Medical support system, medical support method, and medical support program

Abstract

The present invention relates to a medical support system, a medical support method, and a medical support program.SOLUTION: A medical support system 1 includes an acquisition unit, a generation unit, and a display processing unit. The acquisition unit acquires questions about a disease posted by a plurality of users and disease response information corresponding to the questions. A generation part generates answer support for guiding an answer candidate to the question on the basis of the question and the disease correspondence information, and a display processing part performs display processing of the answer support, and transmits the display processing result.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a medical support system, a medical support method, and a medical support program. [Background technology]

[0002] In recent years, various systems have been developed to provide support in the fields of medicine and health.

[0003] For example, Patent Document 1 discloses a health management system that uses chat groups to control a conversation bot that encourages users to improve their health. This health management system not only promotes communication between host users and guest users, but also provides health advice to the host user. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2019-133397 Summary of the Invention [Problem to be solved by the invention]

[0005] As described above, health advice and the like can be provided to users by using the health promotion system of Patent Document 1. However, in the medical field, which requires advanced knowledge and experience, it is difficult for the technology of Patent Document 1 to derive an appropriate answer (diagnosis) to a patient's question (disease, etc.) from the limited information provided by the patient.

[0006] In particular, when determining whether a patient has a rare disease with a small number of cases, even medical professionals with a certain amount of experience have difficulty recognizing the disease. Furthermore, accurately confirming that a disease is rare may require excessive questioning of the patient, which places a burden on the patient.

[0007] The present invention has been made in consideration of the above-mentioned problems with the conventional technology, and its purpose is to provide a medical support system that can provide appropriate and highly accurate answers to patients' questions without placing a burden on the patients. [Means for solving the problem]

[0008] In order to solve the above problems, the present invention provides a medical support system for providing support related to medical care, comprising: the medical support system includes an acquisition unit, a generation unit, and a display processing unit; the acquiring unit acquires questions about diseases posted by a plurality of users and disease response information corresponding to the questions; the generation unit generates answer support for guiding answer candidates to the question based on the question and the disease correspondence information; The display processing unit processes the response support to display, and transmits the display processing result.

[0009] The present invention also provides a medical support method executed by a medical support system for providing support related to medical care, comprising: the medical support system includes an acquisition unit, a generation unit, and a display processing unit; The acquisition unit acquires questions about diseases posted by a plurality of users and disease response information corresponding to the questions; generating answer support for guiding answer candidates to the question based on the question and the disease correspondence information by the generating unit; The display processing unit displays the answer assistance and transmits a result of the display processing.

[0010] The present invention also provides a medical support program for providing support related to medical care, comprising: causing a computer to function as an acquisition unit, a generation unit, and a display processing unit; the acquiring unit acquires questions about diseases posted by a plurality of users and disease response information corresponding to the questions; the generation unit generates answer support for guiding answer candidates to the question based on the question and the disease correspondence information; The display processing unit processes the response support to display, and transmits the display processing result.

[0011] This configuration makes it possible to generate appropriate answer support (support necessary to obtain an answer, additional questions, etc.), and as a result, appropriate and highly accurate answer candidates (potential diseases, test suggestions, potential diagnosis results, etc.) can be provided to patients in response to their questions without placing a burden on them.

[0012] For example, if the answer (diagnosis) to a patient's question is about a rare disease, it is difficult to arrive at an accurate answer (diagnosis) by simply interviewing the patient. Therefore, according to the present invention, by generating answer support using the posted dialogue between users, it is possible to provide highly accurate answer candidates to address the patient's concerns without placing a burden on the patient.

[0013] In a preferred embodiment of the present invention, the acquisition unit acquires the questions and the disease response information posted from a plurality of users in a time-series format in which the questions are displayed in chronological order of posting, and The display processing unit processes the response assistance to be displayed in the time series format and transmits the display processing result.

[0014] By configuring in this way, answer support can be generated using conversations between users posted in chronological order, making it possible to provide highly accurate answer candidates to address the concerns of patients without placing a burden on them.

[0015] For example, by utilizing conversations between users (exchanges of information about diseases posted by multiple users) posted in a chronological format such as a timeline or thread, it is possible to provide highly accurate answer candidates to address the concerns of patients without placing a burden on them.

[0016] In a preferred embodiment of the present invention, the generation unit inputs the question and the disease correspondence information into a generation model, and generates answer support based on the output of the generation model.

[0017] With this configuration, it is possible to generate answer assistance for obtaining more accurate answer candidates by using a specific generation model.

[0018] In a preferred embodiment of the present invention, the question and the disease response information include text information, The generation unit inputs the question and the disease correspondence information into a generation model, and generates the answer support based on the character information extracted by the generation model.

[0019] This configuration makes it possible to generate answer support that uses text information to obtain more accurate answer candidates. For example, text can be extracted by text mining from questions and disease response information posted in a timeline format (chronological format), and appropriate answer support can be generated by using that text and a generation model.

[0020] In a preferred embodiment of the present invention, the medical support system includes a storage unit, the storage unit stores attribute information of the user; The generation unit generates the answer assistance based on the question, the disease response information, and attribute information linked to each user.

[0021] With this configuration, it is possible to generate more appropriate answer support based on the attribute information.

[0022] In a preferred embodiment of the present invention, the generation unit inputs the question, the disease response information, and attribute information associated with each user into the generation model, and selects additional users who can post answer candidates to the question; generating the answer assistant including the selected additional users; The display processing unit displays the reply support including the additional users in a time series format in which the reply support is displayed in chronological order of posting, and transmits the display processing result.

