Program, control method, and electronic medical record system

Abstract

The present invention provides a program, control method, and electronic medical record system that enable physicians to obtain more appropriate summaries of patient examinations. [Solution] The program controls the system that provides electronic medical records. This program causes the computer to perform the following processes: acquiring audio information representing the voice of a doctor during a patient's examination; acquiring patient information from the electronic medical record; and inputting the audio information and patient information into a generating AI (Artificial Intelligence) model and acquiring summary information representing a summary of the examination from the generating AI model.

Description

[Technical Field]

[0001] The present invention relates to a program, a control method, and an electronic medical record system. [Background Art]

[0002] Japanese Patent Laid-Open No. 2005-149083 (Patent Document 1) discloses an electronic medical record input support device. In this electronic medical record input support device, medical information is input by voice. Medical terms are extracted based on the input voice, and the extracted medical terms are output in association with entry items for the electronic medical record. According to this electronic medical record input support device, an electronic medical record can be efficiently created based on the output result. [Prior Art Documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Patent Laid-Open No. 2005-149083 [Summary of the Invention] [Problem to be Solved by the Invention]

[0004] For an electronic medical record, for example, the content of a medical examination performed on a patient needs to be described. If an appropriate summary of the conversation between a doctor and a patient during a medical examination, that is, an appropriate summary of a doctor's examination of a patient, can be obtained in advance, it becomes easier to describe the examination content in the electronic medical record. Patent Document 1 does not disclose a technique for acquiring an appropriate summary of a doctor's examination of a patient.

[0005] The present invention has been made to solve such a problem, and an object of the present invention is to provide a program, a control method, and an electronic medical record system that can acquire a more appropriate summary of a doctor's examination of a patient. [Means for Solving the Problem]

[0006] A program according to a certain aspect of the present invention controls a system for managing electronic medical records. This program causes a computer to perform the following processes: acquiring audio information representing the voice of a doctor during a patient's examination; acquiring patient information from the electronic medical record; and inputting the audio information and patient information into a generating AI (Artificial Intelligence) model and acquiring summary information representing a summary of the examination from the generating AI model.

[0007] When this program is executed by a computer, voice information and patient information are input into a generating AI model, and summary information showing a summary of the examination is obtained from the generating AI model. Therefore, with this program, patient information is also input into the generating AI model in addition to voice information, so more appropriate summary information that takes patient information into account can be obtained from the generating AI model.

[0008] The above program may further have the computer perform processes such as displaying a list of patient information obtainable from the electronic medical record and receiving a selection of patient information from the physician from the list of patient information, and the patient information input into the generated AI model may be the patient information selected by the physician.

[0009] In this program, the patient information input to the generating AI model is patient information selected by the physician. Therefore, this program allows the generating AI model to generate summary information based on the patient information that the physician deems necessary.

[0010] In the above program, patient information may include first patient information relating to the patient's unique characteristics and second patient information relating to the patient's health status that changes over time, and the second patient information input to the generating AI model may be restricted compared to the first patient information.

[0011] In this program, the second patient information input into the generating AI model is restricted compared to the first patient information. Therefore, according to this program, for example, second patient information that has little impact on the current examination is not input into the generating AI model, thus reducing the possibility of generating unexpectedly low-accuracy summary information.

[0012] The above program may further cause the computer to perform processes such as displaying a list of medical dictionaries and receiving a selection of dictionaries from the doctor, and the dictionaries selected by the doctor may be further input into the generating AI model.

[0013] When this program is run by a computer, the dictionary selected by the doctor is further input into the generating AI model. Therefore, according to this program, a more appropriate dictionary is input into the generating AI model, resulting in the acquisition of more accurate summary information.

[0014] The above program may also have the computer perform a process to receive supplementary information from the physician, and this supplementary information may be further input into the generating AI model.

[0015] When this program is executed by a computer, supplementary information is further input into the generating AI model. Therefore, according to this program, the summarization accuracy of the generating AI model is improved by the input of appropriate supplementary information by a physician, resulting in the acquisition of more accurate summary information.

[0016] In the above program, in addition to the dictionary selected by the physician, a dictionary of general medical terminology may also be input into the generative AI model, regardless of the physician's selection.

[0017] When this program is executed by a computer, a dictionary of general medical terminology is added to the generative AI model, in addition to the dictionary selected by the physician, regardless of the physician's selection. Therefore, because the dictionary of general medical terminology is added to the generative AI model regardless of the physician's selection, it is possible to obtain summary information with a certain degree of accuracy.

[0018] The above program may also cause the computer to perform the following processes: storing audio information acquired during the examination, inputting the stored audio information and patient information back into a generating AI model, and then obtaining summary information again from the generating AI model.

[0019] When this program is executed by a computer, the stored voice information and patient information are re-input into the generating AI model, and the summaries are retrieved again from the generating AI model. Therefore, according to this program, after the doctor has reviewed the initially retrieved summaries, the summaries can be regenerated by the generating AI model as needed.

[0020] The above program may also have the computer perform the processes of receiving corrections to the summary information from doctors and sending instructions to the generating AI model to reflect the corrections in the dictionary.

[0021] When this program is executed by a computer, an instruction is sent to the generating AI model to reflect the changes in the dictionary. Therefore, this program reduces the effort required to revise the dictionary because the changes made by doctors to the summary information are reflected in the dictionary.

[0022] A control method according to another aspect of the present invention controls a system for managing electronic medical records. This control method includes a computer acquiring audio information indicating the voice of a doctor during a patient's examination, acquiring patient information about the patient from the electronic medical record, inputting the audio information and patient information into a generating AI model, and acquiring summary information indicating a summary of the examination from the generating AI model.

