Information processing device, information provision method, and program

Abstract

To provide a mechanism that supports the review of a record input by a user who changes his / her action.SOLUTION: An information processing device includes one or a plurality of processors in which the one or the plurality of processors acquire data related to health input by a user who changes his / her action, instruct a learned model to generate a summary of the data related to health, and present the generated summary on a browsing screen.SELECTED DRAWING: Figure 7

Description

[Technical Field]

[0001] The present disclosure relates to an information processing device, an information providing method, and a program. [Background technology]

[0002] Today, there are application programs (hereinafter referred to as "programs") that can be used for health promotion and medical treatment. Programs that do not require approval under the Pharmaceutical and Medical Device Act are called, for example, healthcare apps. On the other hand, programs that require approval under the Pharmaceutical and Medical Device Act are called treatment apps. These programs have a function for recording daily activities, feelings, etc. in a diary-like manner. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 5299965 Summary of the Invention [Problem to be solved by the invention]

[0004] There are times when you want to review past records. However, reviewing these types of records, which allow free description, can be a heavy burden the longer the period covered.

[0005] As one aspect of the present disclosure, a mechanism is provided to support review of records entered by users who change their behavior. [Means for solving the problem]

[0006] The invention described in claim 1 is an information processing device having one or more processors, which acquire health-related data input by a user who is changing their behavior, instruct a trained model to generate a summary of the health-related data, and present the generated summary on a viewing screen. The invention described in claim 2 is an information processing device described in claim 1, wherein the one or more processors instruct the trained model to generate the summary by quoting one or more partial data extracted from the health-related data. The invention described in claim 3 is the information processing device described in claim 2, wherein the one or more processors instruct the trained model to distinguish between one of the partial data and other of the partial data. The invention described in claim 4 is the information processing device described in claim 2, wherein the one or more processors instruct the trained model to generate the summary including the number of occurrences of the partial data. The invention described in claim 5 is the information processing device described in claim 2, wherein the one or more processors instruct the trained model to generate the summary that does not include an estimated expression. The invention described in claim 6 is the information processing device described in claim 2, wherein the one or more processors instruct the trained model to generate the summary that does not include instructional expressions. The invention described in claim 7 is the information processing device described in claim 1, wherein the one or more processors instruct the trained model to generate the summary on a predetermined item basis. The invention described in claim 8 is an information processing device described in claim 7, wherein the one or more processors instruct the trained model to use management items determined according to the program used to record the health-related data as the specified items. The invention described in claim 9 is the information processing device described in claim 1, wherein the one or more processors instruct the trained model on the period for generating the summary. The invention described in claim 10 is the information processing device described in claim 1, wherein the one or more processors present the summary in association with a goal for a period corresponding to the health-related data. The invention described in claim 11 is the information processing device described in claim 1, wherein the one or more processors instruct a second trained model to generate lifestyle challenges based on the health data and present the challenges on a second viewing screen. The invention described in claim 12 is the information processing device described in claim 11, wherein the one or more processors present an input field for a new goal in association with the task. The invention described in claim 13 is the information processing device described in claim 1, wherein the one or more processors instruct a second trained model to generate information indicating improved lifestyle habits based on the health data, and present the information indicating the improved lifestyle habits on a second viewing screen. The invention described in claim 14 is an information processing device described in claim 1, wherein the one or more processors instruct the trained model to generate the summary including a description indicating the number of occurrences of content that appears multiple times. The invention recited in claim 15 is the information processing device recited in any one of claims 1 to 14, wherein the summary includes a content portion and a quotation portion. The invention described in claim 16 is an information processing device described in claim 15, wherein the one or more processors instruct the trained model to prohibit quoting a summary of one or more partial data extracted from the health-related data as a generation condition for the quoted portion. The invention described in claim 17 is an information provision method in which a computer executes the following processes: acquiring health-related data input by a user who is changing their behavior; instructing a trained model to generate a summary of the health-related data; and presenting the generated summary on a viewing screen. The invention described in claim 18 is a program for causing a computer to realize the following functions: acquiring health-related data input by a user who is changing their behavior; instructing a trained model to generate a summary of the health-related data; and presenting the generated summary on a viewing screen. [Effects of the Invention]

[0007] According to one embodiment of the present disclosure, it is possible to support the review of records entered by a user who is changing their behavior. [Brief explanation of the drawings]

[0008] [Figure 1] 1 is a diagram illustrating an example of the overall configuration of an information processing system according to a first embodiment. [Figure 2] FIG. 1 is a diagram illustrating an example of the hardware configuration of a PDT platform. [Figure 3] FIG. 2 is a diagram illustrating an example of status management data stored in an auxiliary storage device of the PDT platform. [Figure 4] FIG. 2 is a diagram illustrating an example of the hardware configuration of a PDT server. [Figure 5] FIG. 10 is a diagram illustrating an example of patient data stored in an auxiliary storage device of the PDT server. [Figure 6] FIG. 2 is a diagram illustrating an example of the hardware configuration of a doctor terminal and a patient terminal. [Figure 7] FIG. 2 is a diagram illustrating an example of a processing sequence according to the first embodiment. [Figure 8] FIG. 10 is a diagram illustrating an example of an operation screen in a patient application. [Figure 9] FIG. 10 is a diagram illustrating screen transitions related to a review input screen. [Figure 10] FIG. 10 is a diagram illustrating an example of a management screen for the usage status of a patient application displayed on the output device of the doctor terminal. [Figure 11] FIG. 10 is a diagram illustrating an example of a patient data viewing screen. [Figure 12] FIG. 2 is a diagram illustrating an example of a prompt description in the first embodiment. [Figure 13] FIG. 10 is a diagram illustrating an example of a JSON file output by a generation AI server. [Figure 14] FIG. 10 is a diagram illustrating a display example of a summary screen in the first embodiment. [Figure 15] FIG. 10 is a diagram illustrating an example of a processing sequence according to the second embodiment. [Figure 16] FIG. 10 is a diagram illustrating a method for reading out an inputted review from the menu screen. [Figure 17] 10A and 10B are diagrams illustrating another method for reading out an inputted review from the menu screen. [Figure 18]FIG. 10 is a diagram illustrating an example of a prompt description in the second embodiment. [Figure 19] FIG. 10 is a diagram illustrating an example of a JSON file output by a generation AI server. [Figure 20] FIG. 10 is a diagram illustrating a display example of a summary screen in the second embodiment. [Figure 21] FIG. 10 is a diagram illustrating a summary screen that does not include a display of the number of occurrences. [Figure 22] FIG. 10 is a diagram illustrating a summary screen including a display of the number of occurrences in the "Total summary" section. [Figure 23] FIG. 10 is a diagram illustrating another example of the patient data viewing screen. [Figure 24] FIG. 10 is a diagram illustrating another display example of the summary screen. [Figure 25] FIG. 10 is a diagram illustrating another example of a home screen display. [Figure 26] FIG. 10 is a diagram illustrating an embodiment in which a patient terminal directly instructs a generation AI server to generate a summary. [Figure 27] FIG. 10 is a diagram illustrating an embodiment in which "review" recorded through a non-patient app and "review" recorded through a patient app are used in combination. [Figure 28] FIG. 10 is a diagram illustrating another embodiment in which "review" recorded through a non-patient app and "review" recorded through a patient app are used in combination. DETAILED DESCRIPTION OF THE INVENTION

[0009] <Terminology> First, terms used in the embodiments described below will be explained. A "program that encourages behavioral change" is a program that is provided with the intention of encouraging a change in a person's behavior. However, this program does not guarantee that a person's behavior will change. Whether or not a person's behavior actually changes depends on the user who uses the program. Therefore, a program that encourages behavioral change can also be said to be a program that supports behavioral change. This type of program can be classified as a medical device or a non-medical device. A medical device program is an example of a program that requires a prescription, while a non-medical device program is an example of a program that does not require a prescription.

[0010] "Program objectives" refer to the effects to be achieved by using a program to promote behavioral change. Program objectives vary depending on the disease for which the program promotes behavioral change is targeted. For example, program objectives may include improving lifestyle habits, maintaining improved lifestyle habits, and maintaining improved numerical values. "Program goal" refers to the goal to be achieved through the use of a program to promote behavioral change. Goals are defined from the perspective of achieving the objective. Goals are classified into qualitative indicators and quantitative indicators. Goals are also classified into indicators defined by measurements or other numerical values ​​and indicators defined by the content of actions. For example, goals can include improving or maintaining habits, behaviors, or numerical values. Program goals can also be further defined by one or more subgoals. Subgoals are smaller-scale targets set to achieve the corresponding goal.

[0011] "Therapeutic App" refers to an application program that has been approved under the Pharmaceuticals and Medical Devices Act. Therapeutic Apps are programs that are classified as medical devices. Therapeutic apps are approved for each disease. Diseases for which approval has already been obtained include hypertension, nicotine addiction, and insomnia. Diseases for which therapeutic apps are currently in development include NASH (non-alcoholic steatohepatitis), diabetes, dyslipidemia, kidney disease, and alcoholism.

[0012] "User" refers to a person who changes their behavior through a program that encourages behavioral change. Users include users who have not yet visited a medical institution and users who have visited a medical institution. Users who have visited a medical institution include users who are subject to medical fee calculations and users who are not subject to medical fee calculations. A "patient" refers to a user who is currently receiving treatment at a medical institution. In other words, a patient refers to a user who has started or is currently undergoing treatment at a medical institution.

[0013] Treatment apps include patient apps and doctor apps. A "Patient App" is an application program that runs on a device operated by a patient (hereinafter also referred to as a "Patient Terminal") and is prescribed to the patient by a doctor. In this sense, the Patient App is also called a PDT (Prescription Digital Therapeutic). The patient app can be downloaded from, for example, an app store. In the embodiment described below, a code required for activation (hereinafter referred to as a "prescription code") is issued by a doctor's prescription. The patient app is used to record patient data (hereinafter also referred to as "patient app data") outside of medical institutions.

[0014] A validity period is set for a patient app at the time of approval. This validity period is determined based on, for example, the period during which the public medical insurance system applies. The validity period is determined depending on the type of disease that the patient app covers. For example, the validity period for a patient app for hypertension is six months, counting from the month following the month in which the patient app was prescribed. However, six months is just an example, and it could be, for example, nine months or 12 months. The validity period may also be determined in days, such as 60 days or 180 days, or in weeks, such as eight weeks or 24 weeks. Needless to say, these values ​​are just examples.

[0015] A "doctor app" refers to an application program that can be used through a terminal operated by a doctor or other medical professional (hereinafter also referred to as a "doctor terminal"). In the embodiment described below, the doctor app runs on a cloud server that can be operated from the doctor terminal. The doctor app is used to view patient data (including patient app data). Note that medical professionals are also referred to as medical workers.

[0016] "Patient data" includes, for example, patient attributes, measurement values, activity records, mood records, physical condition records, medical examination history, patient app operation history, biological characteristics, psychological characteristics, social characteristics, habits, and goal achievement status. However, the items recorded as patient data vary depending on the disease, and do not necessarily have to be all of the exemplified information, but may be some of it or may include other information. Note that the items recorded as patient data are all examples of health-related data.