[0023] This configuration allows the patient to have a dialogue with additional users who can address the patient's questions (for example, experts on a specific disease or users who have had the same disease in the past), thereby providing more appropriate answer candidates.

[0024] In a preferred embodiment of the present invention, the medical support system further includes a calculation unit, the calculation unit calculates a score of the user using the attribute information, associates the score with the user, and stores the score in the storage unit; The generation unit inputs the question, the disease correspondence information, and the score into the generation model, and selects additional users who can post answer candidates to the question.

[0025] With this configuration, it is possible to select more appropriate additional users by utilizing a specific score.

[0026] In a preferred embodiment of the present invention, the medical support system further comprises a storage unit, the storage unit stores test information related to a disease associated with an organization to which the user belongs; the generation unit selects candidate tests for deriving candidate answers to the question based on the question, the disease association information, and the test information; The display processing unit displays the answer support including the test candidates in a time series format in which the answers are displayed in chronological order of posting, and transmits the display processing results.

[0027] With this configuration, it is possible to select test candidates that will lead to appropriate answer candidates.

[0028] In a preferred embodiment of the present invention, the medical support system further comprises a storage unit, The storage unit stores examination information and map information related to diseases associated with an organization to which the user belongs; the generation unit selects candidate tests for deriving candidate answers to the question based on the question, the disease association information, and the test information; The display processing unit displays the candidate examinations together with the user's location information on a map based on the map information, and transmits the display processing results.

[0029] This configuration allows the user to more appropriately identify candidate examinations. For example, if there is a hospital or the like close to the user's location, the user can easily and quickly identify the hospital or the like.

[0030] In a preferred embodiment of the present invention, the generation unit generates a follow-up question for deriving answer candidates for the question based on the question and the disease correspondence information, The display processing unit processes and displays the answer support including the follow-up question in a chronological format in which the answers are displayed in chronological order of posting, and transmits the display processing results.

[0031] By configuring in this way, it is possible to generate follow-up questions that will lead to appropriate answer candidates.

[0032] In a preferred embodiment of the present invention, the generation unit is provided with a plurality of support modes, The normal medical consultation mode, the rare disease response mode, and the infectious disease epidemiology mode can be switched by automatic determination in the generation process of the answer support or manual operation by the user, The search target is changed according to a predetermined support mode.

[0033] This configuration allows for more accurate answer support. For example, after selecting a support mode, more accurate answer support can be obtained by using a search based on Retrieval Augmented Generation (RAG). [Effects of the Invention]

[0034] According to the present invention, a novel technique relating to a medical support system can be provided by executing a predetermined process using disease response information. [Brief explanation of the drawings]

[0035] [Figure 1] 1 shows a block diagram of a system configuration according to an embodiment of the present invention. [Figure 2] 1A and 1B are schematic diagrams illustrating an example of the hardware configuration of an information processing device and a terminal according to an embodiment of the present invention. [Figure 3] 1 shows a flowchart of a processing procedure of a medical support system according to an embodiment of the present invention. [Figure 4] 3 shows an example of each piece of information used in the medical support system according to one embodiment of the present invention. [Figure 5] 10 shows an example of a display screen of a user terminal according to an embodiment of the present invention. [Figure 6] 1 shows a schematic image of an information processing device and a generating device according to an embodiment of the present invention. [Figure 7] 10 shows an example of a display screen displaying a follow-up question according to one embodiment of the present invention. [Figure 8] 1 shows an image of a schematic algorithm for suggesting follow-up questions according to one embodiment of the present invention. [Figure 9] 10 shows an example of a display screen displaying an additional user according to an embodiment of the present invention. [Figure 10] 10 shows an example of a display screen displaying examination candidates according to an embodiment of the present invention. [Figure 11] 10 shows an example of a display screen displaying multiple response support according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0036] The present invention will now be described more fully with reference to the accompanying drawings, in which preferred embodiments are shown, but which may be embodied in many different forms and are not limited to the embodiments set forth herein.

[0037] For example, although the configuration, operation, etc. of the medical support system are described in this embodiment, similar effects can be achieved by an executed method, device, computer program, etc. The program in this embodiment may be provided as a non-transitory computer-readable recording medium, or may be provided so as to be downloadable from an external server.

[0038] Furthermore, in this embodiment, the term "unit" may include, for example, a combination of hardware resources implemented by circuits in a broad sense and software information processing that can be specifically realized by these hardware resources.

[0039] In this embodiment, "information" is represented by, for example, the physical value of a signal representing voltage or current, the high or low value of a signal as a collection of binary bits consisting of 0 or 1, or quantum superposition (so-called quantum bits), and communication and calculations can be performed on a circuit in a broad sense.

[0040] A circuit in the broad sense is a circuit realized by appropriately combining a circuit, circuitry, processor, memory, etc. That is, it includes a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), an LSI (Large Scale Integration), an ASIC (Application Specific Integrated Circuit), an FPGA (Field-Programmable Gate Array), etc.

[0041] <System configuration> Fig. 1 is a block diagram showing a system configuration according to one embodiment of the present invention. As shown in Fig. 1, a medical support system 1 includes an information processing device 10 and a database DB. The medical support system 1 is configured to be able to communicate with a plurality of user terminals 2 (reference numerals 2(a) to 2(c) in Fig. 1) and a generating device 3 via a network NW.

[0042] The information processing device 10 operates as a server that provides the medical support system 1, and the user terminal 2 is a terminal that is operated by a user who uses the medical support system 1. For example, the user may be a patient, a medical professional such as a doctor or a nurse, a person in charge of a medical device manufacturer, a person in charge of a testing institution, or the like.