[0023] In this control method, voice information and patient information are input to a generative AI model, and summary information indicating a summary of the medical consultation is acquired from the generative AI model. Therefore, according to this control method, since patient information is input to the generative AI model in addition to voice information, more appropriate summary information that also takes patient information into consideration can be acquired from the generative AI model.

[0024] An electronic medical record system according to another aspect of the present invention manages electronic medical records. This electronic medical record system includes a first acquisition unit, a second acquisition unit, and a third acquisition unit. The first acquisition unit acquires voice information indicating voice recorded during a patient's medical examination by a physician. The second acquisition unit acquires patient information relating to the patient from an electronic medical record. The third acquisition unit inputs the voice information and the patient information to a generative AI model, and acquires summary information indicating a summary of the medical consultation from the generative AI model.

[0025] In this electronic medical record system, voice information and patient information are input to a generative AI model, and summary information indicating a summary of the medical consultation is acquired from the generative AI model. Therefore, according to this electronic medical record system, since patient information is input to the generative AI model in addition to voice information, more appropriate summary information that also takes patient information into consideration can be acquired from the generative AI model. Effects of the Invention

[0026] According to the present invention, it is possible to provide a program, a control method, and an electronic medical record system that can acquire a more appropriate summary of a patient's medical examination by a physician. Brief Description of the Drawings

[0027] [Figure 1] It is a diagram schematically showing the configuration of the electronic medical record system. [Figure 2] It is a block diagram schematically showing the hardware configuration of an electronic medical record server. [Figure 3] It is a diagram schematically showing a medical record information database. [Figure 4]This is a schematic diagram illustrating the voice information database. [Figure 5] This is a schematic diagram illustrating a dictionary database. [Figure 6] This is a schematic diagram illustrating the prompt database. [Figure 7] This is a block diagram schematically showing the hardware configuration of a user terminal. [Figure 8] This is a flowchart showing the procedure for obtaining a summary of the medical examination results. [Figure 9] This diagram schematically shows an example of the main screen of an electronic medical record system. [Figure 10] This diagram schematically shows an example of a screen that appears when the button to activate the AI ​​summarization function is pressed. [Figure 11] This diagram schematically shows an example of a patient information selection screen. [Figure 12] This diagram schematically shows an example of a screen displayed during the recording of a medical consultation. [Figure 13] This diagram schematically shows an example of a screen displaying summary information. [Figure 14] This is a flowchart showing the procedure for extracting patient information. [Figure 15] This is a flowchart showing the procedure for submitting revision proposals. [Modes for carrying out the invention]

[0028] Hereinafter, an embodiment relating to one aspect of the present invention (hereinafter also referred to as "this embodiment") will be described in detail with reference to the drawings. In the drawings, the same or corresponding parts are denoted by the same reference numerals, and their descriptions will not be repeated. Furthermore, each drawing is schematically depicted with parts omitted or exaggerated as appropriate for ease of understanding.

[0029] [1. Overview of the Electronic Medical Record System] Figure 1 is a schematic diagram showing the configuration of an electronic medical record system 10 according to this embodiment. Referring to Figure 1, the electronic medical record system 10 is a system for managing electronic medical records. The electronic medical record system 10 includes an electronic medical record server 100, a plurality of user terminals 200, and an AI server 300. Each of the electronic medical record server 100 and the AI ​​server 300 is implemented by, for example, a general-purpose computer, and each of the plurality of user terminals 200 is implemented by, for example, a PC (Personal Computer), a tablet, or a smartphone. In the electronic medical record system 10, the electronic medical record server 100, the plurality of user terminals 200, and the AI ​​server 300 are configured to communicate with each other via the Internet N1.

[0030] The services provided by the electronic medical record system 10 are implemented, for example, in medical institutions such as clinics or hospitals. Users of this service are, for example, physicians working at the medical institution that has implemented the service. This service is provided to each user, for example, through an application or browser installed on a user terminal 200.

[0031] The electronic medical record server 100 manages, for example, the electronic medical records of each patient for each medical institution. Each user accesses the electronic medical record server 100 through a user terminal 200 during a patient's examination and edits the electronic medical record of the patient being examined. The electronic medical record includes, for example, basic information about the patient (address, name, gender, and age, etc.), disease / symptoms (disease name and main symptoms, etc.), treatment details (treatment method, prescription and procedures, etc.), and the date of the examination.

[0032] The AI ​​server 300 provides a generative AI model 310. The generative AI model 310 is, for example, composed of a large-scale language model and is built through learning using information periodically acquired from various sites on the Internet N1. The generative AI model 310 is configured to, for example, accept natural language input and generate various responses corresponding to that natural language based on common sense. Examples of generative AI models 310 include, for example, GPT-5.5 and Gemini 3.0.

[0033] The electronic medical records managed by the electronic medical record server 100 must contain information about the medical examinations performed on the patient. If an appropriate summary of the conversation between the doctor and the patient during the examination—that is, an appropriate summary of the doctor's examination of the patient—is available in advance, it will be easier to record the examination details in the electronic medical record.

[0034] In the electronic medical record system 10, a doctor examines a patient in front of a user terminal 200, and the audio recording of the examination is captured by the user terminal 200. Based on the captured audio, a summary of the doctor's examination is generated by the generative AI model 310. This allows the doctor to relatively easily record the examination details in the electronic medical record based on the summary generated by the generative AI model 310. In particular, the electronic medical record system 10 incorporates measures to ensure relatively high accuracy in the summary generated by the generative AI model 310. The electronic medical record system 10 will be described in detail below.

