Text generation device, text generation method, and computer program

Abstract

A document generation device according to the present disclosure comprises: an acquisition means for acquiring document generation information including electronic medical record information and the sharing purpose thereof related to a target patient whose patient information is to be shared with other medical institutions; a reception means for receiving an instruction for generating a medical document for sharing the patient information, the instruction including the document generation information; a generation means for generating a medical document for sharing the patient information on the basis of the document generation information by inputting the instruction to a language model; and an output means for outputting the generated medical document.

Description

Text generation device, text generation method, and recording medium

[0001] The present disclosure relates to a text generation device, a text generation method, and a recording medium.

[0002] There are technologies to reduce the burden of preparing documents for doctors and other medical professionals.

[0003] For example, Patent Document 1 discloses a method of extracting written expressions contained in medical information, interpreting the extracted written expressions as medical facts, extracting definite diseases that have been definitively diagnosed from the medical information, and creating an explanation of the diagnostic basis for the definite diseases that have been definitively diagnosed based on the medical facts.

[0004] JP 2015-138402 A

[0005] When referring a patient to another medical institution, a medical document for sharing patient information may be generated using a template, etc. However, it takes time to input necessary information into the template.

[0006] An example of an objective of the present disclosure is to provide a text generation device or the like that can easily create medical text related to patient information.

[0007] A text generation device in one aspect of the present disclosure includes an acquisition means for acquiring text generation information including electronic medical record information of a target patient whose patient information is to be shared with other medical institutions and a purpose for sharing; a reception means for receiving instructions for generating medical text including the text generation information for sharing the patient information; a generation means for generating medical text based on the text generation information by inputting the instructions into a language model; and an output means for outputting the generated medical text.

[0008] In one aspect of the present disclosure, a text generation method includes a computer acquiring text generation information including electronic medical record information of a patient whose patient information is to be shared with other medical institutions and a purpose for sharing, receiving instructions for generating medical text including the text generation information for sharing the patient information, inputting the instructions into a language model, generating medical text based on the text generation information, and outputting the generated medical text.

[0009] In one aspect of the present disclosure, a recording medium stores a program that causes a computer to acquire electronic medical record information and text generation information including the purpose of sharing of patient information of a target patient whose patient information is to be shared with other medical institutions, accept instructions for generating medical text including the text generation information for sharing the patient information, input the instructions into a language model, thereby generating medical text based on the text generation information, and outputting the generated medical text.

[0010] According to one example of the effect of the present disclosure, medical documents for patients can be easily created.

[0011] FIG. 1 is a block diagram showing the configuration of a text generation device according to the present disclosure. FIG. 2 is a diagram showing a hardware configuration in which the text generation device according to the present disclosure is realized by a computer device and its peripheral devices. FIG. 3 is a diagram showing an example of a medical text generation screen according to the present disclosure. FIG. 4 is a diagram showing an example of a medical text generation screen according to the present disclosure. FIG. 5 is a flowchart showing the text generation operation according to the present disclosure. FIG. 6 is a diagram showing an example of a medical text generation screen according to the present disclosure. FIG. 7 is a block diagram showing the configuration of a text generation device according to the present disclosure. FIG. 8 is a flowchart showing the text generation operation according to the present disclosure.

[0012] Hereinafter, with reference to the drawings, embodiments of a text generation device, a text generation method, a program, and a non-transitory recording medium for recording a program according to the present disclosure will be described in detail. The present embodiments do not limit the disclosed technology.

[0013] First Embodiment Fig. 1 is a block diagram showing the configuration of a text generation device 100 according to the present disclosure. As shown in Fig. 1, the text generation device 100 includes an acquisition unit 101, a reception unit 102, a generation unit 103, and an output unit 104.

[0014] The text generation device 100 is a device for generating medical text for sharing patient information with other medical institutions. One example of medical text is a letter of referral written by a doctor to the doctor at the referring medical institution when the patient is to receive treatment at a different medical institution. Another example of medical text is a report written to report the medical procedure performed at the referring medical institution when the referring medical institution accepts the patient. However, other medical texts may be used as long as they are created to share patient medical information. In this embodiment, a situation in which a letter of referral is generated will be mainly described, but other medical texts may also be used.