[0043] The generating device 3 is an external server device that provides a generative model. The information processing device 10 may store the generative model and generate answer support without using an external device. The medical support system 1 may also configure a generating unit (described later) without using a generative model.

[0044] In this embodiment, the network NW is an IP (Internet Protocol) network, but there is no limitation on the type of communication protocol, and there is also no limitation on the type and scale of the network.

[0045] It should be noted that a general-purpose server computer or a personal computer can be used for the information processing device 10 and the generating device 3. It is also possible to configure the medical support system 1 by implementing the functional components described below on multiple computers.

[0046] A smartphone, a tablet terminal, a personal computer, a wearable device, or the like can be used as the user terminal 2. The user terminal 2 stores a medical support program for the user. In this embodiment, information related to medical support, etc. is displayed on the user terminal 2.

[0047] <Hardware configuration> 2(a) is a diagram showing an example of the hardware configuration of the information processing device 10. The information processing device 10 includes a control unit 11, a storage unit 12, and a communication unit 13 as the hardware configuration.

[0048] The control unit 11 includes one or more processors such as a CPU, and controls the overall operation and processing of the information processing device 10 by executing the medical support program according to the present invention, an OS, and other applications.

[0049] The storage unit 12 is an HDD, SSD, ROM, RAM, etc., and stores the medical support program according to the present invention and data used when the control unit 11 executes processing based on the program. The control unit 11 executes processing based on the medical support program stored in the storage unit 12, thereby realizing the functional configuration described below.

[0050] The communication unit 13 controls communication with the network NW, and performs inputs required to operate the information processing device 10 and outputs related to the operation results.

[0051] 2(b) is a diagram showing an example of the hardware configuration of a terminal 90 (user terminal 2 in FIG. 1). The terminal 90 includes, as its hardware configuration, a control unit 91, a storage unit 92, a communication unit 93, an input unit 94, and an output unit 95.

[0052] The control unit 91 of the terminal 90 includes one or more processors such as a CPU, and controls the overall operation and processing of the terminal 90. The storage unit 92 of the terminal 90 is an HDD, SSD, ROM, RAM, or the like, and stores the above-mentioned medical support program for the user, as well as data and the like used when the control unit 91 executes processing based on the program.

[0053] A communication unit 93 of the terminal 90 controls communication with the network. An input unit 94 of the terminal 90 is a touch panel, a mouse, a keyboard, or the like, and inputs operation requests from the user to the control unit 91. An output unit 95 of the terminal 90 is a display, or the like, and displays the results of processing by the control unit 91, etc.

[0054] Furthermore, the terminal 90 can execute a medical support program for the user to use the medical support system 1. This allows various information, such as information necessary for medical support, that has been processed for display via the display processing unit 105 to be displayed on the output unit 95, and also allows information to be input via the input unit 94.

[0055] The terminal 90 may be configured to be able to access the medical support system 1 provided as a web application by executing a web application instead of a medical support program for the user.

[0056] <Functional configuration> The information processing device 10 executes a medical support program to provide the medical support system 1. As shown in Fig. 2(a), the information processing device 10 includes, as functional components, a receiving unit 101, an acquiring unit 102, a calculating unit 103, a generating unit 104, and a display processing unit 105. These are information processing performed by software (stored in the storage unit 12) specifically realized by hardware (such as the control unit 11).

[0057] The receiving unit 101 receives various information from users such as patients and medical professionals. In this embodiment, user information and the like is registered based on input from users such as patients and medical professionals.

[0058] In this embodiment, users include patients who wish to receive a diagnosis from a doctor, people seeking advice on some medical condition, medical professionals such as doctors and nurses, personnel at medical device manufacturers, personnel at testing institutions, etc. In this specification, organizations such as medical device manufacturers and testing institutions may also be treated as users. Furthermore, medical professionals and personnel at medical device manufacturers may also be patients, and users may also include people who are not patients but have experienced some kind of disease in the past (people with a medical history).

[0059] The acquisition unit 102 acquires information necessary for medical support. In this embodiment, the acquisition unit 102 acquires questions about diseases posted by multiple users in a time series format in which the questions are displayed in chronological order in the order of posting, and disease response information corresponding to the questions. Note that the time series format in this embodiment is, for example, a timeline format or a thread format. In this specification, the timeline format is described as a list in which content posted by users is arranged in chronological order, but in a broad sense, it also includes records of events over time. Furthermore, the thread format displays messages and their replies in groups, but in a broad sense, it can also be interpreted as including records of events over time. Note that the time series format only needs to arrange the posted content in chronological order, and may be displayed in chronological order, such as oldest first or newest first, or in an order determined based on predetermined conditions.

[0060] The calculation unit 103 calculates a score required for medical support. In this embodiment, the calculation unit 103 calculates a score for each user using attribute information associated with the user, and stores the score in the storage unit 12 in association with the user.

[0061] The generation unit 104 generates various information useful for medical support. In this embodiment, the generation unit 104 generates answer support for guiding answers (answer candidates) to users' questions, based on questions about diseases posted in a timeline format (chronological format) by multiple users and disease response information corresponding to the questions.

[0062] Furthermore, the generation unit 104 inputs the question and disease correspondence information from the user into the generation model, and generates answer support based on the output of the generation model. The generation unit 104 inputs the question and disease correspondence information into the generation model, and generates answer support based on character information (text) extracted by the generation model.