[0035] [2. Structure] <2-1. Configuration of the Electronic Medical Record Server> Figure 2 is a schematic block diagram showing the hardware configuration of the electronic medical record server 100. As shown in Figure 2, the electronic medical record server 100 includes a control unit 110, a communication interface 130, and a storage unit 120. Each component is electrically connected via a bus.

[0036] The control unit 110 includes a CPU (Central Processing Unit) 112, RAM (Random Access Memory) 114, and ROM (Read Only Memory) 116, and is configured to control each component according to information processing.

[0037] The communication interface 130 is configured to communicate with the user terminal 200 and the AI ​​server 300 (see Figure 1) via the internet N1. The communication interface 130 consists of, for example, a wired LAN (Local Area Network) module or a wireless LAN module.

[0038] The storage unit 120 is composed of, for example, an auxiliary storage device such as a hard disk drive or a solid-state drive. The storage unit 120 stores, for example, a control program 121, a medical record information database 122, a voice information database 123, a dictionary database 124, and a prompt database 125. Various functions of the electronic medical record server 100 are realized when the control program 121 is executed by the CPU 112 (processor).

[0039] Figure 3 is a schematic diagram of the medical record information DB122. Referring to Figure 3, in the medical record information DB122, patient information (medical record information) for each patient is managed for each medical institution. There are two types of patient information: first patient information and second patient information. First patient information is information about the unique characteristics of the patient. Examples of first patient information include medical history, allergies, contraindicated drugs, and history of side effects. Second patient information is information about the patient's health status that changes over time. Examples of second patient information include test results, disease name, chief complaint / findings, treatment / procedures, regularly used medications, and information from the medical interview. For example, regarding second patient information, the content of each item is managed in chronological order.

[0040] Figure 4 is a schematic diagram of the voice information DB 123. Referring to Figure 4, in the electronic medical record system 10, audio from a doctor's examination of a patient (for example, a conversation between the doctor and the patient) is collected by the user terminal 200, and an audio file is generated in the user terminal 200. The audio file generated in the user terminal 200 is sent to the electronic medical record server 100 along with information about the patient being examined (for example, patient ID (identifier) ​​and patient name), etc., and is stored in the storage unit 120 and managed in the voice information DB 123. In the voice information DB 123, received audio files are managed for each medical institution. In the voice information DB 123, for example, patient ID, patient name, examination date and time, and information about the storage location of the audio file in the storage unit 120 are managed in association with each other. Note that the voice information managed in the voice information DB 123 does not necessarily have to be an audio file that stores the audio itself. The audio collected by the user terminal 200 may be converted to text, and a text file showing the converted text may be managed as voice information in the voice information DB 123.

[0041] Figure 5 is a schematic diagram of the dictionary DB124. Referring to Figure 5, the electronic medical record system 10 has a dictionary of general medical terms (hereinafter also referred to as the "general medical term dictionary") provided by default, and the general medical term dictionary (not shown) is stored in the storage unit 120 of the electronic medical record server 100. In the general medical term dictionary, for example, the reading, notation, category, and registration date of medical terms are managed in correspondence with each other. The "category" indicates the part of speech of the medical term, and the "registration date" indicates the date on which the medical term was registered in the general medical term dictionary. The general medical term dictionary is updated, for example, by the electronic medical record service provider. The general medical term dictionary is sent to the generation AI model 310 and used in the generation AI model 310 for summarizing the examination.

[0042] In the electronic medical record system 10, in addition to the general medical terminology dictionary, each user can register additional dictionaries to be passed to the generation AI model 310. These dictionaries include, for example, medical terms other than general medical terms (e.g., in-hospital terminology, doctor-specific terminology). In the dictionary database 124, registered dictionaries are managed for each medical institution. In the dictionary database 124, the pronunciation, notation, category, and registration date of medical terms are managed in correspondence with each other.

[0043] Figure 6 is a schematic diagram of the prompt DB 125. Referring to Figure 6, the electronic medical record system 10 has one or more types of prompts provided by default, each instructing the AI ​​model 310 to generate a summary of the examination content, and these prompts are stored in the storage unit 120 of the electronic medical record server 100. In addition to the prompts provided by default, the electronic medical record system 10 allows users to register additional prompts that instruct the AI ​​model 310 to generate a summary of the examination content. In the prompt DB 125, each registered prompt is managed for each medical institution. In the prompt DB 125, the prompt ID, prompt name, and prompt content are managed in correspondence with each other.

[0044] <2-2. User Terminal Configuration> Figure 7 is a schematic block diagram showing the hardware configuration of the user terminal 200. As shown in Figure 7, the user terminal 200 includes a control unit 210, a communication interface 230, an operation unit 240, a display 250, a microphone 260, and a storage unit 220. In the user terminal 200, each component is electrically connected via a bus.

[0045] The control unit 210 includes a CPU, RAM, ROM, etc., and is configured to control each component according to information processing. The communication interface 230 is configured to communicate with the electronic medical record server 100 and the AI ​​server 300 via the internet N1. The communication interface 230 is composed of, for example, a wired LAN module or a wireless LAN module.