[0015] 2 is a diagram showing an example of a hardware configuration in which the text generation device 100 according to the present disclosure is realized by a computer device 500 including a processor. As shown in Fig. 2, the text generation device 100 includes a CPU (Central Processing Unit) 501, memories such as a ROM (Read Only Memory) 502 and a RAM (Random Access Memory) 503, a storage device 505 such as a hard disk for storing a program 504, a communication I / F (Interface) 508 for network connection, and an input / output interface 511 for inputting and outputting data.

[0016] The CPU 501 runs an operating system to control the entire text generation device 100 according to the present disclosure. The CPU 501 also reads programs and data into memory from a recording medium 506 attached to a drive device 507, for example. The CPU 501 also functions as the acquisition unit 101, the reception unit 102, the generation unit 103, and the output unit 104 according to the present disclosure, or as part of these units, and executes processing or commands in the flowchart shown in FIG. 5, which will be described later, based on the program.

[0017] The recording medium 506 is, for example, an optical disk, a flexible disk, a magneto-optical disk, an external hard disk, or a semiconductor memory. A part of the recording medium in the storage device is a non-volatile storage device, and the program is recorded therein. The program may also be downloaded from an external computer (not shown) connected to a communication network.

[0018] The input device 509 is realized by, for example, a mouse, a keyboard, built-in key buttons, etc., and is used for input operations. The input device 509 is not limited to a mouse, a keyboard, or built-in key buttons, and may be, for example, a touch panel. The output device 510 is realized by, for example, a display, and is used to check output.

[0019] As described above, the text generation device 100 shown in FIG. 1 is realized by the computer hardware shown in FIG. 2. However, the means for realizing each part of the text generation device 100 in FIG. 1 is not limited to the configuration described above. The text generation device 100 may be realized by a single physically coupled device, or may be realized by two or more physically separated devices connected by wire or wirelessly. For example, the input device 509 and the output device 510 may be connected to the computer device 500 via a network. The text generation device 100 shown in FIG. 1 can also be configured using cloud computing or the like.

[0020] The acquisition unit 101 is a means for acquiring electronic medical record information of a target patient whose patient information is to be shared with another medical institution, and text generation information including the purpose of sharing. Patient information is, for example, information regarding the patient's medical treatment. The acquisition unit 101 acquires electronic medical record information linked to the ID, etc. of the patient for whom a referral letter is to be created, from a server device, etc. of the medical institution. The acquisition unit 101 acquires, as electronic medical record information, at least one of the name of the injury or illness, the progress of symptoms, test results, the progress of treatment, information on prescribed medication, etc. The acquisition unit 101 may also acquire, as electronic medical record information, the findings of the patient by the referring doctor.

[0021] Furthermore, the acquisition unit 101 acquires at least information for the purpose of sharing from a user such as a doctor on the screen of an application program for creating medical documents. Examples of purposes of sharing include, in the case of a patient referral, a request for further treatment for the patient, a request for follow-up observation when the patient moves, a request for an examination, a request for a second opinion, etc. The acquisition unit 101 may also accept text generation information for purposes other than sharing. For example, the acquisition unit 101 may acquire information about the medical institution to which the patient is referred by input from the user. The acquisition unit 101 outputs the acquired text generation information to the acceptance unit 102.

[0022] The reception unit 102 is a means for receiving instructions for generating medical text, including text generation information, for sharing patient information. The reception unit 102 receives text generation instructions, including the text generation information, on a screen of an application program for generating medical text. The instructions may include generating medical text to satisfy specified generation conditions. Examples of the generation conditions include conditions regarding the content and format to be included in the medical text, such as not including speculative content, stating only the reason for referral, using polite language such as "desu" and "masu," and describing the progress in bullet points. The instructions are, for example, prompts used as input for a large-scale language model, but the form of the instructions is not limited to prompts.