[0017] The "patient attributes" include, for example, the patient's name, gender, and date of birth. These pieces of information are examples of basic patient information. "Measurement value" refers to a numerical value measured using a measuring device. The measurement value items recorded as patient app data are defined for each disease. For example, if the disease is hypertension, blood pressure values ​​are recorded as measurement values. Blood pressure values ​​are defined, for example, as systolic blood pressure (i.e., maximum blood pressure) and diastolic blood pressure (i.e., minimum blood pressure). For example, if the disease is nicotine addiction, the measurements recorded would be carbon monoxide (CO) in the breath and nicotine concentration in the saliva.

[0018] The "activity record" is, for example, the operation history of the patient app, the medication record, the meal record, the smoking record, the drinking record, and the exercise record. These records are also records of the user's behavior. The activity record is an example of information related to activity. The "mood record" is, for example, a record of the mood perceived by the user. The mood record is an example of information related to mood. The "physical condition record" is, for example, a record of the physical condition or symptoms that the user feels. The physical condition record is an example of information related to the physical condition.

[0019] The "medical examination history" includes, for example, the treatment start date, the consultation date, the content of the treatment, and advice. The medical examination history is an example of information related to medical examinations. The "patient application operation history" is, for example, a history of operations related to the patient application startup operation, and the input operation of measurement values ​​and reviews. "Biological characteristics" include, for example, whether or not the patient has other illnesses, whether or not they have had injuries undergoing treatment, whether or not they have pain in their knees or feet, whether or not they have had treatment for an illness, the number of years since the illness was diagnosed, how strongly seasoned their food is at home, and the amount of food they eat.

[0020] "Psychological characteristics" include, for example, expectations regarding app treatment, willingness to acquire knowledge about disease treatment, whether reducing salt intake is difficult, whether one believes one's sense of taste cannot be changed, and psychological resistance to leaving food on one's plate. "Social characteristics" include, for example, wake-up time, bedtime, type of work (e.g., shift work, day shift, night shift), days of the week worked, start time of work, time home from work, regular days off, and whether or not there is a heater in the changing room.

[0021] "Habits" include, for example, exercise habits, weighing oneself, checking the calorie count on food labels, choosing low-fat foods, not consuming caffeine after 4 p.m., eating and drinking after 10 p.m., skipping breakfast, snacking, taking a bath one hour before bedtime, stretching or massaging before bedtime, and sleeping for six hours or more. The "goal achievement status" is, for example, information indicating the progress status toward a goal set by a doctor for each patient. The "goal achievement status" may be, for example, information indicating the progress status toward a goal set by the patient himself / herself. The "goal achievement status" may be, for example, information indicating the progress status toward a goal presented by a patient app for each patient. The above-mentioned information can be classified into subjective information and objective information. Patient data is recorded, for example, as text, images (moving images, still images), audio, numbers, and codes.

[0022] The "PDT server" is a server that manages patient data entered through a patient app, etc. The PDT server is also an example of a cloud server. A PDT server is basically provided for each patient app. Therefore, in order for a doctor to view patient data, he / she must log in to the PDT server running the doctor app that is paired with the patient app used by the patient.

[0023] For example, if a patient uses a patient app for hypertension provided by service provider A, the doctor needs to log in to the PDT server operated by service provider A for hypertension. Also, if a patient uses a patient app for hypertension provided by service provider B, the doctor needs to log in to the PDT server operated by service provider B for hypertension.

[0024] In addition, if a patient uses a patient app for nicotine addiction provided by service provider A, the doctor needs to log in to the PDT server operated by service provider A for nicotine addiction. It should be noted that multiple patient apps targeting different diseases may share one PDT server. Furthermore, multiple patient apps provided by different service providers may share one PDT server.

[0025] The PDT platform is a server that manages patient app prescriptions and the usage status of patient apps after prescription. The PDT platform is also an example of a cloud server. The PDT platform is also called an APS (Application Prescription Service) server, meaning that it is a server that provides prescription services for patient apps. The usage status may be, for example, "before use begins," "expired," "in use," "scheduled to end," "ended," or "expired."

[0026] "Before use begins" refers to the state in which the prescription for the patient app has been completed but use on the patient's device has not yet begun. For example, this refers to the state in which the patient has not yet entered the prescription code into the device. "Start expired" refers to a state in which the prescription code has not been confirmed as entered within the period in which the prescription code is valid (for example, within four days including the prescription date).

[0027] "In use" refers to the state in which the patient app installed on the patient device has been activated and is ready for use. Activating the patient app requires the entry of an activation code, such as the prescription code mentioned above. "Scheduled to end" refers to a state in which a medical institution has set the patient to be outside the scope of management within the validity period. For example, this state is set for a patient who will not be seen again.

[0028] "Terminated" refers to a state in which the patient app cannot be used due to, for example, the expiration of the validity period. "Expired" is a state that indicates that a predetermined period has passed since the prescription date. The predetermined period is set to be longer than the validity period. For example, if the validity period is 6 months, the predetermined period is set to 8 months. "Selected medical treatment" is displayed when the validity period has expired but the patient is still eligible for selected medical treatment. Selected medical treatment refers to medical services that patients enrolled in social insurance can receive in addition to insured treatment, treatment not covered by insurance, by paying an additional fee.

[0029] The PDT platform can support multiple patient apps that target the same disease but are provided by different service providers, as well as multiple patient apps that are provided by the same service provider but target different diseases. In this sense, the PDT platform acts as a platform for multiple patient apps. The app, which runs on the PDT platform, manages the usage status of multiple patient apps for different diseases and service providers on a patient-by-patient basis.

[0030] In the embodiments described below, the "medical institution" is assumed to be a medical institution that is covered by health insurance. More specifically, the medical institution refers to a medical institution that is covered by health insurance and to which a doctor who issues a prescription code required to activate a patient app belongs. However, if deregulation allows pharmacists, public health nurses, nurses, nutritionists, hospital staff, and other medical professionals to issue prescription codes, the term "medical institution" also includes facilities and organizations that employ these medical professionals. In the following, these medical professionals will be collectively referred to as "doctors, etc." The issuance of prescription codes is not limited to medical treatment, but may also be for non-insurance treatment (i.e., private treatment) or mixed treatment. Incidentally, medical treatment is not limited to face-to-face treatment, but also includes online treatment.

[0031] <First Embodiment> <Overall system> 1 is a diagram illustrating an example of the overall configuration of an information processing system 1 according to Embodiment 1. The information processing system 1 illustrated in FIG. 1 includes a PDT platform 10, PDT servers 20 (20A, 20B, 20C, . . . 20F), a doctor terminal 30, and a patient terminal 40.

[0032] 1 depicts only one PDT platform 10, although multiple PDT platforms 10 may exist. The PDT platform 10 may be configured with multiple servers connected via a network. In this case, the multiple servers cooperate to provide the services of the PDT platform. The PDT platform 10 and the PDT server 20 are communicatively connected via a network (not shown). The network may be, for example, a LAN (Local Area Network), the Internet, or a mobile communication system (4G, 5G, etc.).

[0033] 1, six PDT servers 20 are connected to one PDT platform 10. However, the number of PDT servers 20 connected to one PDT platform 10 is arbitrary. The PDT server 20 is a server that performs patient authentication, manages patient application data input through the patient application, and provides patient data to the doctor terminal 30. The PDT server 20 is an example of an information processing device. In the case of FIG. 1, the PDT server 20A is a server that manages patient data of a patient application (hereinafter referred to as "patient application A") provided by company A for patients with hypertension. The PDT server 20B is a server that manages patient data of a patient application (hereinafter referred to as "patient application B") provided by Company B for patients with hypertension.

[0034] The PDT server 20C is a server that manages patient data of a patient application (hereinafter referred to as "patient application C") provided by Company A for diabetes patients. The PDT server 20D is a server that manages patient data of a patient application (hereinafter referred to as "patient application D") provided by Company B for diabetes patients. The PDT server 20E is a server that manages patient data of a patient application (hereinafter referred to as "patient application E") provided by Company C for patients with dyslipidemia. The PDT server 20F is a server that manages patient data of a patient application (hereinafter referred to as "patient application F") provided by Company D for patients with NASH.

[0035] 1, a PDT server 20 is prepared for each combination of a disease targeted by a patient app and a service provider that provides the patient app. Therefore, even if the service provider that provides the patient app is the same, if the targeted diseases are different, a different PDT server 20 is prepared. Furthermore, even if the target disease is the same, if the service provider that provides the patient app is different, a different PDT server 20 will be prepared.

[0036] However, it is also possible to provide one PDT server 20 for a plurality of patient applications with different combinations. The PDT platform 10 and the PDT server 20 may be operated by different operators, not necessarily by the same operator. For example, some of the PDT servers 20 may be operated by the same operator as the PDT platform 10.

[0037] The doctor terminal 30 is a terminal operated by a doctor or the like who uses the services provided by the PDT platform 10 and the PDT server 20. In Figure 1, only one representative doctor terminal 30 is depicted. For example, a doctor can view the usage status of the patient app by logging in to the PDT platform 10. Also, a doctor can view patient data by logging in to the PDT server 20. The patient terminal 40 is a terminal operated by a patient. The patient terminal 40 uploads patient application data recorded through the patient application to the corresponding PDT server 20. In Fig. 1, only one patient terminal 40 is depicted as a representative example.

[0038] <Device hardware configuration> <PDTプラットフォーム> 2 is a diagram illustrating an example of the hardware configuration of the PDT platform 10. The PDT platform 10 is a so-called server. 2 includes a processor 11, a semiconductor memory 12, an auxiliary storage device 13, and a communication interface 14. Each device is connected via a bus or other signal lines.

[0039] The processor 11 is a device that realizes various functions through the execution of programs. The semiconductor memory 12 stores UEFI (Unified Extensible Firmware Interface) and the like. The semiconductor memory 12 is also used as an execution area for programs. The processor 11 and the semiconductor memory 12 function as a computer.

[0040] The auxiliary storage device 13 is composed of, for example, a hard disk device or a semiconductor storage. The auxiliary storage device 13 stores an operating system and other programs. The other programs include, for example, an application program that displays a list of usage statuses of the patient app. In addition, the auxiliary storage device 13 also stores data for managing the status management data 130 by prescription code prescribed by a medical institution. The communication interface 14 is an interface for communicating with an external terminal such as a PDT server 20 (see FIG. 1) through a network. The communication interface 14 is compatible with communication standards such as Ethernet (registered trademark), Wi-Fi (registered trademark), and mobile communication systems.

[0041] <PDT Platform Management Data> FIG. 3 is a diagram for explaining an example of the status management data 130 stored in the auxiliary storage device 13 (see FIG. 2) of the PDT platform 10 (see FIG. 1). The status management data 130 shown in FIG. 3 stores a prescription code 130A, a patient ID / patient name 130B, a prescription date / medical treatment date 130C, a medical institution ID / prescriber ID 130D, a patient app name 130E, and a usage status 130F. However, these are just examples, and for example, the patient's medical record number, the patient's gender, the patient's date of birth, age, type of insurance card, insurance number, and version of the patient app may be stored.