[0063] Furthermore, the generation unit 104 can generate answer support based on a question from a user, disease correspondence information, and attribute information linked to each user. The generation unit 104 can input the question, disease correspondence information, and the score calculated by the calculation unit 103 using the attribute information into a generation model, and can also select additional users who can post answer candidates to the question.

[0064] In addition, the generation unit 104 may select candidate tests to derive candidate answers to the question based on the question, disease correspondence information, and test information, or the generation unit 104 may generate additional questions to derive candidate answers to the question based on the question and disease correspondence information.

[0065] The display processing unit 105 processes the display of various information. In this embodiment, the display processing unit 105 processes the display of answer support in a chronological format (such as a timeline format) in which the answers are displayed in chronological order in the order of posting, and transmits the display processing results. The display processing unit 105 can display the answer support including the additional users in a chronological format and transmit the display processing results.

[0066] Furthermore, the display processing unit 105 can display answer support including test candidates in a chronological order in which they are posted, and transmit the display processing results.The display processing unit 105 can also display answer support including follow-up questions in a chronological order, and transmit the display processing results.

[0067] <Database> 1 stores user information about users, organization information about organizations to which users belong, disease information, posted information including content posted in a timeline format, and other information necessary for medical support, etc. Some or all of this information may be stored in the storage unit 12 or the like, or some of this information may be stored in another database or the like.

[0068] The medical support system 1 and the processing performed by each functional component will be described below with reference to FIGS.

[0069] <Overview of the medical support system> The medical support system 1 according to this embodiment is a system for effectively providing answers to questions of a user without placing an excessive burden on the user. For example, when a user suspects that he or she has some kind of disease due to poor health, the medical support system 1 is used to obtain an appropriate answer (diagnosis) by joining a specific community (such as a timeline group) and casually consulting (posting or sending) with community members.

[0070] This medical support system 1 is highly effective when a patient is having trouble finding the cause of an illness even after visiting a hospital or testing facility. For example, this system can be highly effective when a patient is diagnosed with a "cold" after a medical examination at a hospital, but in fact it is a rare illness that even doctors have difficulty diagnosing.

[0071] A variety of users participate in the medical support system 1, including patients, clients, their families, doctors, nurses, public health nurses, physical therapists, clinical engineers, care workers, and other medical professionals, as well as personnel from medical device manufacturers and personnel from testing companies and laboratories that perform various tests to identify diseases. These users pose questions in a chronological format (such as a timeline), and these questions serve as a starting point for a dialogue in which each user shares their views based on their expertise and their own medical history. The following explanation takes as an example a dialogue posted in timeline format.

[0072] The medical support system 1 uses the dialogue between these multiple users to generate answer support to guide users to appropriate answers to their questions, and by providing answer support to the dialogue between multiple users (by providing answer support in the form of a conversational bot to the dialogue between multiple users in a timeline format), it is possible to obtain appropriate and highly accurate answer results (answer candidates) to the questions.

[0073] It should be noted that this medical support system 1 merely guides answer candidates, and for example, the doctor ultimately determines the diagnosis result for a patient from among the answer candidates. In other words, the answer candidates are used as a reference by the doctor when making a diagnosis, and the doctor is not necessarily required to select one result from the answer candidates. On the other hand, if a diagnosis result from a doctor is not required, for example, in the case of a client asking for advice on dieting, the client can simply determine an appropriate answer from among the answer candidates related to dieting. Furthermore, this medical support system 1 can also use the answer candidates as a diagnosis result (or candidate diagnosis results).

[0074] <Registering various information> Fig. 3 is a flowchart showing the processing procedure of the medical support system 1 according to one embodiment. In S201 of Fig. 3, the receiving unit 101 of the information processing device 10 receives various types of information such as user information, organization information, and disease information, and registers the information in the storage unit 12. The receiving unit 101 registers user information based on input from each user. The organization information and disease information can also be registered by an administrator of the medical support system 1, for example.

[0075] As shown in Figure 4(a), user information is managed by user ID and includes information such as the user's name, email address, password for using this system, age, gender, medical history, attribute information related to the user's attributes such as doctor and testing institution, disease ID and organization ID (described later), etc. In addition to this, user information can also include various other information related to the user, such as the user's contact information, address, and family information.

[0076] Here, the medical support system 1 in this embodiment can be used even by users who have not officially registered as members (or the type of member registration can be classified for each user). In this case, when registering user information, the user can omit information that identifies the user personally (for example, name, contact information, etc.). The receiving unit 101 can accept input of confidential information excluding information that identifies the user, and store the user information in the storage unit 12 based on this confidential information. By using such confidential information, the user can obtain the results of disease, diagnosis, etc. without being identified as an individual.

[0077] For some users, the content they post may be about a very sensitive issue, such as a medical condition, and many people feel uncomfortable or anxious about being identified. Therefore, in this embodiment, the service can be used without formal membership registration.

[0078] For example, it is possible to classify the types of member registration into multiple categories, and use this system (or register as a user of this system) by simply entering confidential information according to the classification. In this case, the functions available in this system can be differentiated between users who register using confidential information and users who register by entering all necessary information. It is also possible to classify the types of member registration into free membership and paid membership, with free members only being able to view the information. In this case, free members can register by simply entering confidential information.

[0079] Furthermore, in this embodiment, a user may register as a member by setting up a single account for himself or herself, or a single shared account may be set up for a user such as a patient or a family member of a client, and the family may register as a member. Sharing medical information and test information among family members can contribute significantly to identifying the cause of infectious diseases and rare diseases. In particular, the results of genetic testing can be crucial in identifying the cause of rare diseases, and genetic information shared among family members can be important information in terms of molecular biology and genetics. In this embodiment, user registration can be performed taking such circumstances into consideration. For example, the user and their family members (such as a mother, father, and child, a husband and wife, or a parent and child who require care) can share information within the same account.