[0023] The language model is, for example, a model that generates sentences according to instructions input by a user. The language model may be a model optimized by machine learning. The language model may be an interactive model that receives prompts and responds to them. The language model may be a large-scale language model trained with a large amount of text data, or a model that has undergone transfer learning of the large-scale language model. For example, GPT-2 (Generative Pre-Training-2), GPT-3, or GPT-4 can be used as the large-scale language model. Furthermore, T5 (Text-to-Text Transfer Transformer), BERT (Bidirectional Encoder Representations from Transformers), RoBERTa (Robustly optimized BERT approach), or ELECTRA (Efficiently Learning an Encoder that Classifies Token Replacements Accurately) may be used as the large-scale language model.

[0024] The generation unit 103 is a means for inputting instructions into a language model to generate medical text for sharing patient information based on text generation information. The generation unit 103 may input instructions into a language model stored in the storage device 505, for example. The generation unit 103 may also perform processing for generating medical text using a language model in a system external to the text generation device 100.

[0025] The output unit 104 is a means for outputting the generated medical text to a display device such as a monitor. For example, the output unit 104 outputs the medical text on the screen of an application program for generating the medical text. The output unit 104 may also transmit the generated medical text to a terminal device used by a doctor, or may output the medical text in response to a trigger such as the doctor's operation to generate a medical text for the target patient.

[0026] The output unit 104 may display annotations for technical terms included in the medical text. Technical terms may be stored, for example, in the storage device 505 together with their annotations, and the output unit 104 may compare the technical terms with the terms stored in the storage device 505 and display the annotations associated with these terms. This allows the recipient to understand the content of the medical text even when the medical text is generated for medical institution staff other than doctors who do not have knowledge of the technical terms.

[0027] 3 is an example of a medical document generation screen according to the present disclosure. In the example of FIG. 3, a medical document creation instruction is accepted on a screen displaying the electronic medical record information of the patient for whom the medical document is to be generated. Also, as shown in FIG. 3, the screen displaying the electronic medical record information 31 includes a display 32 including a button for selecting whether or not to generate a medical document for sharing the patient information of this patient.

[0028] The output unit 104 may also display electronic medical record information that serves as the basis for the information contained in the medical text. In this case, for example, the instructions input by the generation unit 103 to the language model may include indicating the basis for the information contained in the medical text. The information contained in the medical text may include the patient's symptom progression, test results, treatment progression, information on prescribed medications, and the doctor's findings on the patient. The output unit 104 may, for example, display a link to the cited electronic medical record information contained in the medical text, making it possible to output the electronic medical record information. The output unit 104 may also output information that identifies the cited electronic medical record information. The information that identifies the electronic medical record information may, for example, be information such as the date and time of input of the electronic medical record information or the person who input it.

[0029] Fig. 4 is an example of a medical text generation screen according to the present disclosure. The example of Fig. 4 is an example of displaying the created medical text. In the example of Fig. 4, the medical text can be edited by the user on the screen. The medical text in Fig. 4 is an example, and other medical texts may be output.

[0030] The operation of the sentence generation device 100 configured as above will be described with reference to the flowchart of FIG.

[0031] 5 is a flowchart showing an outline of the operation of the text generation device 100 according to the present disclosure. The processing according to this flowchart may be executed based on program control by the processor described above. The processing according to this flowchart is triggered by an operation for creating a medical text for a target patient.

[0032] As shown in Figure 5, first, the acquisition unit 101 acquires text generation information including electronic medical record information of a patient whose patient information will be shared with other medical institutions and a purpose for sharing (step S101). Next, the reception unit 102 receives an instruction to generate a medical text for sharing the patient information, including the text generation information (step S102). The generation unit 103 generates a medical text based on the text generation information by inputting the instruction into a language model (step S103). Finally, the output unit 104 outputs the generated medical text (step S104). This completes the text generation operation of the text generation device 100.