[0042] The prescription code 130A is issued every time a prescription is notified from the doctor terminal 30. The patient ID / patient name 130B is the patient ID and patient name registered at the time of prescribing the patient app. In the case of FIG. 3, the patient name of patient ID "12543" is "Mr. A", the patient name of patient ID "12544" is "Mr. B", and the patient name of patient ID "12545" is "Mr. C".

[0043] The prescription date / treatment date 130C is the date on which the doctor examined the patient. In this embodiment, the treatment date on which the doctor prescribed the patient app is referred to as the "prescription date" to distinguish it from other treatment dates. In the case of FIG. 3, only the prescription date is stored. In the case of FIG. 3, the prescription date of the patient app for "Mr. A" and "Mr. B" is "2024 / 5 / 28", and the prescription date of the patient app for "Mr. C" is "2024 / 2 / 24". The medical institution ID / prescriber ID 130D is an ID that identifies the medical institution and prescriber that prescribed the patient app. In the case of Figure 3, the prescription codes "12345" and "23456" were prescribed by the same doctor at the same medical institution.

[0044] The patient app name 130E is the name of the patient app prescribed by a doctor or the like. However, as long as it is possible to identify the patient app, an identification code of the patient app may be stored. In the case of FIG. 3, patient app A has been prescribed to "Mr. A" and "Mr. B." Patient app D has been prescribed to "Mr. C." The usage status 130F is information indicating the usage status of the patient app. The usage status stores one of the following: before use, start deadline expired, in use, scheduled to end, finished, and expired. In the case of FIG. 3, all patient apps are "in use."

[0045] <PDTサーバ> FIG. 4 is a diagram illustrating an example of the hardware configuration of the PDT server 20. As shown in FIG. 4 includes a processor 21, a semiconductor memory 22, an auxiliary storage device 23, and a communication interface 24. Each device is connected via a bus or other signal lines. The processor 21 is a device that realizes various functions through the execution of programs. The semiconductor memory 22 stores UEFI and the like. The semiconductor memory 22 is also used as an execution area for programs. The processor 21 and the semiconductor memory 22 function as a computer.

[0046] The auxiliary storage device 23 is composed of, for example, a hard disk device or a semiconductor storage. The auxiliary storage device 23 stores an operating system and other programs. Among the other programs, there is, for example, a doctor app. The doctor app is an application program that generates a browsing screen for patient data corresponding to patients treated by doctors and the like.

[0047] The other programs also include an application program and a prompt that instruct the generation AI server 50 (see FIG. 7) to generate a summary of the "review". In addition, the auxiliary storage device 23 also records patient data 230 recorded through the patient app The communication interface 24 is an interface for communicating with an external terminal such as the PDT platform 10 through a network. The communication interface 24 is compatible with communication standards such as Ethernet (registered trademark), Wi-Fi (registered trademark), and mobile communication systems.

[0048] <Management data of the PDT server> FIG. 5 is a diagram for explaining an example of patient data 230 stored in the auxiliary storage device 23 (see FIG. 4) of the PDT server 20. Note that the patient data 230 shown in FIG. 5 assumes data of a hypertension patient. The patient data 230 shown in FIG. 5 stores measured values and other information recorded through a patient app for hypertension.

[0049] In the case of FIG, 5, the patient data 230 stores a prescription code 230A, a patient ID / patient name 230B, and patient app data 230C. In the case of the present embodiment, the patient app data 230C means the management data of the patient app and the data recorded by the patient through the patient app. [[ID=二十]]The prescription code 230A records the prescription code 130A (see FIG. 3) issued by the PDT platform 10 (see FIG. 1). The prescription code 130A is registered by the patient at the start of using the patient app (that is, at the time of new registration). The patient ID / patient name 230B is, for example, the patient ID and the patient name of the medical institution that prescribed the patient app. The patient ID and the patient name are acquired from the PDT platform 10, for example, when authenticating the prescription code.

[0050] 5 includes a measurement value / measurement date and time 230C1, a review / input date and time 230C2, a current step of a treatment program 230C3, a treatment program practice history 230C4, a behavioral goal 230C5, a target value 230C6, an outpatient record 230C7, and a medication record 230C8. Note that the illustrated items do not need to be all of the patient application data 230C, and may be only a part of the patient application data 230C, or may include other items. The measured blood pressure value and the date and time when the blood pressure was measured are recorded in the measurement value / measurement date and time 230C1. The blood pressure value is given as a systolic blood pressure value and a diastolic blood pressure value.

[0051] The review / input date and time 230C2 records a review of the day's actions and the input date and time. In the review, for example, the level of physical condition, stress level, length of sleep, weight, and details of the actions taken are recorded. The details of the actions are composed of the category of the actions taken and text input. Other label candidates include, for example, "sleep," "stress," "moderate drinking," and "other." In this embodiment, the category of the actions taken is recorded by checking buttons labeled "reduce salt," "weight loss," "exercise," etc. In the text input, it is possible to freely input details and emotions that cannot be recorded by checking buttons.

[0052] The current step 230C3 of the treatment program records a step indicating the progress of the treatment program provided by the patient app. In this embodiment, the treatment program is composed of three steps. The three steps are, for example, "acquiring knowledge," "practicing behavioral goals," and "making behavior a habit." The treatment program practice history 230C4 records the history of learning and behavior practiced in accordance with the treatment program. For example, in the case of "Step 1," the learning history records whether or not each learning item has been taken. For example, in the case of "Step 2," the behavior history records the practice status of each goal, such as "reducing salt intake," "reducing weight," "exercise," "sleep," "stress," "alcohol consumption," and "quitting smoking." For example, in the case of "Step 3," the practice status of the behavior set by the patient as a goal is recorded.

[0053] The behavioral goal 230C5 records behaviors that the patient has set as goals. The goals here include one or more behavioral goals that the patient has selected from among the goals presented by the patient app according to the progress of the treatment program. For example, the behavioral goal 230C5 records a behavioral goal that the patient has set in step 2 or 3 of the treatment program. A behavioral goal refers to a behavioral goal that should be achieved in order to achieve a target. The behavioral goal is not limited to a behavioral goal selected from the behaviors presented by the patient app, and may include behaviors individually set by the patient.

[0054] In target value 230C6, a value set by a doctor or the like in the "blood pressure reduction target value field" 313 (see FIG. 11) on the patient data viewing screen 310 (see FIG. 11) is recorded. The "blood pressure reduction target value field" 313 is an example of a quantitative target in that it is given as a numerical value. The outpatient record 230C7 records appointments with the doctor, the next appointment date, medication timing, etc. The appointments with the doctor record appointments that the patient should make or be aware of before the next medical examination. In this embodiment, appointments with the doctor can only be recorded through the patient app. Note that appointments with the doctor may also be recorded only from the doctor terminal 30 (see FIG. 1). The medication record 230C8 records information about taking prescribed medication.

[0055] <Doctor's terminal / Patient's terminal> Figure 6 is a diagram illustrating an example of the hardware configuration of the doctor terminal 30 and the patient terminal 40. The hardware configuration of the doctor terminal 30 and the hardware configuration of the patient terminal 40 are basically the same. For this reason, in Figure 6, they are expressed in the format of "symbols of elements constituting the doctor terminal 30 / symbols of elements constituting the patient terminal 40". The doctor terminal 30 / patient terminal 40 shown in Figure 6 includes a processor 31 / 41, a semiconductor memory 32 / 42, an auxiliary storage device 33 / 43, an input interface 34 / 44, an input device 35 / 45, an output interface 36 / 46, an output device 37 / 47, and a communication interface 38 / 48. Each device is connected via a bus or other signal lines.

[0056] The processor 31 / 41 is a device that realizes various functions through the execution of a program. The semiconductor memory 32 / 42 stores UEFI and the like. The semiconductor memory 32 / 42 is also used as an execution area for programs. The processor 31 / 41 and the semiconductor memory 32 / 42 function as a computer. The auxiliary storage device 33 / 43 is configured by, for example, a hard disk drive or semiconductor storage, and stores an operating system and other programs.

[0057] In the case of the doctor terminal 30, the auxiliary storage device 33 stores a client certificate for accessing the PDT platform 10 (see FIG. 1) and a client certificate for accessing the PDT server 20 (see FIG. 1). In the case of the patient terminal 40, a client certificate for accessing the PDT server 20 (see FIG. 1) is recorded in the auxiliary storage device 43. In addition, the auxiliary storage device 43 also records patient application data 230C (see FIG. 5). The input interface 34 / 44 uses, for example, USB (=Universal Serial Bus) or Bluetooth (registered trademark) for connection to the input device 35 / 45. The input device 35 / 45 may be, for example, a keyboard, a mouse, or a touch panel.

[0058] The output interfaces 36 / 46 may be, for example, HDMI (=High-Definition Multimedia Interface) (registered trademark) or a LAN interface for connection to the output devices 37 / 47. The output devices 37 / 47 may be, for example, a monitor or a printer. The communication interface 38 / 48 is an interface for communicating with an external terminal via a network, and is compatible with Ethernet (registered trademark), Wi-Fi (registered trademark), mobile communication systems, and other communication standards.

[0059] <Patient data viewing screen output> The output process of the patient data viewing screen in the information processing system 1 (see FIG. 1) will be described below with reference to FIGS. Fig. 7 is a diagram illustrating an example of a processing sequence according to Embodiment 1. The symbol S in Fig. 7 represents a step.

[0060] <Step 101> Patients record measurements and other data from the patient app's operation screen. Patients are an example of users who change their behavior. Patients record, for example, measurements, reflections, behavioral goals, target values, outpatient records, and medication records. These records are examples of health-related data. 8 is a diagram illustrating an example of an operation screen in the patient application. The operation screen shown in FIG. 8 can be transitioned from the home screen 401. The measurement value input screen 402 is, for example, a screen for inputting blood pressure values. Measurement values ​​can be input manually or through data linkage with a blood pressure monitor. However, data linkage with a blood pressure monitor is limited to cases where the blood pressure monitor can be linked with a patient app.

[0061] The review input screen 403 allows users to record their physical condition level, stress level, sleep duration, weight, activity details, mood, and physical condition. As mentioned above, the activity details include records of button operations corresponding to the category of activity that was successfully performed. The activity details also include text input. 9 is a diagram illustrating screen transitions related to the review input screen 403. The review input screen 403 is displayed by operating the input button 401A for "Today's review" provided on the home screen 401. Since it is "Today's review," it is assumed that input is made for each day. Note that the input button 401A shown in FIG. 9 is labeled "Enter."

[0062] 9 is composed of two screens. The first screen has fields for inputting the level of physical condition, the level of stress, the amount of sleep, and the weight. The second screen, which is displayed when the input on the first screen is completed, has a recording field 403A for actions related to lifestyle habits and a recording field 403B in which the user can freely record. Recording field 403A shown in Fig. 9 contains the question "Please tell us about the habits you were aware of or were able to act on today (multiple answers allowed)" and four buttons labeled "Lose weight," "Reduce salt intake," "Exercise," and "Other." In Fig. 9, the buttons corresponding to "Reduce salt intake" and "Other" are selected.