[0046] The control unit 240 is configured to accept input from the user. The control unit 240 is composed of, for example, a touch panel, a keyboard, and part or all of a mouse. The display 250 is configured to display various screens. The display 250 is composed of, for example, a liquid crystal monitor or an organic EL (Electro Luminescence) monitor. The display 250 displays, for example, various screens provided by the electronic medical record server 100. The microphone 260 is configured to collect sound and generate an audio file. The microphone 260 collects, for example, a conversation between a doctor and a patient during a medical examination.

[0047] The storage unit 220 is, for example, an auxiliary storage device such as a hard disk drive or a solid-state drive. The storage unit 220 stores, for example, a control program 222. Various functions in the user terminal 200 are realized when the control program 222 is executed by the CPU (processor) of the control unit 210.

[0048] [3. Improving the accuracy of summaries of medical examination results] As mentioned above, electronic medical records need to include, for example, the details of the examination performed on the patient. However, summarizing the examination results during or after the examination and entering those results into the electronic medical record requires a certain amount of effort. This effort can be reduced if an appropriate summary of the conversation between the doctor and the patient during the examination, that is, an appropriate summary of the doctor's examination of the patient, is obtained in advance. In the electronic medical record system 10, the generating AI model 310 generates a summary of the examination based on the audio collected during the examination. Therefore, users of the electronic medical record system 10 can relatively easily record the details of the examination in the electronic medical record based on the summary generated by the generating AI model 310.

[0049] In the electronic medical record system 10, the conversation between the doctor and the patient during the examination is recorded by the user terminal 200. The audio file generated through the recording is sent from the user terminal 200 to the electronic medical record server 100. The electronic medical record server 100 sends the received audio file to the generating AI model 310, causing the generating AI model 310 to generate summary information that shows a summary of the examination. Furthermore, as will be described in detail later, in order to cause the generating AI model 310 to generate summary information, the electronic medical record server 100 sends at least a portion of the patient information concerning the patient being examined to the generating AI model 310 in addition to the audio file. That is, the audio file and patient information are input to the generating AI model 310, and summary information is obtained from the generating AI model 310. According to the electronic medical record system 10, since patient information is also input to the generating AI model 310 in addition to the audio file, more appropriate summary information that takes patient information into account can be obtained from the generating AI model 310. Note that the audio information sent from the electronic medical record server 100 to the generating AI model 310 does not necessarily have to be an audio file. If the audio collected at the user terminal 200 is converted to text, and the text file representing the converted text is managed as audio information in the audio information DB 123, the text file may be sent from the electronic medical record server 100 to the generated AI model 310 instead of the audio file.

[0050] In the electronic medical record system 10, the electronic medical record server 100 manages the electronic medical records that hold each patient's information, and the electronic medical record server 100 initiates the instruction to the generating AI model 310 to summarize the examination content. Since the electronic medical record server 100, which is the entity that issues the instruction to the generating AI model 310 for summarization, manages the electronic medical records that hold each patient's information, it is possible to easily send patient information to the generating AI model 310 in addition to the audio file when issuing the instruction. If the entity that issues the instruction to the generating AI model 310 for summarization and the entity that manages the electronic medical records are different, it is not easy to send patient information to the generating AI model 310 in addition to the audio file.

[0051] [4. Operation] Figure 8 is a flowchart illustrating the procedure for obtaining a summary of the medical examination results. Referring to Figure 8, the processes shown in the left flowchart are executed by the control unit 210 of the user terminal 200 when the electronic medical record is open on the user terminal 200. The processes shown in the right flowchart are executed by the control unit 110 of the electronic medical record server 100.

[0052] Referring to the flowchart on the left side of Figure 8, the control unit 210 of the user terminal 200 determines whether or not it has received a command from the user to start recording audio during the examination (step S100). The command to start recording is received, for example, through the electronic medical record screen. If it is determined that no command to start recording has been received (NO in step S100), the control unit 210 waits until it receives a command to start recording.

[0053] Figure 9 is a schematic diagram showing an example of the main screen of an electronic medical record system. Referring to Figure 9, screen IM1 is displayed on the display 250 of the user terminal 200. Screen IM1 is the main screen of the electronic medical record system. For example, area TR1 contains the patient's chief complaint and findings determined by the physician during the examination, and area TR2 contains the procedures and actions performed by the physician on the patient during the examination. Screen IM1 is provided with a button BT1 for activating a function that causes the generating AI model 310 to generate a summary of the examination content (hereinafter also referred to as the "AI summary function"). When button BT1 is pressed, the AI ​​summary function is activated.

[0054] Figure 10 schematically shows an example of a screen displayed when the button (button BT1) for activating the AI ​​summarization function is pressed. Referring to Figure 10, screen IM2 is displayed on the display 250 of the user terminal 200. On screen IM2, a slide-in panel PN1 is displayed on the main screen. The slide-in panel PN1 slides in from the right when, for example, button BT1 (see Figure 9) is pressed.

[0055] The slide-in panel PN1 is equipped with a pull-down button PD1. When the pull-down button PD1 is pressed, a pull-down menu is displayed showing a list of prompts for summarizing the examination content. The prompts displayed in the pull-down menu are, for example, prompts managed in prompt DB125. The user can select the desired prompt from the multiple prompts included in the pull-down menu. Based on the prompt selected by the user, the examination content is summarized. The slide-in panel PN1 is also equipped with a button BT2. Button BT2 is used to instruct the start of recording the examination content. When button BT2 is pressed, the instruction to start recording is received.

[0056] Referring again to the flowchart on the left side of Figure 8, if it is determined that a recording start instruction has been received in step S100 (YES in step S100), the control unit 210 controls the display 250 to display a patient information selection screen (hereinafter also referred to as the "patient information selection screen") (step S110). The patient information selection screen is a screen for inputting patient information, dictionary information, and supplementary information to be sent to the generating AI model 310 for summarizing the examination content, for example.