[0033] In the text generation device 100, the generation unit 103 inputs instructions into a language model to generate medical text based on text generation information. This allows doctors to generate medical text by simply performing operations to generate medical text, without having to input patient information. This makes it possible to easily generate medical text for sharing patient information. Furthermore, it is possible to reduce variation in the quality of medical text depending on the creator.

[0034] Modifications of the first embodiment described above will be described. In the present disclosure, the modifications can be implemented in any combination.

[0035] [Variation 1 of the First Embodiment] In the first embodiment, the text generation device 100 may generate medical text using a template text of a medical text prepared in advance. In this case, the acquisition unit 101 acquires the template text of a medical text prepared in advance. The template text is prepared, for example, for each purpose of sharing the referral letter, the medical department of the referred patient, and the progress of the patient, and is stored, for example, in the storage device 505. The reception unit 102 receives an instruction to generate medical text using the template text as a generation condition. The generation unit 103 then generates medical text by applying the text generation information to the template text. The template text is data including text from the medical text, excluding patient-specific information. Here, the specific information is information unique to each patient, including personal information such as the patient's name, date of birth, and electronic medical record information. Generating medical text using the template text allows doctors or the referred medical institution to generate medical text in a format familiar to them.

[0036] [Variation 2 of the First Embodiment] In the first embodiment, the text generation device 100 may generate medical text using sample medical texts prepared in advance. In this case, the acquisition unit 101 acquires the sample medical texts prepared in advance. The sample texts are samples of medical texts used in the past, and are prepared, for example, for each doctor who creates medical texts or each medical institution to which the patient is referred, and are stored in the storage device 505. The reception unit 102 receives, as a generation condition, an instruction to generate medical texts using the sample texts. The generation unit 103 generates medical texts by inputting the instruction and the sample texts into a language model. Generating medical texts using the sample texts makes it possible to generate medical texts that are familiar to the doctor or the medical institution to which the patient is referred.

[0037] [Third Modification of the First Embodiment] The receiving unit 102 may receive an instruction to adopt or not adopt the output medical text. For example, the receiving unit 102 receives a response from the doctor on whether or not to adopt the output text on the screen on which the medical text is output. Furthermore, if the receiving unit 102 receives an instruction not to adopt the output medical text, the receiving unit 102 may further receive an instruction to generate medical text by changing the text generation conditions. In this case, the receiving unit 102 may receive at least one of the text length, the summarization level, and the electronic medical record information to be included in the medical text as conditions for regenerating the medical text. The text length is indicated, for example, by the number of characters or the number of lines. The summarization level indicates, for example, the extent to which the text is summarized relative to the length of the original text. When the receiving unit 102 receives an instruction to regenerate the medical text, the generating unit 103 inputs an instruction to generate medical text using the received generation conditions into the language model and regenerates the medical text.

[0038] In this modification, the receiving unit 102 receives a decision as to whether or not to adopt the output medical text. This allows the doctor to confirm the content of the medical text. Furthermore, if the receiving unit 102 receives a decision not to adopt the output medical text, it receives a decision to change the medical text generation conditions and generate the medical text. This makes it possible to regenerate the medical text in a way that reflects the doctor's intention to make corrections.

[0039] [Fourth Modification of the First Embodiment] The text generation device 100 of the first embodiment has been described as creating a letter of referral that a doctor writes to the doctor at the referring medical institution when a patient is to receive treatment at a different medical institution. When the referring medical institution accepts the patient, the text generation device 100 may generate a report of the medical procedure performed at the referring medical institution.

[0040] In this case, the acquisition unit 101 acquires the electronic medical record information of the admitted target patient and sentence generation information including the sharing purpose. Specifically, the acquisition unit 101 acquires the electronic medical record information from a server device of the referred medical institution. The acquisition unit 101 also acquires from the user on the screen of an application program for creating medical sentences that the sharing purpose is a medical procedure report. The reception unit 102 then receives instructions from the language model to generate medical sentences including this sentence generation information. The generation unit 103 then inputs the instructions into the language model to generate a report. The output unit 104 outputs the generated report to a display or the like.