[0063] Recording field 403B shown in FIG. 9 includes an explanatory note saying "Please write freely," an example entry, and a text entry field. The text entered in recording field 403B is an example of health-related data entered by the user. In the case of FIG. 9, the text entered in the text entry field is "I tried cooking low-salt food. I'm glad it turned out delicious." Record field 403B, which allows free entry, can also be used as a kind of diary or memo, so the content entered in record field 403B is not necessarily limited to matters related to health. The contents entered in the record field 403B can be viewed on the doctor terminal 30 (see FIG. 1).

[0064] Returning to the explanation of Figure 8. Start program screen n is the screen used for "Step 1" of the treatment program. "Step 1" is a learning stage aimed at acquiring knowledge about the disease. In "Step 1," the patient is first asked to enter information about their employment status, preferences, etc. Next, the patient's characteristics are analyzed from the entered information. Once the analysis results are obtained, knowledge regarding "reducing salt intake," "weight loss," "exercise," "sleep," "stress," "alcohol," "tobacco," "understanding the disease," and "blood pressure measurement" is presented according to the patient's characteristics. In addition, some of the information presented to patients will be adjusted depending on whether or not kidney dysfunction is indicated, their usual exercise habits, alcohol consumption, etc. The recommended duration for "Step 1" is two weeks.

[0065] "Step 2" is the stage where behavioral goals are put into practice. First, in "Step 2," behavioral goals appropriate for the patient are presented from among "reducing salt," "reducing weight," "exercise," "sleep," "stress," "alcohol," and "tobacco." The patient confirms the presented actions and records the results of the actions. Regarding the behavioral goals in this step, the displayed content of each item is partially adjusted based on the information entered by the patient themselves. In "Step 2," specific behavioral goals are presented for multiple behavioral categories that represent lifestyle habits that will improve the disease. Examples of behavioral categories include "reducing salt intake," "weight loss," "exercise," "sleep," "stress," and "alcohol." For example, behavioral goals in the "reducing salt intake" category include "avoiding salt intake through snacks and rice crackers," "avoiding drinking the broth of udon or ramen noodles," "avoiding eating at convenience stores or eating out for lunch," "avoiding eating within two hours before bed," "walking for 20 minutes," and "getting at least six hours of sleep."

[0066] "Step 3" is the stage where behavior becomes a habit. The patient sets the blood pressure target themselves. The blood pressure target is an example of an achievement goal. The patient records the progress of achieving the blood pressure target. In "Step 3," behaviors that the patient himself evaluated as being highly effective in "Step 2" are presented as behavioral goals with priority. The patient sets the presented behavior as a behavioral goal and aims to make the behavior a habit.

[0067] The behavioral goal list screen 405 is a screen that can be displayed after "Step 2." A list of behavioral categories is displayed on the behavioral goal list screen 405. The degree of adherence is also displayed for each category. The social support screen 406 is a screen that can be displayed after "Step 2". A list of social supports is displayed on the social support screen 406. The social support screen 406 displays treatment know-how that is useful in real life.

[0068] The My Data screen 407 is a screen used to check the progress of steps, measurement values, review, salt intake status, knowledge, and behavior. The measurement value screen 408 is a screen on which the measurement values ​​recorded up to now are displayed as a trend graph or numerical values. The salt intake screen 409 is a screen that can be displayed after "Step 2". On the salt intake screen 409, it is possible to record salt intake, record dietary details, and view the recorded details. The recorded details can be displayed as a trend graph or numerical values. The outpatient record screen 410 is a screen on which the patient can record appointments made with the doctor at the time of the visit, record the next visit date, record the time of day when prescribed medicines will be taken, and view the recorded contents. The medication record input screen 411 is a screen on which medication records can be recorded and the recorded contents can be viewed.

[0069] <Step 102> Returning to the explanation of Figure 7. The patient terminal 40 uploads patient app data at predetermined times. One of the predetermined times is when new data is recorded or updated. This upload is performed when the patient app remains logged in to the PDT server 20. Another predetermined time is when the patient signs in to the PDT server 20. This upload is performed when the patient app logs out of the PDT server 20 and then logs in to the PDT server 20 again.

[0070] <Step 103> The PDT server 20 stores the uploaded patient application data, which constitutes part of the patient data 230 (see FIG. 5). Steps 101 to 103 shown in FIG. 7 are executed in response to a recording operation by a patient to the patient application.

[0071] <Step 104> When a doctor or the like logs in to the PDT platform 10 via the doctor terminal 30, the PDT platform 10 presents a confirmation screen for the usage status of the patient app to the doctor terminal 30. FIG. 10 is a diagram illustrating an example of a management screen 300 for the use status of patient applications displayed on the output device 37 (see FIG. 6) of the doctor terminal 30 (see FIG. 6).

[0072] The management screen 300 shown in FIG. 10 is an example of a management screen that displays, in a list, the usage status of patient apps prescribed to patients by a doctor or the like who operates the doctor terminal 30. The management screen 300 may display only the usage status of patient apps prescribed by the doctor or the like operating the doctor terminal 30, or may display the usage status of patient apps prescribed by the medical institution to which the doctor or the like operating the doctor terminal 30 belongs. In other words, the management screen 300 may include the usage status of patient apps prescribed by other doctors or the like who belong to the same medical institution.

[0073] In addition, the management screen 300 may be configured to display the usage status of patient apps prescribed by other doctors or other medical institutions. This display function is useful, for example, when a patient requests a second opinion. This display can be realized, for example, by a doctor operating the doctor terminal 30 providing information identifying the patient under treatment (for example, a medical record number or a patient name) to the PDT platform 10. Furthermore, technically, it is also possible to include information on all patients managed by the PDT platform 10 on the management screen 300. In that case, however, it is desirable to be able to narrow down or rearrange the patients displayed on the management screen using the ID of the doctor operating the doctor terminal 30, etc.

[0074] The management screen 300 shown in FIG. 10 is composed of an information field 301 , a current date and time 302 , a “Details” button 303 , a “New prescription” button 304 , and a “Close” button 305 . Management information of the patient app prescribed to each patient is displayed in the information field 301. In the case of Fig. 10, the information field 301 is displayed in a table format. Incidentally, information for each patient is displayed in the rows, and management items of the patient app are displayed in the columns. FIG. 10 shows, as management items, examples of medical record number / patient name 301A, application name 301B, usage status 301C, prescription date 301D, and prescription code expiration date 301E.

[0075] The medical record number / patient name 301A displays the medical record number and the patient name, which are examples of information for identifying a patient to whom the patient application has prescribed a medication. The information for identifying a patient may include a user ID, an address, etc. The application name 301B displays the name of the patient application prescribed to the patient. In this embodiment, four applications are displayed: patient application A, patient application B, patient application C, and patient application D. The application name 301B may include version information.

[0076] Note that if multiple patient apps are approved for a patient, the information will be displayed in different rows. In addition to or in addition to the application name 301B, the name of the disease targeted by the patient application may be displayed. The usage status 301C is used to display the usage status of the patient app. In the case of Fig. 10, the usage status 301C displays three options: "Before starting usage," "In use," and "Ended." However, as mentioned above, the usage status 301C also displays "Start deadline expired," "Scheduled to end," "Expired," etc.

[0077] The prescription date 301D displays the date on which the patient app was prescribed by a doctor or the like. The last date on which the prescription code is valid is displayed in prescription code expiration date 301E. In this embodiment, the date three days after prescription date 301D is displayed in prescription code expiration date 301E. The current date and time 302 is the date and time when the patient application usage status management screen 300 is viewed.

[0078] The "Details" button 303 is a button used to display a patient data viewing screen. The "Details" button 303 is arranged on each row corresponding to a patient. When the "Details" button 303 is operated, the patient data viewing screen for the patient corresponding to the "Details" button 303 is displayed. The "New Prescription" button 304 is used when prescribing a patient app to a patient. When the "New Prescription" button 304 is operated, a screen for inputting the app name, patient name, gender, date of birth, etc. is displayed. The name of the prescribing doctor is also registered at the same time. The "close" button 305 is a button for closing the management screen 300 shown in FIG.

[0079] <Step 105> Returning to the explanation of Figure 7. The doctor terminal 30 accepts the selection of a patient on the management screen 300 (see FIG. 10) described above. Specifically, the doctor terminal 30 accepts the operation of the "Details" button 303 (see FIG. 10). For example, the doctor terminal 30 accepts the operation of the "Details" button 303 corresponding to Mr. H, for whom more than five months have passed since the prescription date. "Mr. H" here is an example of a specific user. The doctor terminal 30 that has accepted the operation of the "Details" button 303 accesses the PDT server 20 of the patient app prescribed to the patient corresponding to the operated "Details" button 303.

[0080] It is assumed that the doctor or the like has already logged in to the PDT server 20 to be accessed via the doctor terminal 30. In this access, the prescription code of the patient app prescribed to the patient corresponding to the operated “Details” button 303 is notified. The doctor terminal 30 may request the PDT platform 10 to issue a one-time token required to access the corresponding PDT server 20. In this case, the doctor terminal 30 accesses the PDT server 20 using the one-time token returned from the PDT platform 10. When this method is adopted, doctors and other personnel can access patient app data of PDT servers 20 to which they are not logged in.

[0081] <Step 106> The PDT server 20, which is accessed by the doctor terminal 30, generates a patient data viewing screen for the selected patient. FIG. 11 is a diagram illustrating an example of a patient data viewing screen 310. The viewing screen 310 shown in FIG. 11 is composed of a “patient information field” 311, a “measurement result display field” 312, a “blood pressure reduction target value field” 313, a “review information display field” 314, an “activity record display field” 315, a “Go to examination date registration screen” button 316, and an “Exit” button 317.

[0082] The "patient information column" 311 shown in FIG. 11 indicates that the medical record number is "13002," the patient name is "Mr. H," the age is "60 years old," and the sex is "male." The average blood pressure measured at home is displayed numerically and graphically in the "measurement result display column" 312 shown in Figure 11. Daily measurement values ​​can be entered on the measurement value input screen 402 (see Figure 8).

[0083] In the case of FIG. 11, the average values ​​for two intervals, "8 weeks to 4 weeks ago" and "4 weeks ago to the day before," are shown. Here, the day before means the day before the current date and time when the viewing screen 310 is opened. Doctors and other medical professionals can check the progress of treatment by viewing the numerical values ​​and other information displayed in the "measurement result display column" 312. 11 indicates that the target value for home systolic blood pressure (maximum blood pressure) is less than 125 mmHg, and the target value for diastolic blood pressure (minimum blood pressure) is less than 75 mmHg. Note that blood pressure targets may differ depending on gender, the presence of other diseases, age, and other factors. 11, a "change" button for the target value is provided in the "blood pressure reduction target value field" 313. Only a doctor or the like can change the target time.

[0084] The "retrospective information display field" 314 shown in FIG. 11 displays records of mood and physical condition. The retrospective information is the content entered in the record field 403B (see FIG. 9). In the example shown in FIG. 11, a four-day retrospective is displayed. For example, the display shows that the physical condition on May 27 was good, the physical condition on May 26 was normal, the physical condition on May 25 was extremely poor, and the physical condition on May 24 was poor. Additionally, the beginning of the text entered in the record field 403B corresponding to each date is displayed. To read the entire text in the "retrospective information display field" 314, simply operate the "list" button 314A. The "retrospective information display field" 314 also includes a "summary" button 314B. Operating the "summary" button 314B displays a summary of the retrospective text for, for example, the last 30 days on the screen. The processing operations initiated by operating the "summary" button 314B will be described later.