[0057] Figure 11 is a schematic diagram showing an example of a patient information selection screen. Referring to Figure 11, the patient information selection screen is displayed, for example, as a slide-in panel PN1. That is, when the button BT2 (see Figure 10) that instructs the start of recording is pressed on the slide-in panel PN1, the display content on the slide-in panel PN1 switches to the patient information selection screen.

[0058] On the patient information selection screen, checkbox CB1 is associated with each of several types of patient information. The user selects the checkbox CB1 associated with the patient information they wish to use for summarizing the examination content. The user can select one or more types of patient information. The selected patient information is used for summarizing the examination content. For example, in addition to the currently selected prompt, a system prompt is sent to the generating AI model 310, and this system prompt contains instructions to summarize the examination content using the selected patient information.

[0059] Furthermore, a pull-down button PD2 is provided on the patient information selection screen. When the pull-down button PD2 is pressed, a pull-down menu showing a list of dictionaries is displayed. The dictionaries displayed in the pull-down menu are, for example, dictionaries managed in dictionary DB124. The user can select one or more dictionaries from the one or more dictionaries included in the pull-down menu. The selected dictionaries are used to summarize the examination results. For example, the system prompt includes instructions to summarize the examination results using the selected dictionaries.

[0060] Furthermore, the patient information selection screen includes area TR4 for inputting supplementary instructions regarding the summary of the examination content. Users can input information that was not fully explained in the selected prompt into area TR4. For example, the system prompt includes instructions to summarize the examination content based on the supplementary instructions entered in area TR4. In other words, a prompt that includes the content of the supplementary instructions entered in area TR4 and the content of the selected prompt becomes a prompt that instructs the generating AI model 310 to create a summary. For example, by entering supplementary instructions in area TR4, a physician can have the generating AI model 310 create a summary of the examination content that more accurately reflects the physician's intentions, which would not be possible with only the selected prompt.

[0061] On the patient information selection screen, button BT6 is provided. When button BT6 is pressed, the input content regarding patient information, dictionary information, and supplementary instructions is saved, and the recording of the consultation begins.

[0062] Referring again to the flowchart on the left side of Figure 8, when the patient information selection screen is displayed, the control unit 210 determines whether the selection and input of patient information has been completed (step S120). The selection and input of patient information is determined to be completed when button BT6 (see Figure 11) is pressed. If it is determined that the selection and input of patient information has not been completed (NO in step S120), the control unit 210 waits until the selection and input of patient information is completed. The selection result of patient information on the patient information selection screen is saved as the selection result for that patient and may be used at the next consultation of that patient. In this case, if the previous selection result is saved, for example, the previous selection result may be entered in advance when the slide-in panel PN1 is displayed, or if a recording start instruction is received in step S100, the processing in steps S110 and S120 may be omitted and the processing in step S130 may be executed.

[0063] On the other hand, if it is determined in step S120 that the selection and input of patient information, etc., has been completed (YES in step S120), the control unit 210 starts recording and transcribing the audio (audio of the medical examination) collected through the microphone 260 (step S130).

[0064] Figure 12 schematically shows an example of a screen displayed during recording of a medical examination. Referring to Figure 12, the display 250 of the user terminal 200 shows screen IM3. On screen IM3, a slide-in panel PN1 is displayed on the main screen. The slide-in panel PN1 has an area TR3 where the collected audio is sequentially transcribed. For example, if the cursor (not shown) is moved away from the slide-in panel PN1 during recording, the slide-in panel PN1 slides out to the right after a predetermined time has elapsed. Then, for example, if the cursor is placed at the far right, the slide-in panel PN1 slides in again to the left. The slide-in panel PN1 has a button BT4 for pausing recording and a button BT3 for ending recording. When button BT4 is pressed, recording is paused, and when button BT3 is pressed, recording is stopped and the summarization process of the medical examination content begins.

[0065] Referring again to the flowchart on the left side of Figure 8, once recording and transcription begin in step S130, the control unit 210 determines whether or not it has received a recording termination command from the user (step S140). As described above, a recording termination command is received when button BT3 (see Figure 12) is pressed. If it is determined that a recording termination command has not been received from the user (NO in step S140), the control unit 210 waits until it receives a recording termination command.

[0066] On the other hand, when it is determined that the user has given an instruction to end the recording (YES in step S140), the control unit 210 generates an audio file based on the collected audio and controls the communication interface 230 to send information about the patient being examined (patient ID, patient name, etc.), the generated audio file, information indicating the selection result of the prompt on the pull-down button PD1, and each piece of information entered on the patient information selection screen (selection result regarding patient information, selection result regarding the dictionary, and entered supplementary information) to the electronic medical record server 100 (step S150).

[0067] Subsequently, the control unit 210 determines whether or not it has received summary information indicating a summary of the medical examination from the electronic medical record server 100 (step S160). If it is determined that summary information has not been received (NO in step S160), the control unit 210 waits until it receives the summary information. On the other hand, if it is determined that summary information has been received (YES in step S160), the control unit 210 controls the display 250 to display the received summary information (step S170).