[0041] FIG. 6 is an example of a medical text generation screen according to the present disclosure. The example in FIG. 6 is an example of a generated report text displayed. The parts in brackets in FIG. 6, such as [Medical History], [Physical Findings], [Medical History and Physical Findings Revealed at Our Hospital], and [Blood Tests and Imaging Tests], are inserted by applying electronic medical record information. The report text in FIG. 6 is an example, and other texts may be output.

[0042] [Second Embodiment] Next, a second embodiment of the present disclosure will be described in detail with reference to the drawings. Below, descriptions of content that overlaps with the above description will be omitted to the extent that the description of this embodiment is not unclear. As with the computer device shown in FIG. 2, the functions of each component in each embodiment of the present disclosure can be realized not only by hardware but also by a computer device or software based on program control.

[0043] 7 is a block diagram showing the configuration of a text generation device 110 according to the present disclosure. The text generation device 110 will be described with reference to FIG. 7 , focusing on the differences from the text generation device 100. In this embodiment, a case is assumed in which a text detailing symptoms is generated by inserting patient-specific information into a medical text based on electronic medical record information. The text generation device 110 includes an acquisition unit 111, a reception unit 112, a generation unit 113, an output unit 114, an extraction unit 115, and a setting unit 116.

[0044] The extraction unit 115 is a means for extracting problematic terms from the generated medical text. The extraction unit 115 analyzes the generated medical text to identify and extract problematic terms. Examples of problematic terms include discriminatory terms that distinguish between men and women, or inappropriate terms related to the mind, body, or illness. The problematic terms may be stored in, for example, the storage device 505, and the extraction unit 115 may extract these terms by comparing them with terms stored in the storage device 505.

[0045] The output unit 114 outputs the extracted problem terms. For example, the output unit 114 may highlight the extracted problem terms or output the problem terms in a manner that allows the user to correct them on the screen.

[0046] The setting unit 116 is a means for setting not using the extracted problematic term as a condition for generating medical text. For example, the setting unit 116 sets the receiving unit 112 to automatically receive an instruction to generate medical text without using the problematic term when the user performs an operation to generate medical text. Furthermore, the setting unit 116 may set not using the extracted problematic term as a generation condition when the user sets the extracted problematic term as a problematic term.

[0047] The operation of the sentence generation device 110 configured as above will be described with reference to the flowchart of FIG.

[0048] 8 is a flowchart showing an outline of the operation of the sentence generation device 110 according to the present disclosure. Note that the processing according to this flowchart may be executed based on program control by the processor described above. The flow of S111 to S114 is the same as S101 to S104.

[0049] As shown in FIG. 8 , first, the acquisition unit 111 acquires sentence generation information including the electronic medical record information of a target patient whose patient information will be shared with other medical institutions and the purpose of sharing (step S201). Next, the reception unit 112 receives instructions for generating medical sentences, including the sentence generation information (step S202). The generation unit 113 generates medical sentences based on the sentence generation information by inputting the instructions into a language model (step S203). Next, the output unit 114 outputs the generated medical sentences (step S204). Next, the extraction unit 115 extracts problematic terms from the generated medical sentences (step S205). Finally, the setting unit 116 sets the requirement that the extracted problematic terms not be used as a condition for generating medical sentences (step S206). This completes the sentence generation operation of the sentence generation device 110.

[0050] In the sentence generation device 110, the extraction unit 115 extracts problematic terms from the generated sentence. Then, the setting unit 116 sets the requirement for generating medical sentences that the extracted problematic terms not be used. This makes it possible to generate sentences that do not include problematic terms.

[0051] Although the present invention has been described above with reference to the embodiments, the present invention is not limited to the above-described embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

[0052] For example, although multiple operations are described in a sequential order in the form of a flowchart, the order of description does not limit the order in which the multiple operations are performed. Therefore, when implementing each embodiment, the order of the multiple operations can be changed within the scope that does not affect the content.