[0085] The "activity record display column" 315 shown in Figure 11 displays activity records such as the number of days the patient app was used, the number of days a salt reduction in meals was recorded, and the number of days a weight loss was recorded. The "Go to consultation date registration screen" button 316 shown in FIG. 11 is a button for opening a consultation date registration screen (not shown). The "Exit" button 317 shown in FIG. 11 is a button for closing the patient data viewing screen 310.

[0086] <Step 107> Returning to the explanation of Figure 7. The doctor terminal 30 accepts an operation to display a summary of the review through the patient data viewing screen 310 (see FIG. 11). Specifically, it accepts an operation of the "Summary" button 314B (see FIG. 11). Note that the processing operations from step 107 onwards are executed only when the "Summary" button 314B is operated. The doctor terminal 30 notifies the PDT server 20 of the acceptance of the operation of the "Summary" button 314B.

[0087] <Step 108> The PDT server 20 instructs the generation AI (=Artificial Intelligence) server 50 to generate a summary of the review, points for improvement, and good points. The generation AI server 50 here is a server that outputs content using a trained model that has learned the input / output relationships that take documents, images, sounds, and other content as input and outputs summaries of those contents through machine learning.

[0088] However, the "review" assumed in this embodiment is based on text input, and therefore the generation AI server 50 of this embodiment uses a trained model specialized for natural language processing. 7, the generation AI server 50 is a server different from the PDT server 20. However, the PDT server 20 may also serve as the generation AI server 50.

[0089] Specifically, the PDT server 20 provides a prompt including target data to the generation AI server 50 . In this embodiment, the target data is text data entered in record field 403B (see FIG. 9). The target data in this embodiment is the content of the "Looking Back" entered up to one month prior to the date of operation of "Summary" button 314B (see FIG. 11). In this embodiment, one month is 30 days. However, one month may also be 28 days.

[0090] In this embodiment, the prompt consists of a "command," a "limit," an "output example," and a "monthly review." The "monthly review" refers to the target data mentioned above. FIG. 12 is a diagram illustrating an example of a prompt description in the first embodiment. The "command" shown in Figure 12 consists of the following four statements:

[0091] One of the commands is, "Summarize the reflections starting from "#1 month's reflection" below and display them in "Total summary." Here, "Total summary" is an example of the content part. This command instructs the generation of a summary of the 30 days' worth of reflections and the output destination of the generated summary. The "summary" here refers to a sentence that expresses the main points or gist of the 30 days' worth of descriptions entered in record field 403B (see Figure 9).

[0092] The content of the generated summary depends on the generation AI server 50. Therefore, if a different generation AI server 50 is used to generate the summary, the content and expression may differ. Furthermore, even if the same generation AI server 50 is used, the content and expression of the generated summary may differ depending on the time the summary is generated and the version of the generation AI server 50. The instructions in this embodiment do not include an instruction that the "summary" should be generated by quoting from the original text.

[0093] One of the commands is, "In the 'quote', please include one sentence of specific actions related to 'reducing salt intake,' 'weight loss,' 'exercise,' 'sleep,' 'stress,' and 'drinking / smoking.' If you quote more than one sentence, please use commas to clearly indicate the separator." Here, the 'quote' is an example of a quotation. Incidentally, "reducing salt intake" and the like correspond to the behavioral goals of "Step 2" and "Step 3" of the treatment program managed by the patient app. In other words, "reducing salt intake" and the like are examples of management items determined according to the patient app. However, in the case of Figure 12, "drinking alcohol" and "smoking" are combined into one item.

[0094] Here, "reduced salt" is an example of a predetermined item. "Reduce salt intake" is an example of information necessary for medical treatment, and also an example of information necessary for health promotion. Information necessary for medical treatment includes, for example, blood pressure and weight. Information necessary for health promotion includes, for example, behavioral goals managed by a patient app.

[0095] This instruction instructs that statements related to each behavioral goal be extracted in quotations from the "Monthly Review" data. Because it is a citation format, the original data description is extracted as is. Therefore, the quoted description is an example of the partial data that makes up the "Monthly Review" data. The extracted description is used as the basis for the content displayed in the "Total summary" column.

[0096] The extracted descriptions are also used to confirm the patient's descriptions related to the behavioral goals. Furthermore, when extracting multiple descriptions, it is instructed to clearly indicate the separation of each description using commas or the like. Incidentally, instead of commas, periods, bullets, spaces, boundary lines, colons, semicolons, periods, brackets, dates, etc. may also be used. An example of a boundary line is a slash. The separator may also be two or more characters, such as "* / ".

[0097] One of the instructions is, "Please do not summarize the quote, but write it in its original form, even if there are errors." This instruction confirms the role of "quotation" as evidence. One of the instructions is, "If a behavior is repeated, count the number of times and include it in the 'quote'." This instruction allows the frequency of occurrence of the quoted partial data to be understood. Without this instruction, it would be possible to understand whether a description appeared only once in a one-month review, or whether it appeared ten times. For example, descriptions that appear frequently can be used to confirm that they are behaviors or feelings that the patient pays close attention to or is of high interest to.

[0098] The "restrictions" shown in Figure 12 consist of the following three statements: One of the restrictions is "Please keep the output to about 400 characters." In this embodiment, "output" here refers to the number of characters output according to the "output example." This restriction is set for the convenience of checking the summary. Even if the number of characters is reduced compared to the original text and the content is organized, if the number of characters is large (for example, 2000 characters), it will take time to understand the content. In particular, the "summary" in this embodiment is generated so that doctors and others can understand the content of the "reflection" recorded by the patient. For this reason, the total number of characters is limited.

[0099] One of the restrictions is that "summary" should only include the contents written in the diary and should not include any content related to instruction." This restriction is an instruction for generating a summary for each item, such as "reduced salt intake," as specified in the output example. Here, "summary" is also an example of the content portion, similar to the "Total summary" described above. In this embodiment, "Total summary" is a summary of the entire review for one month, while "summary" is a summary for each item. Note that this restriction is also common to "Total summary."

[0100] In this embodiment, "only the content written in the diary" means "only the content written in the retrospective." Here, "only the content" is an instruction regarding the identity with the content written in the retrospective. However, there is no command like that for "quote." In other words, expressions that only quote from the original text or summaries of the quoted content are not prohibited. However, as a limitation on expression, it is required that instructional expressions not be included. Instructional expressions include advice on lifestyle habits, etc. If information or value judgments not contained in the original text, such as instructional expressions, are allowed to be mixed in, the reliability of the "summary" will be impaired. In other words, there is a possibility that doctors and other medical professionals may misunderstand the content of the review. For this reason, this restriction lists instructional expressions as examples of expressions that should be avoided.

[0101] One of the restrictions is to refrain from expressions such as "It seems that..." or "It is inferred that..." and to display only the facts. This restriction is also an instruction for generating a summary by category, such as "reduced salt," as specified in the output example. This restriction is also common to "Total summary." In the present embodiment, expressions such as "it seems that..." and "it is inferred that..." are examples of inferential expressions. Inferential expressions refer to expressions that include fact-based estimations, inferences, predictions, evaluations, and other information other than facts. As with instructional expressions, if information or value judgments not present in the original text are allowed to be mixed in, the reliability of the "summary" will be undermined, so this restriction gives examples of presumptive expressions that should be avoided.

[0102] The "output example" shown in Figure 12 consists of the following two statements: One of the instructions is "Please output in JSON format as shown below." JSON stands for JavaScript Object Notation. One of the statements is, "The JSON file output result should end with [start], [end]." An example format is also provided in the "Example Output" section.

[0103] The "Looking back" data used to generate a summary is described in the "Looking back for one month" shown in Fig. 12. In the present embodiment, one month has 30 days. Therefore, the "Looking back" data for the most recent 30 days is described in the "Looking back for one month."

[0104] <Step 109> Returning to the explanation of Figure 7. The generation AI server 50 generates a summary of the retrospective in accordance with the prompts, and returns the generated summary to the PDT server 20. 13 is a diagram illustrating an example of a JSON file output by the generation AI server 50. The JSON file is written in accordance with the format specified in the prompt.

[0105] For example, the "Total summary" outputs the following summary of the 30-day review: "Average number of steps, slightly insufficient sleep time. High stress, slightly high blood pressure. Attention to reducing salt intake and adjusting portion sizes in meals." This output does not include any instructional or inferential expressions. Furthermore, the output content is based only on the content and facts written in the review. The part "My walking frequency is average" is an excerpt from the "summary" in the exercise category: "My walking frequency is average, and I make a conscious effort to walk." The part "I feel like I don't get enough sleep" is a summary of the "summary" in the sleep category, "It's difficult to get enough sleep on weekdays, so I feel like I don't get enough sleep."

[0106] "I have a lot of stress and my blood pressure is a little high." is a summary of "I have a lot of stress from work and family, and my blood pressure is a little high." in the "summary" of the stress category. "I am conscious of reducing salt intake and adjusting the amount of food I eat" is a summary of "I am conscious of reducing salt intake when I eat" in the "summary" of the salt reduction category and "I am conscious of reducing the amount of food I eat when I eat" in the "summary" of the weight loss category.

[0107] The content of the "summary" for each category is a summary of the content of the corresponding "quote." For example, in the "summary" for the low-salt category, "I am conscious of reducing salt intake in my meals" corresponds to the summary "For breakfast, I have rice porridge with seven herbs. To limit salt intake [4], I don't add salt [8], and I ask everyone to add their own. If I adjust it myself, I become too concerned about it and it doesn't taste good." In Figure 13, the "quote" contains three sentences separated by periods. Each sentence here corresponds to a portion of data quoted verbatim from the retrospective data. The numbers in square brackets indicate the number of times the same character string appears during the review.

[0108] <Step 110> The PDT server 20 presents the acquired summary to the doctor terminal 30 viewing the patient data of "Mr. H." The summary here is displayed, for example, as part of the patient data viewing screen 310 (see FIG. 11). However, the PDT server 20 may also display the acquired summary in a separate window. The separate window display format can be, for example, a pop-up display on the patient data viewing screen 310, or a separate tab.

[0109] 14 is a diagram illustrating a display example of the summary screen 320 according to Embodiment 1. The summary screen 320 shown in FIG. 14 corresponds to the JSON file shown in FIG. In the example display of the summary screen 320 shown in Figure 14, a disclaimer has been added to the end of the sentence, such as "This sentence was generated by AI. The content is not 100% guaranteed." This sentence is added by the PDT server 20.

[0110] <Summary> In this embodiment, a doctor or the like can open a summary screen 320 (see FIG. 14) of the "review" entered through the patient app by operating the patient data viewing screen 310 (see FIG. 11). Specifically, a doctor or the like can display a summary of one month's worth of text entered in the record field 403B (see FIG. 9) on the doctor terminal 30 (see FIG. 1). As a result, it becomes possible to efficiently review the main points and gist of the "review" content, which can be extremely long in total.

