a drug history creation business support program.

Further, the user terminal 13 may transmit the drug history created by the pharmacist to the server 11. By registering the created new drug history in the database, the server 11 can present the drug history to the pharmacist via the user terminal 13 the next time the new prescription is read.

Further, the server 11 may automatically create the newly created drug history based on the contents of the newly read prescription and the past drug history. The drug history is created using a predetermined algorithm. The server 11 transmits the created new drug history to the user terminal 13. The pharmacist confirms the drug history created on the server 11, and corrects the drug history created on the server 11 according to the result of the explanation to the patient and the confirmation of the situation. Alternatively, if the pharmacist determines that there is no problem with the drug history created by the server 11 as a result of the explanation to the patient and the confirmation of the situation, the pharmacist determines the drug history created by the server 11. When the pharmacist confirms the drug history, the confirmed drug history is registered in the server 11 and the receipt computer 14 from the user terminal 13.

In this way, the server 11 of the present embodiment can support the drug history creation work by automatically creating the drug history, and can further reduce the load of the pharmacy work.

It should be noted that the description of this embodiment is an example in all respects and should be considered not to be restrictive. The scope of the invention is indicated by the claims, not by the embodiments described above. Further, the scope of the present invention includes the scope equivalent to the claims.

For example, the prescription reading program may be executed not on the server 11 but on the reception terminal 12, the user terminal 13, or the receipt computer 14.

1 ... Prescription reading system 11 ... Server 12 ... Reception terminal 13 ... User terminal 14 ... Receipt 51 ... Network I / F
52 ... Storage unit 53 ... CPU
54 ... RAM

The server 11 stores the reading result information in the database for each patient. The pharmacist can refer to all the past reading results for each patient by using the application program of the user terminal 13. When the information of the user terminal 13 is updated in the process of S17 of FIG. 3, the pharmacist confirms the contents of the prescription, explains the prescribed medicine to the patient via a telephone or the like, and the patient's taking status. Can be confirmed. The pharmacist records these contents as a drug administration history management record (hereinafter referred to as “drug history”) using the user terminal 13. The drug history includes subjective information (contents spoken by the patient, subjective symptoms) (S: Subjective), objective information (prescription contents, changes, pharmacist's observations, test data) (O: Objective), evaluation and guidance. (Evaluation / judgment as a pharmacist from S and O and guidance content based on it) (A: Assessment ), and plan (points to ask next time and points to note) (P: Plan). The pharmacist can easily create a new drug history by referring to the past information in addition to the current prescription after the information of the user terminal 13 is updated in the process of S17. As a result, the prescription reading system 1 of the present embodiment also functions as a drug history creation business support program.

Claims (6)

Receiving process to receive prescription data and
A layout type determination process for determining a layout type based on the prescription data received in the reception process, and a layout type determination process.
The text area determination process for determining the text area based on the layout type determined in the layout type determination process, and the text area determination process.
A text recognition process for recognizing text data using a machine-learned character recognition algorithm for the text area determined by the text area determination process, and a text recognition process.
An extraction process for extracting words including the name and quantity of a drug using a machine-learned word extraction algorithm for the text data recognized by the text recognition process.
A prescription reading program that causes the information processing device to execute. Further, a correction process for correcting the error of the text data recognized by the text recognition process by the machine-learned algorithm is executed.
The extraction process is performed on the text data corrected by the correction process.
The prescription reading program according to claim 1. The correction process is performed using a correction process using a drug name dictionary and a pattern dictionary showing a specific pattern included in the prescription.
The prescription reading program according to claim 2. The machine-learned word extraction algorithm
The first algorithm, which is learned to determine and tag the format of each part of the relevant text in a prescription, and
A second algorithm that is learned to understand the relationships between unique keywords in prescriptions,
Have,
The prescription reading program according to any one of claims 1 to 3. The layout type determination process is performed by comparing with the layout type stored in the database.
When there is no layout type that matches the database, the operation of the text area is accepted from the user, and the registration process to be registered in the database is further executed.
The prescription reading program according to claim 4. A receiver that receives prescription data and
A layout type determination unit that determines the layout type based on the prescription data received by the reception unit, and a layout type determination unit.
A text area determination unit that determines a text area based on the layout type determined by the layout type determination unit, and a text area determination unit.
A text recognition unit that performs text data recognition processing for the text area determined by the text area determination unit, and a text recognition unit.
An extraction unit that extracts words including the name and amount of a drug from the text data recognized by the text recognition unit using a machine-learned algorithm, and an extraction unit.
A prescription reader equipped with.

Images