[0080] 4(b), the organization information includes information such as the name of the organization such as XX Institution or XX Hospital, contact information for the organization, address of the organization, test information on tests that the organization can handle, disease IDs on diseases that correspond to the tests that the organization can handle, etc. In this embodiment, the user ID and organization ID are linked so that information on the organization to which the user belongs can be known.

[0081] In addition, as shown in Figure 4(c), disease information is managed by disease ID, including information such as the disease name (e.g., skin disease or liver function disease), symptoms that may occur with each disease, and candidate tests for identifying the disease.

[0082] In this embodiment, disease information is managed by disease names such as skin diseases and liver function diseases, but it can also be managed by larger concepts (higher concepts, general diseases), such as "infectious diseases" which include skin diseases, or "surgical diseases" which include liver function diseases, or it can be managed by narrowing down skin diseases and liver function diseases to more specific disease names and managing them as smaller concepts (lower concepts, detailed diseases). <User Submissions>

[0083] In S202 of Fig. 3, a user posts various information in a timeline format. Specifically, the user posts various information to a community of the medical support system 1. The community here refers to a chat group (also called a conversation group or community group) to which multiple users belong, created by the user (or created in advance as a specific group). A user can join this community group and have conversations with multiple users.

[0084] FIG. 5 shows an example of a display screen on a user terminal according to this embodiment. As shown in the figure, the display screen W10 displays conversations between users A, B, and C. In this embodiment, as shown in FIG. 5, user A first posts (or transmits) a question (question, request for diagnosis) saying, "I've had trouble with skin sores for the past two months. I'd like to know how to treat it." In FIG. 5, user A also posts an image of the skin he or she wishes to have examined or checked along with his or her question.

[0085] User B then posts some kind of answer or disease-related information (disease response information) to User A's question, saying, "Your skin is turning red, so it might be disease X. Have you been touching it?" User C then posts another disease-related information, saying, "Based on the number of days that have passed and the state of moisture, it might be disease X."

[0086] In this embodiment, a question from user A is the starting point, and multiple users such as user B and user C post disease response information, leading to a dialogue. In this specification, the starting post from user A is called a question, and the posts from user B and user C in response to this question are called disease response information. The content of the posts can be roughly classified into questions and disease response information. Also, in this embodiment, in FIG. 5, an image is attached in addition to text as the post from user A, and the text and image together can be regarded as a question from user A.

[0087] <Getting post content> Next, in S203, the acquisition unit 102 of the information processing device 10 acquires posted content from multiple users. Specifically, the acquisition unit 102 acquires, as posted content, the text information and skin image information of user A, the text information of user B, and the text information of user C in Fig. 5. The posted content of each user acquired by the acquisition unit 102 is stored in the storage unit 12 as posted information.

[0088] As shown in Fig. 4(d), posted information is managed by a posting ID, including information such as the user ID, the posting destination indicating each community group, the posted content, the posting date and time, etc. As shown in Fig. 5, the posted content is not limited to character information (text) and may include information such as images.

[0089] Also, in Figure 5, the acquisition unit 102 acquires text information and image information as the posted content, but the information to be acquired is not limited to these and may be various other information, such as URL information if a linked URL has been posted, or file information if an attached file has been included.

[0090] In this embodiment, the posted content acquired by the acquisition unit 102 is stored in the storage unit 12 as posted information. However, for example, it is also possible to execute generation of answer support, which will be described later, without storing the posted information in the storage unit 12.

[0091] <Generating Answer Support> In S204, the generation unit 104 generates answer support from the posted content. Specifically, the generation unit 104 generates answer support for guiding answer candidates to the question based on the question included in the posted content and the disease response information.

[0092] In this embodiment, the calculation unit 103 calculates a score used to generate answer support. As will be described in detail later, the calculation unit 103 calculates a score for a user using attribute information, associates the score with the user, and stores the score in the storage unit 12. The generation unit 104 can then generate answer support using the score.

[0093] 6 shows a schematic image of the information processing device 10 and the generation device 3 according to the embodiment. As shown in the figure, the generation unit 104 transmits the question and disease correspondence information acquired by the acquisition unit 102 to the generation device 3. The answer support generated by the generation device 3 is then displayed by the display processing unit 105, which will be described later, and the display processing result is transmitted to the user terminal 2.

[0094] The generation device 3 is configured to include a natural language processing (NLP (Natural Language Processing)) model. The generation device 3 is preferably configured to include a large-scale language model (LLM (Large Language Model)). The natural language processing model realizes computer processing of data input as natural language. The type of natural language processing model employed in this embodiment is not limited, but examples include ChatGPT (Chat Generative Pre-trained Transformer), BERT (Bidirectional Encoder Representations from Transformers), etc.

[0095] The generating device 3 can utilize a server device that provides a response generation model and / or an image generation model. The generating device 3 can also be configured using multiple devices. The information processing device 10 can also be configured to include the generating device 3.

[0096] The generation unit 104 of the information processing device 10 can generate a generation command instructing the generation of answer support as a sentence in natural language based on the question and disease correspondence information acquired by the acquisition unit 102, and transmit the command together with the question and disease correspondence information to the generation device 3. In this embodiment, answer candidates for the user's question can be obtained by using the answer support generated by this generation command. For example, if it is difficult to generate answer support (if a relevant answer cannot be found immediately), answer support may be generated by multiple interactions between the generation unit 104 and the generation device 3. Furthermore, in order to obtain more accurate answer support, predetermined preprocessing or the like can be performed in advance on various data stored in the storage unit 12.