[0111] Furthermore, in this embodiment, the summary screen 320 displays not only the overall summary "Total summary," but also individual "summaries" for each category, making it possible to review the main points and gist of the "review" content for each category. Furthermore, the summary in this embodiment outputs only the content of the original text written in the review, and does not include instructional expressions or estimated expressions, thereby preventing the inclusion of information not entered by "H." As a result, the content of the summary allows the user to accurately understand the content entered by "H" in the "review."

[0112] In addition, the summary screen 320 in this embodiment includes a "quote" that displays a description of specific actions related to each category in their original form, so that the description that served as the basis for the generated summary can also be confirmed. Furthermore, if a "quote" contains multiple citations, each citation is displayed with a period to distinguish it from the others, making it easier for doctors and other medical professionals to identify the boundaries of each citation. Furthermore, in the summary screen 320 of this embodiment, when a description that appears repeatedly in the "Looking Back" is quoted in a "quote," the number of occurrences is displayed in association with the corresponding description. This allows doctors and other medical professionals to confirm the content that "Mr. H" repeatedly wrote in the "Looking Back." Furthermore, differences in the frequency of occurrence of each description can be confirmed as differences in the specific number of occurrences.

[0113] <Embodiment 2> A second embodiment will be described. The information processing system according to this embodiment is the same as the information processing system 1 (see FIG. 1) described in the first embodiment. However, in this embodiment, it is assumed that the patient checks the summary of the review that he or she inputs on the patient terminal 40 (see FIG. 1). Therefore, the information processing system does not necessarily require the PDT platform 10 (see FIG. 1) or the doctor terminal 30.

[0114] <Review viewing screen output> The output process of the review viewing screen on the patient terminal 40 (see FIG. 1) will be described below with reference to FIGS. Fig. 15 is a diagram illustrating an example of a processing sequence according to the second embodiment. Note that in Fig. 15, parts corresponding to those in Fig. 7 are assigned the same reference numerals.

[0115] <Steps 101-103> Steps 101 to 103 are the same as those in the first embodiment. That is, the patient records measurement values, reviews, and the like through the patient app running on the patient terminal 40. The patient terminal 40 then uploads the patient app data to the PDT server 20. The PDT server 20 also stores the patient app data.

[0116] <Step 111> The patient terminal 40 accepts a display operation of the inputted review through the patient application. FIG. 16 is a diagram for explaining a method for reading out the inputted review from the menu screen 421. In FIG. The menu screen 421 can be read from the home screen 401 (see FIG. 8). The menu screen 421 shown in FIG. 16 has a "Blood Pressure Measurement" button, a "Review" button, a "Start Program" button, an "Activity Goal List" button, and the like arranged thereon.

[0117] When the "Looking Back" button is operated on the menu screen 421, a viewing screen 422 consisting of a list of entered reviews is displayed. In the case of Fig. 16, the reviews dated 4 / 26, 4 / 25, 4 / 23, and part of the reviews dated 4 / 5 are displayed. Furthermore, a "Summary" button 422A is arranged at the top of the viewing screen 422. The "Summary" button 422A is a button for instructing the display of a summary of the review.

[0118] Fig. 17 is a diagram for explaining another method for reading out the inputted review from the menu screen 421. In Fig. 17, the parts corresponding to Figs. 8 and 16 are assigned the same reference numerals. In the case of Fig. 17, the "My Data" button is operated on the menu screen 421. When My Data is operated, the My Data screen 407 (see Fig. 8) is displayed. As shown in Fig. 8, the My Data screen 407 displays a list of various recorded data. Blood pressure data and retrospective data are displayed on the My Data screen 407 shown in Fig. 17. A "Summary" button 422A is provided in the retrospective data.

[0119] <Step 112> Returning to the explanation of FIG. The patient terminal 40 accepts the display of the summary of the review. Specifically, it accepts the operation of the "Summary" button 422A. Note that the processing operations from step 112 onwards are executed only when the "Summary" button 422A is operated. For example, the patient terminal 40 notifies the PDT server 20 of the acceptance of the operation of the "Summary" button 422A.

[0120] <Step 113> The PDT server 20 instructs the generation AI server 50 to generate a review summary and feedback information. In this embodiment, since the viewer is the patient himself / herself, the instructions include not only a summary of the review but also the generation of feedback information for the patient. The feedback information in this embodiment includes lifestyle habits that need improvement and improved lifestyle habits. In this embodiment, improved lifestyle habits also include lifestyle habits for which efforts to improve are recognized.

[0121] In the case of Figure 15, instructions from the PST server 20 are output to a single generation AI server 50. However, it is also possible to provide separate generation AI servers for generating the retrospective summary and for generating feedback information for the patient. The generation AI server used to generate feedback information is an example of a second trained model. As shown in Figure 15, the generation AI server that generates both the retrospective summary and the feedback information is both the first trained model and the second trained model.

[0122] Fig. 18 is a diagram illustrating an example of a prompt description in embodiment 2. The prompt shown in Fig. 18 also includes a "command," a "restriction," an "output example," and a "monthly review." The "command" shown in Figure 18 consists of the following two statements. One of the commands is "A month's worth of reflection is a diary entry written by the patient in the app." This command indicates that the target of processing is "Reflection." The generation of a summary is indicated by "Total summary" in the "Example output."

[0123] One of the instructions was, "In the "#1MonthReview" below, praise the patient for their efforts in improving their lifestyle habits and encourage them to complete the tasks from the next session onwards." "Praise the patient for their efforts in improving their lifestyle habits" is an example of an instruction to generate "improved lifestyle habits" in the feedback information. In this embodiment, the instruction goes beyond pointing out the improved lifestyle habits to also include the use of expressions that praise the patient.

[0124] Furthermore, "Please encourage the person to complete the assignment from the next time onwards" is an example of an instruction to generate "lifestyle habits that need improvement" in the feedback information. In this embodiment, the instruction is not limited to pointing out lifestyle habits that need improvement, but also instructs the adoption of expressions that encourage improvement. In the case of FIG. 18, unlike the first embodiment, commands related to "quote" and commands related to outputting the number of times the action is repeated are not included.

[0125] The "restriction" shown in Figure 18 consists of the following single statement: The restriction here is "Please limit output to approximately 400 characters." This content is the same as the prompt in the first embodiment. On the other hand, it does not include the generation of a summary as in the first embodiment, one of the reasons being that the viewer is the patient himself / herself. The output example shown in FIG. 18 is the same as the prompt in the first embodiment. The format of the output example shown in FIG. 18 is the same as the prompt in the first embodiment.

[0126] <Step 114> Returning to the explanation of FIG. The generation AI server 50 generates the reflective summary and feedback information according to the prompts. FIG. 19 is a diagram illustrating an example of a JSON file output by the generation AI server 50. In the case of the output example shown in FIG. 19, not only the "Total summary" shown in the format of the output example is included, but also "Improvements" and "Successes" corresponding to the prompt instructions.

[0127] The "Total summary" summarizes the past month, outputting the following: "Over the past month, I have been consciously working on my diet and exercise, and although there were times when I felt stressed, my blood pressure figures have improved by increasing the number of steps I take and being careful to reduce salt intake. It is important to continue to eat a balanced diet and exercise, and to avoid accumulating stress." This content is consistent with the content of the abstract and citations presented on the doctor terminal 30 (see FIG. 1).

[0128] In the case of Figure 19, the "Improvements" section, which indicates the issues, outputs the following: "If you cannot take the time to cook at home in order to reduce salt intake, try to use less of the seasonings that come with the meal. Also, try to incorporate exercise into your work schedule to avoid accumulating stress." This output includes statements encouraging specific actions, such as "If..., try to do..." and "Try to do...." Additionally, in "Successes," which indicates improved lifestyle habits, the output states, "By increasing the number of steps, your blood pressure readings have improved. You've also started to consider reducing salt intake when eating out." In this example output, no expression praising the patient's efforts can be seen, but it would be possible to include an expression encouraging the patient to make the habit a habit, such as, "By increasing the number of steps, your blood pressure readings have improved. You've also started to consider reducing salt intake when eating out." or, at the end of the sentence, "Let's keep it up."

[0129] <Step 115> Returning to the explanation of FIG. The PDT server 20 presents the acquired summary and feedback information to the patient terminal 40 operated by "Mr. H." 20 is a diagram illustrating a display example of the summary screen 430 in Embodiment 2. The summary screen 430 shown in FIG. 20 corresponds to the JSON file shown in FIG.

[0130] 20 also displays a target period 431. The display columns of the summary screen 430 are divided into three sections: a "Summary" column 432, an "Improvement Points" column 433, and an "Good Things" column 434. The columns correspond to "Total summary," "Improvements," and "Successes." In addition, a disclaimer such as "This text was generated by AI. The content is not 100% guaranteed" has been added to the end of the summary screen 430 shown in Figure 20. The PDT server 20 adds this statement.

[0131] <Summary> In this embodiment, the patient "Mr. H" can open a summary screen 430 (see FIG. 20) of the "Looking Back" input through the patient app by operating the patient terminal 40 (see FIG. 1). Specifically, the patient "Mr. H" can display a summary of one month's worth of text input in the record field 403B (see FIG. 9) on the patient terminal 40 (see FIG. 1). As a result, it becomes possible to efficiently review the main points and gist of the "Looking Back" content, which can be extremely long in total.

[0132] In this embodiment, the summary screen 430 includes not only a summary of the review but also a description of lifestyle habits that need to be improved and lifestyle habits that have been improved. Therefore, the patient can specifically check the issues with his / her lifestyle habits and the lifestyle habits that have been improved, in addition to the summary of the review that he / she inputted. As a result, "Mr. H" can become aware of the issues with his / her lifestyle habits. At the same time, "Mr. H"'s sense of competence can be enhanced through feedback on his / her improved lifestyle habits.

[0133] <Other embodiments> (1) Although the embodiments of the present disclosure have been described above, the technical scope of the present disclosure is not limited to the scope of the above-described embodiments. It is clear from the claims that various modifications or improvements to the above-described embodiments are also included in the technical scope of the present disclosure.

[0134] (2) The processor in the above-described embodiments refers to a processor in a broad sense, and includes general-purpose processors (e.g., CPUs (=Central Processing Units)) as well as dedicated processors (e.g., GPUs (=Graphical Processing Units), ASICs (=Application Specific Integrated Circuits), FPGAs (=Field Programmable Gate Arrays), programmable logic devices, etc.). Furthermore, the operations of the processors in each of the above-described embodiments may be performed by multiple processors working together, rather than by a single processor. The order in which the processors perform the operations is not limited to the order described above, and may be changed individually.

[0135] (3) In the above-described embodiment, the calculation requirements for medical fees under the public medical insurance system are assumed, but information regarding the applicability of medical fees under the private medical insurance system may also be displayed on the work screen of the doctor terminal 30 (see FIG. 1). For example, if the requirements for approval of calculation of medical fees under the private medical insurance system require the use of a specific app or medical treatment that refers to patient data entered through a specific app, the explanation of the above-described embodiment can be applied.

[0136] (4) The above-described embodiment assumes that the target period of the "review" for generating summaries and the like is specified in advance. However, the target period may be specified by the doctor or patient who instructs the display of the summary. For example, when the "Summary" button 314B (see FIG. 11) or the "Summary" button 422A (see FIGS. 16 and 17) is operated, a screen for specifying the target period may be displayed. The target period may be specified not only in number of days, but also in number of months or years. Also, it may be possible to specify a specific month (e.g., May) or a specific year (e.g., 2023).