[0053] In each embodiment of the present disclosure, the generation units 103 and 113 may generate medical sentences using a plurality of different language models, and the output units 104 and 114 may output the medical sentences generated by each of them. In this case, the reception unit 102 may receive a request to adopt the medical sentences generated by one of the language models. As the plurality of language models, for example, a large-scale language model released as open source and a language model using a proprietary engine may be used. In this case, the user can select the better medical sentence.

[0054] A part or all of the above-described embodiments can be described as, but not limited to, the following supplementary notes.

[0055] (Supplementary Note 1) A text generation device comprising: an acquisition means for acquiring text generation information including electronic medical record information of a patient whose patient information is to be shared with other medical institutions and a purpose for sharing; a reception means for receiving instructions for generating medical text including the text generation information for sharing the patient information; a generation means for generating the medical text based on the text generation information by inputting the instructions into a language model; and an output means for outputting the generated medical text.

[0056] (Supplementary Note 2) The text generation device according to Supplementary Note 1, wherein the medical text is an introduction text for introducing the target patient to another medical institution.

[0057] (Appendix 3) The medical text is a report for reporting the medical procedure performed at the referred medical institution when the referred medical institution accepts the target patient.

[0058] (Supplementary Note 4) The text generation device according to any one of Supplementary Notes 1 to 3, wherein the electronic medical record information includes information on medical findings input by a doctor.

[0059] (Appendix 5) The text generation device described in any of Appendices 1 to 4, wherein the reception means receives an instruction to generate a medical text on a screen displaying the patient's electronic medical record information for generating a medical text to share the patient information.

[0060] (Supplementary Note 6) The text generation device according to any one of Supplementary Notes 1 to 5, wherein the output means displays the generated medical text in an editable manner.

[0061] (Supplementary Note 7) The text generation device according to any one of Supplementary Notes 1 to 6, wherein the output means displays electronic medical record information that is the basis for the information included in the medical text.

[0062] (Supplementary Note 8) The text generation device according to any one of Supplementary Notes 1 to 7, wherein the output means displays annotations for technical terms included in the medical text.

[0063] (Appendix 9) The text generation device described in any of Appendices 1 to 8, wherein the acquisition means further acquires information about the medical institution to which the patient is referred as the text generation information, and the generation means generates a medical text addressed to the medical institution using the language model.

[0064] (Appendix 10) The text generation device described in any of Appendices 1 to 9, wherein the receiving means further receives instructions including conditions for generating medical text for the target patient, and the generating means generates medical text for sharing patient information based on the generation conditions.

[0065] (Appendix 11) The text generation device described in Appendix 10, wherein the acquisition means further acquires a template text of a medical text prepared in advance, the reception means receives an instruction to generate the medical text using the template text as the generation condition, and the generation means inputs the instruction and the template text into a language model and generates medical text for sharing the patient information by applying electronic medical record information to the template text.

[0066] (Appendix 12) The text generation device described in Appendix 10, wherein the acquisition means further acquires sample sentences of medical texts prepared in advance, the reception means receives instructions for generating the medical text using the sample sentences as the generation conditions, and the generation means generates the medical text by inputting the instructions and the sample sentences into a language model.

[0067] (Supplementary Note 13) The text generation device according to any one of Supplementary Notes 1 to 12, wherein the accepting means further accepts whether or not to adopt the output medical text.

[0068] (Appendix 14) The text generation device described in Appendix 13, wherein the generation means generates medical text using a plurality of different language models, the output means outputs the medical text generated, and the reception means receives a request as to which language model the text generated using should be adopted.

[0069] (Appendix 15) The text generation device described in Appendix 13, wherein, when the acceptance means accepts that the medical text will not be adopted, the acceptance means further accepts that the medical text will be regenerated by changing the generation conditions, and the generation means regenerates the medical text under the generation conditions.

[0070] (Appendix 16) The text generation device described in Appendix 15, wherein the receiving means further receives at least one of the length of the text, the degree of summary, and the electronic medical record information to be included in the medical text as generation conditions for regenerating the medical text, and the generation means regenerates the medical text using the generation conditions.