[0097] Furthermore, the generation unit 104 inputs the question and disease correspondence information into a generation model and generates answer support based on the character information extracted by the generation model. For example, the generation unit 4 (generation device 3) can extract character information (text) included in the question and disease correspondence information by text mining.

[0098] In this embodiment, the generation unit 104 inputs the question and disease correspondence information into a generation model (generation device 3) and generates answer support based on the output of the generation model, but it is not necessary to use a generation model, and the generation unit 104 of the information processing device 10 can also generate answer support. For example, the information processing device 10 can store predetermined algorithms and tables for generating answer support in the storage unit 12, and the generation unit 104 can generate answer support using these algorithms and tables.

[0099] <Additional Questions> In this embodiment, various answer support can be generated by the generation unit 104. Here, follow-up questions will be explained. FIG. 7 shows an example of a display screen on which follow-up questions according to this embodiment are displayed. As shown in the figure, the generation unit 104 generates follow-up questions to derive answer candidates for a question based on the question and disease response information, and the display processing unit 105 displays the answer support including the follow-up questions in a timeline format and transmits the display processing results to the user terminal 2.

[0100] A follow-up question is a question that provides missing information to obtain answer candidates in response to a question from user A. In Fig. 7, in addition to the contents posted by users B and C, if the illness of user A can be identified by confirming with user A the situation, such as "Do you remember if you ate XX two months ago?" or "Has there been any change in the condition of your skin sores in the past two months," these questions are follow-up questions to assist in answering.

[0101] In this embodiment, additional questions also include asking questions to user B and user C in order to obtain further information (disease response information, answer candidates, etc.) from user B and user C. For example, in Fig. 7, additional questions such as "Is there any reason other than the redness of the skin that led you to conclude that it is disease XX?" and "Based on the number of days that have passed and the state of wetness, it seems possible that it is not disease XX but disease XX. What do you think?" can be asked to user B.

[0102] 8 shows a schematic image of the algorithm used to generate follow-up questions in this embodiment. In this embodiment, the algorithm and table for generating follow-up questions can be stored in the storage unit 12. In this figure, multiple diagnoses X, Y, and Z are shown, along with questions (questions A to E) and tests (tests P and Q) for obtaining these diagnoses.

[0103] For example, if we focus on diagnosis X, we can obtain diagnosis X by asking the user questions A and B, obtaining answers from the user, and then conducting test P and confirming the test results. In this embodiment, by providing answer support to fill in the missing questions, etc. required to obtain a diagnosis in a timeline-style dialogue, highly accurate answer candidates can be provided without placing a burden on the user.

[0104] 8, depending on the answer from user A to question C, a branch may occur, such as question D or question E. In this embodiment, the generation unit 104 can generate these questions as additional questions (answer support) when they are missing from the timeline-style dialogue or when branched questions are needed.

[0105] In this embodiment, a generative model is used that has been trained in advance on this algorithm and its diagnostic results. As described above, answer assistance can also be generated using the algorithm of Fig. 8 that has been stored in advance in the storage unit 12, without using a generative model.

[0106] <Additional User> Next, additional users as answer support will be described. Fig. 9 shows an example of a display screen displaying additional users according to this embodiment. The generation unit 104 inputs the question, disease response information, and attribute information linked to each user into the generation device 3 (generation model), selects additional users who can post answer candidates to the question of user A, and generates answer support including the selected additional users.

[0107] These additional users may be, for example, users with a medical history similar to (or similar to) the illness of user A, doctors familiar with the illness, personnel from testing institutions, personnel from manufacturers, etc. In other words, the additional users are users who can derive answer results by interacting with multiple users in the community.

[0108] In this embodiment, after user A confirms the information of the additional user (candidate user to be added), information regarding approval to join the community is sent to the additional user selected from the candidate users to be added, and by receiving approval to join from the additional user, interaction with the additional user becomes possible. In this medical support system 1, this approval to join is not a required process, and for example, user A may select an additional user, who may then automatically join the community.

[0109] Alternatively, participation approval can be sent to all users who are candidates for additional users, and users who can participate in the community can be selected from among the multiple candidates for additional users who have received approval. In this case, multiple users can be selected to participate in the community.

[0110] As described above, in this embodiment, the calculation unit 103 calculates a predetermined score, and this score can be used to select additional users. The calculation unit 103 calculates a user's score using attribute information, associates the score with the user, and stores it in the storage unit 12 (not shown). The generation unit 104 then inputs the question, disease correspondence information, and score into a generative model, and can select additional users who can post answer candidates to the question based on the output from the generative model (generation device 3).

[0111] In this embodiment, the scores can include an occupation score according to the occupation, such as a doctor, an employer score related to the doctor's workplace, a medical history score related to the patient's medical history, and an evaluation score related to the evaluation of the user's posts. Each of these pieces of information related to a user can be scored, and these multiple scores can be calculated based on a predetermined logical formula. For example, the score for a certain user can be calculated using the logical formula "X (occupation score + employer score) + Y (medical history score) + Z (evaluation score) = score, (X, Y, and Z are predetermined coefficients)."

[0112] For example, a score can be calculated based on the disease ID of the user posting the question and the disease IDs of other users (such as a higher score if the diseases or symptoms are similar). Also, if the scores are relative, additional user candidates can be identified in descending order of the scores for the user posting the question, or users posting questions can be scored and users with scores close to the scores can be identified as additional user candidates.