[0137] (5) In the above embodiment, the prompt "output example" includes a description at the end that indicates the period covered by the summary. However, it is also possible to extract only the “review” data corresponding to the target period and output it to the generation AI server 50.

[0138] (6) In the above-described embodiment, when the "Summarize" button is pressed, a summary of the "Review" data is generated in accordance with a prepared prompt. However, the doctor or the patient may be able to specify the contents of the "commands" and "restrictions" to be written in the prompt. For example, when the "Summary" button 314B or the "Summary" button 422A is operated, a screen for specifying the contents of the "commands" or the "restrictions" may be displayed on the operation screen. The specification here includes the selection of options.

[0139] (7) In the case of the first embodiment described above, when a description that appears repeatedly in the "Looking Back" is quoted in the "quote," the number of occurrences is displayed in association with the corresponding description. However, the number of occurrences may be counted not only for the same expression but also for similar expressions as long as the meaning is the same.

[0140] (8) In the case of the first embodiment described above, when a description that appears repeatedly in the "Looking Back" is quoted in the "quote," the number of occurrences is displayed in association with the corresponding description. However, the number of occurrences does not have to be displayed in the summary. 21 is a diagram illustrating a summary screen 320 that does not include a display of the number of occurrences. In FIG. 21, the parts corresponding to those in FIG. 14 are denoted by the same reference numerals.

[0141] (9) In the case of the first embodiment described above, when a description that appears repeatedly in the "Looking Back" is quoted in the "quote," the number of occurrences is displayed in association with the corresponding description. However, the number of occurrences may be displayed for the "Total summary" or "summary" descriptions. 22 is a diagram illustrating a summary screen 320 that includes a display of the number of occurrences in the "Total summary." In FIG. 22, the same reference numerals are used to denote parts corresponding to those in FIG.

[0142] (10) In the first embodiment described above, the summary screen 320 (see FIG. 14) is displayed as a separate window. However, the retrospective summary may also be displayed as part of the patient data view screen 310 (see FIG. 11). 23 is a diagram illustrating another example of the examination date registration screen 310A. In FIG. 23, parts corresponding to those in FIG. 11 are denoted by the same reference numerals.

[0143] The examination date registration screen 310A shown in FIG. 23 is displayed when, for example, the "Go to examination date registration screen" button 316 (see FIG. 11) is operated. The consultation date registration screen 310A is composed of a "patient information column" 311, a column 321 displaying appointments with the patient, an insurance calculation information column 322, a consultation history column 323, a "return to previous screen" button 324, and an "end consultation" button 325.

[0144] In the case of FIG. 23, the display field 321 for the appointment with the patient displays the "history of the appointment contents" and the "review summary" made between the doctor and the patient at the previous consultation. The "promise details" can be registered by the patient through the patient app, based on what the doctor or other medical professional has communicated to them by hand or orally, or the doctor or other medical professional can register it through the doctor screen. In the case of Figure 23, the target period is "April 30, 2024 to May 29, 2024." The promise was "I will pay attention to my diet and try to improve my exercise and sleep." Meanwhile, the "reflection summary" was "My walking frequency is average, but my sleep time is a little insufficient. I have a lot of stress and my blood pressure is a little high. I am conscious of reducing salt intake and adjusting the amount of food I eat."

[0145] In this way, by displaying the contents of the "Last Appointment" and the "Review Summary," which is the patient's record, side by side, doctors and other medical professionals can easily understand how the patient is progressing with the previous appointment. The date of the current consultation and a selection button for whether or not you wish to have insurance calculated for this consultation are displayed in the insurance calculation information field 322. In the insurance calculation information field 322 in Figure 23, "Yes" is selected.

[0146] Furthermore, the medical examination history column 323 displays a number of medical examination dates, including the most recent medical examination date, and the results of insurance calculations. The "return to previous screen" button 324 is a button for returning to the patient data viewing screen 310 (see FIG. 11). The "End Consultation" button 325 is a button for closing the consultation date registration screen 310A and ending the consultation. Closing the consultation date registration screen 310A also means closing the viewing screen 310 (see FIG. 11).

[0147] (11) The summary screen 430 (see FIG. 20) in the second embodiment described above displays only the contents of the JSON file generated by the generation AI server 50 (see FIG. 15). That is, three columns are displayed: a “Summary” column 432, an “Improvement Points” column 433, and a “Good Points” column 434. However, a new goal input field may be displayed in association with the "improvement points" field 433. Fig. 24 is a diagram illustrating another display example of the summary screen 430. In Fig. 24, parts corresponding to those in Fig. 20 are assigned the same reference numerals.

[0148] In the case of the summary screen 430 shown in Fig. 24, a new goal input field 435 is arranged between the "points for improvement" field 433 and the "good things" field 434. In the case of Fig. 24, the new goal input field 435 has the sentence "Next goal (set a behavioral goal)" and a blank field for inputting a goal arranged therein. Displaying a new goal input field 435 makes it possible to encourage patients to take autonomous action. Also, by confirming the achievement of the goals they themselves entered in the "Good points" field 434, it is possible to give the patient intrinsic motivation and a sense of competence.

[0149] (12) In the case of the second embodiment described above, the display of the summary of the "review" is premised on the operation of the "summary" button 422A (see FIGS. 16 and 17). However, if the "review" summary is generated automatically without any operation by the patient, a mechanism may be adopted in which a message is displayed on the home screen 401 (see FIG. 9) indicating that a new summary is available for viewing. Fig. 25 is a diagram illustrating another display example of the home screen 401. In Fig. 25, parts corresponding to those in Fig. 9 are assigned the same reference numerals.

[0150] 25, a summary notification field 401B is added below the input button 401A. Note that this summary notification field 401B is added when a review summary is automatically generated. Therefore, if there is no review summary that the patient has not confirmed, only the input button 401A is displayed, similar to the home screen 401 shown in FIG. The summary notification field 401B shown in Fig. 25 has a message saying "April review summary is available" and an "Open" button. In this example, when the "Open" button is operated, the patient application displays the summary screen 430 (see Figs. 20 and 24).

[0151] (13) In the above embodiment, the PDT server 20 instructs the generation AI server 50 to generate a summary. However, the doctor terminal 30 or the patient terminal 40 may instruct the generation AI server 50 to generate a summary. 26 is a diagram illustrating an embodiment in which a patient terminal 40 directly instructs the generation AI server 50 to generate a summary. In FIG. 26, parts corresponding to those in FIG. 15 are assigned the same reference numerals.

[0152] 26, the processing operations of step 101 and steps 111 to 113 are all executed by the patient terminal 40. Specifically, when a review summary display is received through the patient app, the patient terminal 40 instructs the generation AI server 50 to generate a review summary and feedback information. Since the request originates from the patient terminal 40, the generation AI server 50 returns the generated summary and feedback information directly to the patient terminal 40.

[0153] The patient terminal 40 also presents the acquired summary and feedback information. This mechanism can also be employed when a mechanism for uploading patient application data to the PDT server 20 is employed, as in the above-described embodiment. 26 can also be used when viewing a review summary on the doctor terminal 30. However, this is only when the doctor terminal 30 can access or already stores the review data from which the summary is to be generated. The doctor terminal 30 and the patient terminal 40 here are examples of information processing devices.

[0154] (14) In the above-described embodiment, a viewing screen for blood pressure values ​​recorded through a patient app prescribed for the treatment of hypertension is assumed, but for other diseases, viewing screens for different measurement values ​​are prepared. For example, in the case of nicotine addiction, a viewing screen is prepared that displays, for example, the amount of nicotine intake, the number of times a medium containing nicotine has been taken, the number of times a medium containing nicotine has been taken, and the measured value of the CO concentration in the breath.

[0155] In addition, if the person is an alcoholic, a screen for viewing the measured amount of alcohol intake is prepared. In addition, for NASH or obesity, a screen for viewing weight measurements, for example, is provided. In addition, for insomnia, a viewing screen is provided for recording, for example, the amount of sleeping pills taken and the total score of the Pittsburgh Sleep Quality Index (PSQI), which is defined by 24 items. Other diseases contemplated include, for example, kidney disease, cancer, chronic heart failure, attention deficit hyperactivity disorder, depression, tinnitus, delayed grief disorder, opioid-induced constipation, post-mastectomy pain syndrome, bronchial asthma, and nephrotic syndrome.

[0156] (15) In the above-described embodiment, it is assumed that all of the “reviews” from the start date of recording are recorded through the patient app. In other words, it is assumed that the “reviews” are recorded by a program located in a medical device. However, in addition to the "reflection" recorded through a program located in a medical device, the "reflection" recorded through a program located in a non-medical device may also be included. The program located in the non-medical device is a program that has a "reflection" recording function.

[0157] In the following, programs that are classified as non-medical devices will be referred to as "non-patient apps." FIG. 27 is a diagram illustrating an embodiment in which the "review" recorded through the non-patient app and the "review" recorded through the patient app are used in combination. The horizontal axis in Figure 27 is the time axis. The right end of the time axis is the consultation date, and the left end is the past tense. Figure 27 shows up to 7 months ago. In the case of Figure 27, the prescription date of the patient app is between three months ago and four months ago. Before being prescribed the patient app, the patient (i.e., the user) records a "review" via the non-patient app. The "review" may be recorded only on the patient terminal 40, only on the management server 60, or on both the patient terminal 40 and the management server 60. The management server 60 is, for example, a server operated by a business operator that provides the non-patient app.

[0158] Non-patient apps in this context are sometimes called healthcare apps. Healthcare apps are programs whose primary purpose is to record health information and are not medical devices. In the case of Fig. 27, the "review" recorded through the non-patient app before the patient app was prescribed is imported from the patient terminal 40 or the management server 60 to the PDT server 20. Of course, when importing, the identity of the user is confirmed by a user account or the like.

[0159] Therefore, on the day of the consultation, the PDT server 20 stores not only about four months of "reviews" after the patient app is prescribed, but also about three months of "reviews" before the patient app is prescribed. In other words, the PDT server 20 stores about seven months of "reviews" of the patient. In the case of the above-described embodiment, the "review" summaries displayed on the doctor terminal 30 and the patient terminal 40 cover at most four months after the prescription date. However, in the case of the embodiment described with reference to FIG. 27, it is also possible to view a summary of the "review" before starting treatment using the patient app.

[0160] (16) In the above-described embodiment, it is assumed that the consultation date falls within the insured medical treatment period. However, it may also be possible to view summaries of “reviews” recorded through a program classified as a non-medical device even after the insured medical treatment period. FIG. 28 is a diagram illustrating another embodiment in which the "review" recorded through the non-patient app and the "review" recorded through the patient app are used in combination.

[0161] The horizontal axis in Figure 28 is the time axis. The right end of the time axis is the consultation date, and the left end is the past tense. Figure 28 shows up to 7 months ago. In the case of Figure 28, the prescription date in the patient app is 7 months ago, and the insured medical treatment period expired within the period 2 months before the consultation date. In the case of Figure 28, after the expiration of the insured medical treatment period, a medical treatment period not covered by insurance (i.e., outside the insured medical treatment period) begins.