[0071] (Appendix 17) A text generation device described in any of Appendices 1 to 16, further comprising: an extraction means for extracting problematic terms from the generated medical text; and a setting means for setting the condition for generating medical text to not use the extracted problematic terms.

[0072] (Supplementary Note 18) The text generation device according to any one of Supplementary Notes 1 to 17, wherein the language model is a model optimized by machine learning.

[0073] (Supplementary Note 19) A text generation method in which a computer acquires text generation information including electronic medical record information of a patient whose patient information is to be shared with other medical institutions and a purpose of sharing, receives an instruction for generating a medical text including the text generation information for sharing the patient information, inputs the instruction into a language model, thereby generating the medical text based on the text generation information, and outputs the generated medical text.

[0074] (Appendix 20) A recording medium storing a program that causes a computer to execute the following steps: acquire electronic medical record information of a patient whose patient information is to be shared with other medical institutions and sentence generation information including the purpose of sharing; receive instructions for generating medical sentences including the sentence generation information for sharing the patient information; generate the medical sentences based on the sentence generation information by inputting the instructions into a language model; and output the generated medical sentences.

[0075] Some or all of the configurations described in Supplements 2 to 18, which are dependent on Supplement 1 described above, may also be dependent on Supplements 19 and 20 in the same dependency relationship as Supplements 2 to 18. Not limited to Supplements 1, 19, and 20, some or all of the configurations described as Supplements may also be dependent on various hardware, software, various recording devices for recording software, or systems, within the scope of each of the above-mentioned embodiments.

[0076] 100, 110 Text generation device 101, 111 Acquisition unit 102, 112 Reception unit 103, 113 Generation unit 104, 114 Output unit 115 Extraction unit 116 Setting unit 500 Computer device 501 CPU 502 ROM 503 RAM 504 Program 505 Storage device 506 Recording medium 507 Drive device 508 Communication interface 511 Input / output interface 512 Bus

Claims

1. A means for acquiring electronic medical record information and text generation information including the purpose of sharing for patients who are to be shared with other medical institutions. A receiving means that receives instructions for generating medical documents for sharing patient information, which include the aforementioned document generation information, A generation means that generates the medical document based on the document generation information by inputting the aforementioned instructions into a language model, A document generation device comprising output means for outputting the generated medical document.

2. The document generation device according to claim 1, wherein the medical document is a referral document for referring the patient to another medical institution.

3. The output means displays annotations for technical terms contained in the medical document, as described in claim 1.

4. The document generation apparatus according to claim 1, wherein the receiving means further receives whether or not to adopt the outputted medical document.

5. The generation means generates medical texts using multiple different language models, The output means outputs the generated medical documents, The text generation apparatus according to claim 4, wherein the receiving means receives a request to decide which language model generated the text to adopt.

6. If the aforementioned receiving means receives a request not to adopt the request, it will further receive a request to regenerate the medical document with different generation conditions. The document generation apparatus according to claim 4, wherein the generation means regenerates the medical document under the generation conditions.

7. The aforementioned receiving means further receives, as generation conditions for regenerating the medical document, at least one of the length of the document, the degree of summary, and the electronic medical record information to be included in the medical document. The document generation apparatus according to claim 6, wherein the generation means regenerates medical documents under the generation conditions.

8. An extraction means for extracting problematic terms from the generated medical text, The text generation apparatus according to claim 1, further comprising: setting means for setting that the extracted problem terms should not be used as a condition for generating medical text.

9. Computers We acquire electronic medical record information and text generation information including the purpose of sharing for patients who are to be shared with other medical institutions. The system receives instructions to generate medical documents for sharing patient information, which include the aforementioned document generation information. By inputting the above instructions into the language model, the medical document is generated based on the document generation information. A document generation method that outputs the generated medical document.

10. We acquire electronic medical record information and text generation information including the purpose of sharing for patients who are to be shared with other medical institutions. The system receives instructions to generate medical documents for sharing patient information, which include the aforementioned document generation information. By inputting the above instructions into the language model, the medical document is generated based on the document generation information. A computer program that causes a computer to output the generated medical document.