[0113] In this way, in this embodiment, by obtaining the information missing to obtain a diagnosis in a timeline-style dialogue from posts (disease response information) from additional users, it is possible to provide highly accurate answers (answer candidates) without placing a burden on the user.

[0114] <Test candidate> Next, we will explain test candidates as answer support. As shown in FIG. 10, the generation unit 104 selects test candidates to derive answer candidates to a question based on the question, disease response information, and test information related to diseases linked to the organization to which the user belongs. In FIG. 10, posts from users A, B, and C are used to provide test candidates such as an XX institution for AAA disease testing and an XX company for BBB confirmation testing. For example, various tests are expected, such as health checkups, genetic tests, and allergy tests.

[0115] In this embodiment, it is preferable to have a plurality of candidate tests so that the user can select one, and more preferably two or three candidate tests are selected. Alternatively, the number of candidate tests may be one.

[0116] In this embodiment, detailed information about a test can be confirmed by selecting (touching) one of the candidate tests. After confirming the detailed information, User A (e.g., a patient) can select the test that he or she will undergo.

[0117] For example, if a patient wishes, testing can be performed while ensuring the patient's anonymity. In this case, even if user information is registered in the medical support system 1, the patient can undergo testing without providing their personal information to the testing institution (providing only the information necessary for the test). The patient can then check the test results from the medical support system 1.

[0118] Candidate tests can also be identified using the algorithm in Figure 8. That is, as shown in Figure 8, if it is known that Question A and Question B have been asked in a community post, the possibility of obtaining Diagnosis X can be increased by suggesting Test P to the patient as answer support (depending on the test results) as the next step.

[0119] While Figure 8 shows that test P is necessary to obtain diagnosis X, it is possible that diagnosis X or other diagnostic results may be obtained by performing other tests according to, for example, multiple other algorithms. In this embodiment, in addition to the algorithm shown in the schematic image of Figure 8, multiple other algorithms can also be used to select more optimal test candidates.

[0120] Furthermore, in this embodiment, candidate tests can be identified using the user's location information and the location information of testing institutions, etc. For example, a regional map including the locations of candidate tests (such as the locations of hospitals that can perform tests necessary to determine the user's illness, etc.) is displayed on the user terminal 2, and candidate tests (testing institutions) can be identified by the user's selection input (such as by touching the screen). Furthermore, the generation unit 104 can identify candidate tests as answer support by appropriately combining the above-mentioned multiple processes and location information.

[0121] <Ensuring evidence for response support> In this embodiment, to ensure that medical answers are factual, the medical support system 1 preferentially (selectively) searches public databases such as peer-reviewed papers, medical manuals, and various guidelines based on predetermined conditions to obtain answer support. For example, in this embodiment, a search using Retrieval Augmented Generation (RAG) can be used. Note that the medical support system 1 can also use non-public information such as internal information from a predetermined organization.

[0122] The medical support system 1 can output (display) answer support while indicating citations and sources along with candidate answers to the user's questions, thereby addressing issues such as a lack of scientific evidence in the explanation, outdated information, the occurrence of hallucination, etc. As a result, it is possible to provide the user with more appropriate answer support that indicates the evidence for the answer explanation.

[0123] The generative model according to this embodiment can utilize a multimodal model that can handle not only text data but also various types of data such as images and audio. In addition, by indexing (keywording) various types of data stored in a database, it is possible to further perform multivariate analysis on keywords extracted by text mining or the like, thereby narrowing down and utilizing the keywords.

[0124] <Selection of various modes> The generation unit 104 according to this embodiment can be configured to be able to switch between multiple modes, such as a "normal medical consultation mode," a "rare disease response mode for rare diseases," and an "infectious disease epidemiology mode for infectious diseases," depending on the generation process and situation of the answer support by the generation device 3 (generative model). The generation unit 104 can also generate answer support after one of the multiple modes is selected by a selection operation from the user.

[0125] The generation unit 104 can switch between support modes such as a normal medical consultation mode, a rare disease response mode, and an infectious disease epidemiology mode, either automatically during the response support generation process or manually by the user. The knowledge base to be used (stored in the storage unit 12), search targets, display items, etc. can be changed depending on the selected mode. For example, in this embodiment, if it is determined that no highly relevant information is found in the medical consultation mode, the mode is automatically switched to the rare disease response mode.

[0126] Specifically, in the normal medical consultation mode, a patient as a user posts a question on a timeline or the like, and the generation unit 104 uses the generation device 3 (generative model) to search knowledge (databases, websites, etc.) based on the question and extract highly relevant information. The generation unit 104 can then generate answer support by combining the information obtained through the search (answer candidates, etc.) with a trained generative model such as LLM. In this embodiment, the normal medical consultation mode may be set as the initial setting.

[0127] In the rare disease response mode for a rare disease, a patient as a user posts a question (inquiry) on a timeline or the like, and the generation unit 104 causes the generation device 3 (generative model) to search knowledge (knowledge base, database, website, etc.) based on the question, and as a result, no highly relevant information is extracted. For example, the system can be configured to switch from the normal rare disease response mode to the rare disease response mode for a rare disease at this time, or the system can be set to the rare disease response mode for a rare disease from the beginning. At this time, the generation unit 104 generates an additional question for the user and posts it on the timeline.