[0068] Figure 13 is a schematic diagram showing an example of a screen displaying summary information. Referring to Figure 13, the display 250 of the user terminal 200 shows screen IM4. On screen IM4, a slide-in panel PN1 is displayed on the main screen. The slide-in panel PN1 has an area TR5 where summary information received from the electronic medical record server 100 is displayed. The slide-in panel PN1 is provided with a button BT7. When button BT7 is pressed, the summary displayed in area TR5 is transferred to area TR1. The user can, for example, modify the summary transferred to area TR1 as needed and record the appropriate content in the electronic medical record.

[0069] Furthermore, the slide-in panel PN1 is equipped with a button BT5. Button BT5 becomes pressable when summary information is acquired. Before summary information is acquired, button BT5 is not pressable (see Figures 10 and 12). When button BT5 is pressed, summary information is regenerated based on the recorded audio file. The recorded audio file is managed, for example, in the audio information DB 123. When button BT5 is pressed, the patient information selection screen (see Figure 11) is displayed again. Then, the prompt selected by the pull-down button PD1, the information entered on the patient information selection screen, and the recorded audio file are sent from the electronic medical record server 100 to the generation AI model 310, and summary information is generated again. The generated summary information is sent from the electronic medical record server 100 to the user terminal 200. In this way, according to the electronic medical record system 10, after the user confirms the initially acquired summary information, the summary information can be regenerated by the generation AI model 310 as needed.

[0070] Referring to the flowchart on the right side of Figure 8, the control unit 110 of the electronic medical record server 100 determines whether or not it has received various information from the user terminal 200 (information about the patient being examined (patient ID, patient name, etc.), generated audio files, information indicating the selection result of the prompt on the pull-down button PD1, and various pieces of information entered on the patient information selection screen (selection result regarding patient information, selection result regarding the dictionary, and entered supplementary information)) (step S200). If it is determined that various information has not been received (NO in step S200), the control unit 110 waits until it receives various information. On the other hand, if it is determined that various information has been received (YES in step S200), the control unit 110 executes the patient information extraction process based on the selection result regarding the received patient information (step S210).

[0071] Figure 14 is a flowchart showing the procedure for extracting patient information. The process shown in this flowchart is executed by the control unit 110 of the electronic medical record server 100 in step S210 of Figure 8.

[0072] Referring to Figure 14, the control unit 110 of the electronic medical record server 100 extracts all information regarding the first patient information from the selected patient information (step S300). The control unit 110 then extracts only the information from the second patient information from the selected patient information, going back from the present to a predetermined period (step S310). The predetermined period may be a time period such as 3 months or 1 year, or a frequency period such as the most recent 1 time or the most recent 3 times. In addition, different predetermined periods may be set for each type of second patient information. Thus, in the electronic medical record system 10 according to this embodiment, the second patient information to be transmitted to the generation AI model 310 is restricted compared to the first patient information. Note that the restriction on the second patient information compared to the first patient information is not necessarily limited to a time period. The restriction on the second patient information compared to the first patient information may be, for example, a restriction on the number of items.

[0073] Referring again to the flowchart on the right side of Figure 8, once the patient information extraction process is complete, the control unit 110 controls the communication interface 130 to send the extracted patient information, received audio files, selected dictionaries, selected prompts, a general medical terminology dictionary, and system prompts to the generating AI model 310 (step S220).

[0074] The control unit 110 determines whether or not it has received summary information from the generation AI model 310 (step S230). If it is determined that summary information has not been received (NO in step S230), the control unit 110 waits until it receives the summary information. On the other hand, if it is determined that summary information has been received (YES in step S230), the control unit 110 controls the communication interface 130 to send the summary information to the user terminal 200 (step S240).

[0075] Thus, in the electronic medical record system 10, audio files and patient information are input into the generating AI model 310, and summary information is obtained from the generating AI model 310. Therefore, according to the electronic medical record system 10, since patient information is also input into the generating AI model 310 in addition to the audio files, more appropriate summary information that takes patient information into account can be obtained from the generating AI model 310.

[0076] Furthermore, in the electronic medical record system 10, the patient information input to the generation AI model 310 is patient information selected by the user (physician). Therefore, according to the electronic medical record system 10, the generation AI model 310 can generate summary information based on the patient information that the user deems necessary.

[0077] Furthermore, in the electronic medical record system 10, the second patient information input to the generating AI model 310 is restricted compared to the first patient information input to the generating AI model 310. Therefore, according to the electronic medical record system 10, for example, second patient information that has little impact on the current examination is not input to the generating AI model 310, thus reducing the possibility of generating unexpectedly low-accuracy summary information.

[0078] Furthermore, in the electronic medical record system 10, the dictionary selected by the user is further input into the generation AI model 310. Therefore, according to the electronic medical record system 10, a more appropriate dictionary is input into the generation AI model 310, making it possible to obtain more accurate summary information.

[0079] Furthermore, in the electronic medical record system 10, supplementary information is further input into the generation AI model 310. Therefore, according to the electronic medical record system 10, the summarization accuracy of the generation AI model 310 is improved by the input of appropriate supplementary information by the user, thereby enabling the acquisition of more accurate summary information.

[0080] Furthermore, in the electronic medical record system 10, in addition to the dictionary selected by the user, a general medical terminology dictionary is also input into the generating AI model 310, regardless of the user's selection. Therefore, according to the electronic medical record system 10, since the general medical terminology dictionary is input into the generating AI model 310 regardless of the user's selection, it is possible to obtain summary information with a certain level of accuracy.

[0081] Furthermore, in the electronic medical record system 10, the recorded (stored) audio file and patient information are input again into the generating AI model 310, and summary information is retrieved again from the generating AI model 310. Therefore, according to the electronic medical record system 10, after the physician has reviewed the initially retrieved summary information, the AI ​​model 310 can regenerate the summary information as needed.