[0162] 28, it is assumed that the patient app continues to be used after the validity period has expired in order to record the "review" outside the insured medical treatment period. In other words, it is assumed that the patient app continues to be used as a non-patient app after the validity period has expired. For this reason, not only the "review" for the insured medical treatment period but also the "review" for the period outside the insured medical treatment period are recorded in the PDT server 20.

[0163] However, the "review" may be recorded on the patient terminal 40 (see FIG. 27) or the management server 60 (see FIG. 27) using a non-patient app different from the patient app used during the insured medical treatment period. If the "review" recorded through the non-patient app is not stored in the PDT server 20, the "review" is imported from the patient terminal 40 or the management server 60 to the PDT server 20, and then the summary generation process described above is started.

[0164] (17) In the above-described embodiment, the summary of the "review" is generated by the generation AI server 50 (FIGS. 7, 15, etc.). However, if the PDT platform 10 (see Figure 1), PDT server 20 (see Figure 1), doctor terminal 30 (see Figure 1), and patient terminal 40 (see Figure 1) have a trained model, any of these terminals may generate a "review" summary.

[0165] (18) In the above-described embodiment, the patient app is assumed to be a program located in a medical device, but it may be a program located in a non-medical device. In other words, it may be assumed that all "reviews" are recorded through a program located in a non-medical device. In this case, similarly to the above-described embodiment, the generation AI server 50 (see FIGS. 7 and 15) or the like may generate the "review."

[0166] (19) In the above-described embodiment, a summary of the "review" of the most recent month is generated. However, the period of the "review" for which the summary is to be generated may be freely specified by the doctor, patient, etc. through a user interface. For example, the period from the last medical consultation date to the present day, or the period from the first medical consultation date to the present day, may be specified.

[0167] (20) In the above embodiment, the case where the summary is generated is a "review." However, the name of the data for which a summary is to be generated is not limited to "review," but may be "diary" or "memo."

[0168] (21) In the above embodiment, the case where the app used to record the “review” for which a summary is to be generated is a patient app or a healthcare app has been described. However, the type of app used to record health-related data is not limited, and the type of app may be, for example, a diary app, an image app, or an audio app.

[0169] (22) In the above embodiment, text data is assumed to be the target for generating summaries, but any type of data can be used as long as it is health-related data. For example, image data, audio data, or numerical data can be used. For example, in the case of treating dermatitis, the changes in images of the affected area (e.g., the inflamed area) captured by the patient may be presented as a summary. The summary here may be, for example, an extracted image showing the change. In addition, in the case of treating mental illness, the summary may be a summary of characteristic parts of audio or video recorded by the patient that meet predetermined criteria. The summary here may be, for example, an extracted characteristic utterance that should be noted from the perspective of promoting health or treating the patient. In the case of audio, the extracted audio is not limited to being played back, and it can also be displayed as text.

[0170] (23) The prompts described in the above embodiments command the output of summaries and quotations for each item, such as “reduced salt intake,” managed by the patient app, etc. In other words, the items indicated by the prompts are fixedly determined depending on the patient app, etc. However, doctors and patients may want to selectively check the patient's records related to the goals they are currently working on. For this reason, summaries may be presented in association with the goals for the target period. For example, a patient app may manage "reducing salt intake," "weight loss," "exercise," "sleep," "stress," and "alcohol / smoking." If the patient's behavioral goals for the past month were "reducing salt intake" and "weight loss," it may present only summaries related to "reducing salt intake" and "weight loss."

[0171] (24) In the above embodiment, the case where the generation of a summary is instructed to the generation AI server 50 (see FIG. 7) has been described. That is, the case where the generation of a summary is instructed to a dedicated server has been described. However, a trained model that has been machine-learned to understand the input / output relationship in which documents, images, sounds, and other content are input and summaries of those contents are output may be implemented in the PDT platform 10 (see FIG. 1), PDT server 20 (see FIG. 1), doctor terminal 30 (see FIG. 1), and patient terminal 40 (see FIG. 1). In this case, it is not necessary to instruct the generation AI server 50 to generate a summary. For example, the summary may be generated in the doctor terminal 30 or patient terminal 40 that outputs the summary.

[0172] <Summary> The disclosed examples described in the above-mentioned embodiments are shown below. (((1))) An information processing device having one or more processors, which acquire health-related data input by a user who is changing their behavior, instruct a trained model to generate a summary of the health-related data, and present the generated summary on a viewing screen. This information processing device can assist users who change their behavior in reviewing records entered by them.

[0173] (((2))) The information processing device described in (((1))), wherein the one or more processors instruct the trained model to generate a summary by citing one or more partial data extracted from health-related data. According to this information processing device, it is possible to avoid mixing in information that has not been entered by a user whose behavior is changing.

[0174] (((3))) The information processing device described in (((2))), wherein the one or more processors instruct the trained model to distinguish between one piece of partial data and other pieces of partial data. According to this information processing device, citation units can be presented in an identifiable manner.

[0175] (((4))) The information processing device described in (((2))), wherein one or more processors instruct the trained model to generate the summary including the number of occurrences of the partial data. According to this information processing device, the relative position of a specific user can be grasped from the viewpoint of the similarity or difference in frequency distribution.

[0176] (((5))) The information processing device according to (((2))), wherein the one or more processors instruct the trained model to generate a summary that does not include the estimated expression. According to this information processing device, it is possible to avoid mixing in information that has not been entered by a user whose behavior is changing.

[0177] (((6))) The information processing device according to (((2))), wherein the one or more processors instruct the trained model to generate a summary that does not include instructional expressions. According to this information processing device, it is possible to avoid mixing in information that has not been entered by a user whose behavior is changing.

[0178] (((7))) The information processing device described in (((1))), wherein the one or more processors instruct the trained model to generate a summary in units of predetermined items. This information processing device can support a user who is changing his or her behavior in reviewing each item of the records.

[0179] (((8))) The information processing device described in (((7))), wherein the one or more processors instruct the trained model to use lifestyle behaviors determined according to a program used to record health-related data as predetermined items. This information processing device can assist in reviewing records related to items according to the program.

[0180] (((9))) The information processing device according to (((1))), wherein the one or more processors instruct the trained model on the period for generating the summary. According to this information processing device, it is possible to clarify the target period used for generating a summary.

[0181] (((10))) The information processing device according to (((1))), wherein the one or more processors present a summary in association with a goal for a period corresponding to the health-related data. According to this information processing device, the goal and summary for the corresponding period can be presented in a comparable manner.

[0182] (((11))) The information processing device described in (((1))), wherein the one or more processors instruct the second trained model to generate lifestyle challenges based on health data and present the challenges on the second viewing screen. This information processing device can present lifestyle-related issues in addition to summaries.

[0183] (((12))) The information processing device according to (((11))), wherein the one or more processors present an input field for a new goal in association with the task. This information processing device can support improvements in lifestyle habits.

[0184] (((13))) The information processing device described in (((1))), wherein the one or more processors instruct the second trained model to generate information indicating improved lifestyle habits based on health data, and present the information indicating the improved lifestyle habits on the second viewing screen. This information processing device can help people establish good lifestyle habits.

[0185] (((14))) The information processing device described in (((1))), wherein the one or more processors instruct the trained model to generate the summary including a description indicating the number of occurrences of content that appears multiple times. This information processing device can assist in understanding the frequency of occurrence of citation units.

[0186] (((15))) The information processing device according to any one of (((1))) to (((14))), wherein the summary is composed of a content portion and a quote portion. According to this information processing device, the reliability of the summary can be improved.

[0187] (((16))) The information processing device described in (((15))), wherein the one or more processors instruct the trained model to prohibit the citation of one or more partial data extracted from health-related data by summarizing them as a generation condition for the quoted portion. According to this information processing device, it is possible to present quoted portions in which the original data is quoted as is.

[0188] (((17))) An information provision method in which a computer performs the following processes: acquiring health-related data entered by a user who is changing their behavior; instructing a trained model to generate a summary of the health-related data; and presenting the generated summary on a viewing screen. This information providing method can assist users who are changing their behavior in reviewing records entered by them.

[0189] (((18))) A program that enables a computer to acquire health-related data entered by a user who is changing their behavior, instruct a trained model to generate a summary of the health-related data, and display the generated summary on a viewing screen. This program can assist in reviewing records entered by users who are changing their behavior. [Explanation of symbols]

[0190] 10...PDT platform, 20, 20A, 20B, 20C, 20D, 20E, 20F...PDT server, 30...doctor's terminal, 40...patient's terminal, 50...generation AI server, 60...management server

Claims

1. having one or more processors; the one or more processors: Obtain health data entered by users who want to change their behavior; instructing the trained model to generate a summary of the health-related data; presenting the generated summary on a viewing screen; Information processing device.

2. the one or more processors: instructing a trained model to generate the summary by citing one or more partial data segments extracted from the health-related data; The information processing device according to claim 1 .

3. the one or more processors: Instructing the trained model to distinguish between one of the partial data and another of the partial data; The information processing device according to claim 2 .

4. the one or more processors: Instructing the trained model to generate the summary including the number of occurrences of the partial data; The information processing device according to claim 2 .

5. the one or more processors: instructing the trained model to generate the summary without including estimated expressions; The information processing device according to claim 2 .

6. the one or more processors: instructing the trained model to generate the summary without including instructional expressions; The information processing device according to claim 2 .

7. the one or more processors: Instructing the trained model to generate the summary on a predetermined item basis; The information processing device according to claim 1 .

8. the one or more processors: Instructing the trained model as the predetermined items management items determined according to a program used to record the health-related data; The information processing device according to claim 7 .

9. the one or more processors: Instructing the trained model on a time period for generating the summary. The information processing device according to claim 1 .

10. the one or more processors: presenting the summary in relation to a time period goal corresponding to the health-related data; The information processing device according to claim 1 .

11. the one or more processors: Instructing a second trained model to generate lifestyle challenges based on the health data; Presenting the assignment on a second viewing screen; The information processing device according to claim 1 .

12. the one or more processors: Present an input field for a new goal in relation to the task; The information processing device according to claim 11.

13. the one or more processors: instructing a second trained model to generate information indicative of improved lifestyle habits based on the health-related data; presenting information indicating the improved lifestyle habits on a second viewing screen; The information processing device according to claim 1 .

14. the one or more processors: Instructing the trained model to generate the summary including a description indicating the number of occurrences of content that appears multiple times. The information processing device according to claim 1 .

15. The abstract includes a content section and a citation section. The information processing device according to any one of claims 1 to 14.

16. the one or more processors: Instructing the trained model to prohibit citing a summary of one or more partial data extracted from the health-related data as a generation condition for the quoted portion. The information processing device according to claim 15.

17. The computer Obtaining health-related data entered by a behavior-changing user; instructing the trained model to generate a summary of the health-related data; presenting the generated summary on a viewing screen; Information provision method to carry out.

18. On the computer, A function to capture health data entered by a user who modifies their behavior; Instructing a trained model to generate a summary of the health-related data; a function of displaying the generated summary on a viewing screen; A program to achieve this.

Images