[0082] In addition, the electronic medical record system 10 is equipped with a function that, when a summary of the examination content is obtained and the user modifies the summary, suggests revisions to the dictionary used for the summary based on the modified content. For example, when the obtained summary is transcribed to area TR1 (see Figure 13), and the user modifies the entries in area TR1, revisions to the dictionary used for the summary of the examination content are suggested as needed.

[0083] Figure 15 is a flowchart illustrating the procedure for submitting a revision proposal. Referring to Figure 15, the process shown in the left flowchart is executed by the control unit 210 of the user terminal 200 when the electronic medical record created using the summary of the examination content is finalized (for example, when an instruction to save the electronic medical record or to end the examination is received). The process shown in the right flowchart is executed by the control unit 110 of the electronic medical record server 100.

[0084] Referring to the flowchart on the left side of Figure 15, the control unit 210 of the user terminal 200 determines whether the content of the summary information generated by the generation AI model 310 has been modified (step S400). For example, it is determined whether the summary information has been modified after being transferred to the summary information area TR1 (see Figure 13). If it is determined that the content of the summary information has not been modified (NO in step S400), the process shown in this flowchart ends.

[0085] On the other hand, if it is determined that the content of the summary information has been modified (YES in step S400), the control unit 210 controls the communication interface 230 to send the modified summary information (for example, the content of area TR1 (see Figure 13)) to the electronic medical record server 100 (step S410).

[0086] Subsequently, the control unit 210 determines whether or not it has received the results of the review regarding the necessity of dictionary revision from the electronic medical record server 100 (step S420). For example, if the review results indicate that dictionary revision is necessary, the review results include a dictionary revision proposal. If it is determined that the results of the review regarding the necessity of dictionary revision have not been received (NO in step S420), the control unit 210 waits until it receives the results of the review regarding the necessity of dictionary revision.

[0087] On the other hand, if it is determined that the results of the review regarding the necessity of dictionary correction have been received (YES in step S420), the control unit 210 determines whether or not the review results indicate that dictionary correction is necessary (step S430). If it is determined that the review results indicate that dictionary correction is not necessary (NO in step S430), the process shown in this flowchart ends.

[0088] On the other hand, if the review results indicate that dictionary revision is necessary (YES in step S430), the control unit 210 controls the display 250 to display the proposed dictionary revision received from the electronic medical record server 100 (step S440). The control unit 210 then determines whether the user has accepted the revision proposal (step S450). This determination is made, for example, by whether or not the button to accept the revision has been pressed.

[0089] If the proposed revision is not accepted (NO in step S450), the process shown in this flowchart ends. On the other hand, if the proposed revision is accepted (YES in step S450), the control unit 210 controls the communication interface 230 to send instructions for revision according to the proposal to the electronic medical record server 100 (step S460).

[0090] Referring to the flowchart on the right side of Figure 15, the control unit 110 of the electronic medical record server 100 determines whether or not it has received the revised summary information from the user terminal 200 (step S500). If it is determined that the revised summary information has not been received (NO in step S500), the control unit 110 waits until it receives the revised summary information.

[0091] On the other hand, if it is determined that the corrected summary information has been received (YES in step S500), the control unit 110 controls the communication interface 130 to send a prompt instructing the user to consider the necessity of dictionary correction and the correction policy, as well as the corrected summary information received from the user terminal 200 to the generating AI model 310 (step S510).

[0092] The control unit 110 determines whether or not it has received the results of the assessment regarding the need for dictionary correction from the generating AI model 310 (step S520). If it is determined that the results of the assessment regarding the need for dictionary correction have not been received (NO in step S520), the control unit 110 waits until it receives the results of the assessment regarding the need for dictionary correction. On the other hand, if it is determined that the results of the assessment regarding the need for dictionary correction have been received (YES in step S520), the control unit 110 controls the communication I / F 130 to transmit the received assessment results to the user terminal 200 (step S530).

[0093] Subsequently, the control unit 110 determines whether or not it has received instructions from the user terminal 200 to make corrections according to the proposal (step S540). If it is determined that no instructions for correction have been received (NO in step S540), the control unit 110 waits until it receives instructions for correction. On the other hand, if it is determined that instructions for correction have been received (YES in step S540), the control unit 110 controls the communication interface 130 to send instructions to the generation AI model 310 to apply the corrections according to the proposal to the dictionary (step S550).

[0094] The control unit 110 determines whether or not it has received the corrected dictionary from the generated AI model 310 (step S560). If it is determined that the corrected dictionary has not been received (NO in step S560), the control unit 110 waits until it receives the corrected dictionary. On the other hand, if it is determined that the corrected dictionary has been received (YES in step S560), the control unit 110 updates the dictionary DB 124 so that the contents of the received corrected dictionary are reflected (step S570).

[0095] Thus, in the electronic medical record system 10, when a user modifies a summary generated by the generation AI model 310, an instruction is sent to the generation AI model 310 to reflect the modification in at least some of several types of dictionaries. Therefore, according to the electronic medical record system 10, the user's modifications to the summary information are reflected in the dictionaries, thus reducing the effort required to modify the dictionaries.

[0096] [5. Features] As described above, in the electronic medical record system 10, audio files and patient information are input into the generating AI model 310, and summary information showing a summary of the examination is obtained from the generating AI model 310. Therefore, according to the electronic medical record system 10, since patient information is also input into the generating AI model 310 in addition to the audio files, more appropriate summary information that takes patient information into account can be obtained from the generating AI model 310.

[0097] [6. Other Embodiments] The concept of the above embodiment is not limited to the embodiment described above. Below, an example of another embodiment to which the concept of the above embodiment can be applied will be described.

[0098] <6-1> In the above embodiment, only patient information selected by the user was sent to the generating AI model 310. However, the patient information sent to the generating AI model 310 does not necessarily have to be selected by the user. For example, all patient information managed in the electronic medical record for the patient being examined may be sent to the generating AI model 310, and a summary of the examination content may be generated based on all patient information. Alternatively, all patient information managed in the electronic medical record for the patient being examined may be sent to the generating AI model 310, and the generating AI model 310 may select the necessary patient information and generate a summary of the examination content based on the selected patient information.

[0099] <6-2> In the above embodiment, the second patient information transmitted to the generating AI model 310 was limited to second patient information from the present to a predetermined period. However, the second patient information transmitted to the generating AI model 310 does not necessarily have to be limited. All second patient information managed in the electronic medical record for patients currently being examined may be transmitted to the generating AI model 310.

[0100] <6-3> In the above embodiment, the general medical terminology dictionary was sent to the generating AI model 310 regardless of the user's selection. However, the general medical terminology dictionary does not necessarily have to be sent to the generating AI model 310 every time. The general medical terminology dictionary is also subject to selection, and it may be sent to the generating AI model 310 only if the general medical terminology dictionary is selected.

[0101] Embodiments of the present invention have been described illustratively above. That is, a detailed description and accompanying drawings have been disclosed for illustrative purposes. Therefore, some of the components described in the detailed description and accompanying drawings may not be essential for solving the problem. Consequently, the mere fact that these non-essential components are described in the detailed description and accompanying drawings does not mean that they should be immediately assumed to be essential.

[0102] Furthermore, the above embodiments are merely illustrative in every respect of the present invention. The above embodiments can be improved or modified in various ways within the scope of the present invention. For example, at least a part of the configuration of one embodiment may be combined with at least a part of the configuration of any other embodiment. In other words, in carrying out the present invention, specific configurations can be appropriately adopted depending on the embodiment. [Explanation of Symbols]

[0103] 10 Electronic medical record system, 100 Electronic medical record server, 110, 210 Control unit, 112 CPU, 114 RAM, 116 ROM, 120, 220 Storage unit, 121, 221 Control program, 122 Medical record information DB, 123 Voice information DB, 124 Dictionary DB, 125 Prompt DB, 130, 230 Communication I / F, 200 User terminal, 240 Operation unit, 250 Display, 260 Microphone, 300 AI server, 310 Generated AI model, BT1-BT7 Buttons, CB1 Checkbox, IM1-IM4 Screen, N1 Internet, PD1, PD2 Pull-down buttons, PN1 Slide-in panel, TR1-TR5 Area.

Claims

1. A program that controls a system for managing electronic medical records, A process to acquire audio information representing the voice of a doctor during a patient examination, A process for obtaining patient information relating to the aforementioned patient from the electronic medical record, A program that causes a computer to input the aforementioned voice information and the aforementioned patient information into a generating AI (Artificial Intelligence) model, and to perform the process of obtaining summary information representing a summary of the examination from the generating AI model.

2. A process to display a list of patient information obtainable from the electronic medical record, The computer is further instructed to perform the process of receiving the selection of patient information from the patient information list from the physician, The program according to claim 1, wherein the patient information input to the generating AI model is the patient information selected by the physician.

3. The patient information includes first patient information relating to the patient's unique characteristics and second patient information relating to the patient's health status which changes over time. The program according to claim 1 or claim 2, wherein the second patient information input to the generated AI model is restricted compared with the first patient information.

4. A process to display a list of medical dictionaries, The computer is further instructed to perform the process of receiving a selection of a dictionary from the doctor, which is included in the list of dictionaries. The program according to claim 2, wherein the dictionary selected by the physician is further input to the generating AI model.

5. The computer is further instructed to perform a process to receive supplementary information from the aforementioned physician. The program according to claim 4, wherein the supplementary information is further input to the generated AI model.

6. The program according to claim 4 or 5, wherein, in addition to the dictionary selected by the physician, a dictionary of general medical terms is further input to the generating AI model, regardless of the physician's selection.

7. A process for storing the audio information acquired during the aforementioned medical examination, The program according to claim 1 or claim 2, further causing the computer to perform the process of inputting the stored voice information and patient information back into the generating AI model, and then obtaining the summary information again from the generating AI model.

8. A process for receiving corrections to the summary information from the physician, The program according to claim 4 or 5, further causing the computer to perform the process of sending an instruction to the generating AI model to reflect the content of the aforementioned modifications in the dictionary.

9. A control method for controlling a system that manages electronic medical records, Computers To obtain audio information that shows the voice of a doctor during a patient examination, To obtain patient information concerning the aforementioned patient from the electronic medical record, A control method comprising inputting the aforementioned voice information and patient information into a generating AI (Artificial Intelligence) model, and obtaining summary information from the generating AI model that represents a summary of the examination.

10. An electronic medical record system for managing electronic medical records, A first acquisition unit that acquires audio information representing the voice of a doctor during a patient examination, A second acquisition unit that acquires patient information relating to the aforementioned patient from the electronic medical record, An electronic medical record system comprising: a third acquisition unit that inputs the aforementioned voice information and the aforementioned patient information into a generating AI (Artificial Intelligence) model and acquires summary information indicating a summary of the examination from the generating AI model.

Images