[0128] In this mode, the search prioritizes more specialized knowledge (medical encyclopedias, research paper search sites, etc.) based on multiple follow-up questions and the user's responses to them. Keywords are extracted (text mining, indexing) from the obtained information using multivariate analysis, etc., and the obtained keywords (for example, the name of a potential disease) are used to search for more specialized knowledge (case report papers, checklists, clinical guidelines from academic societies, etc.) that contain information on the likelihood of the disease. The information obtained from the search and information on the citation source (evidence) can then be provided to the user.

[0129] In the infectious disease epidemiology mode for infectious diseases, patients as users post questions on a timeline or the like, and the generation unit 104, based on the post, preferentially searches for knowledge about infectious diseases (infectious disease information, public health center observation data, past patient questions, etc.) and guides (posts) the user to hospitals or medical institutions that can perform tests for the disease in question. The user then undergoes testing at a hospital or medical institution. After that, the user can post their subjective symptoms along with the test results on the timeline to receive more accurate answer support.

[0130] In this way, by using a plurality of modes according to the user's illness, etc., it is possible to generate answer support with higher accuracy.

[0131] In this embodiment, highly specialized information is extracted in advance and stored in the storage unit 12, and after being converted into text data as preprocessing for using the generative model, this data can be used as knowledge.

[0132] In addition, in this embodiment, map information (location information) can be associated with the outbreak status of infectious diseases, and the outbreak status of infectious diseases can be displayed on a map. For example, in this embodiment, infection information from patients who have previously used this medical support system 1, information on the infection status from external institutions, etc. can be displayed on the map in real time (or at predetermined intervals) by linking with GPS or the like. In this case, the storage unit 12 may have the predetermined map information, or predetermined map information on an external server can be used by API linkage or the like. Such map information showing the outbreak status of infectious diseases can be effectively used, particularly in the infectious disease epidemiology mode described above.

[0133] As a specific example, in the infectious disease epidemiology mode of this medical support system 1, if there is a high incidence of a specific infectious disease in a certain area, the user can post on their timeline that they suspect the specific infectious disease by referring to a map (allowing the information to be shared with other users), or the generation device 3 can search for specific events that have occurred in the area (such as a water pipe burst or river water pollution), and the search results can be used to provide the user with answer support such as the cause of the infectious disease that has occurred and how to deal with it.

[0134] <Display processing> 3, the display processing unit 105 performs display processing of answer support. As described above, the display processing unit 105 displays the answer support generated by the generation unit 104 in a timeline format (chronological format) and transmits the display processing result to the user terminal 2. For example, in FIG. 5, multiple users, User A, User B, and User C, are having a conversation in a specific community group, and the display processing unit 105 transmits answer support in a timeline format so as to join this conversation.

[0135] Specifically, the display processing unit 105 in this embodiment displays answer support including additional questions in a timeline format, as shown in Fig. 7, and transmits the display processing results. The display processing unit 105 can also display answer support including additional users in a timeline format, as shown in Fig. 9, and transmit the display processing results. The display processing unit 105 can also display answer support including test candidates in a timeline format, as shown in Fig. 10, and transmit the display processing results.

[0136] The display processing unit 105 can also display all or any of the follow-up questions, additional users, and test candidates in a timeline format as answer support, and transmit the display processing results. As shown in FIG. 11 , the display processing unit 105 displays the follow-up questions in a timeline format and transmits the display processing results to the user terminal 2 of user A. Thereafter, the patient's answers to the follow-up questions are displayed in a timeline format. Then, based on the answer results to the follow-up questions from user A, the display processing unit 105 further displays the additional users as answer support, and transmits the display processing results to the user terminal 2.

[0137] In FIG. 11, user D is added to the community as an additional user, and user D posts disease response information. User A answers user D's question. Then, based on the answer to user D's question, the display processing unit 105 can further display candidate tests as answer support, and transmit the display processing results to the user terminal 2.

[0138] In this manner, in this embodiment, multiple types of answer support can be displayed as appropriate. The generation unit 104 in this embodiment generates additional answer support based on posts (answer results) from users in response to answer support, and the display processing unit 105 displays the additional answer support in a timeline format and transmits the display processing results.

[0139] The display screen W10 in FIG. 11 is an example, and the order and number of additional questions, additional users, and test candidates as answer support will vary depending on the posted content, and optimal answer support will be generated as appropriate.

[0140] The user can obtain an answer result (such as a disease) from the answer support provided by this series of steps. For example, if the answer result to the user's question is a specific disease, the display processing unit 105 can display a map on which the number of patients with that disease is plotted, and transmit the display processing result to the user terminal 2. The display processing unit 105 can display a map of a specific area using the location information of the user terminal 2. By performing such display processing, it is possible to confirm how many patients exist in which area (in areas further subdivided into the specific area) in a specific area related to the user.

[0141] In this embodiment, information regarding the number of patients linked to location information (for example, information about users whose diseases were identified as response results) is stored in advance in the memory unit 12 of the medical support system 1, and the above map can be displayed using this information regarding the number of patients.

[0142] As described above, the medical support system 1 according to the present invention generates various types of answer support by executing predetermined processing using the posted content, such as disease response information posted in chronological order, thereby realizing appropriate and highly accurate answers to the patient's questions without placing a burden on the patient.

[0143] Furthermore, in this embodiment, the medical support system has been mainly described, but it goes without saying that the same effects as the present invention can also be obtained from a medical support method or a medical support program having the same characteristics as this system.

[0144] In this embodiment, the medical field has been mainly described, but the same effect as the present invention can also be obtained in various other cases of support, such as sports, learning, education, and tourist guidance. [Explanation of symbols]

[0145] 1 Medical support system 2. User terminal 3 Generator 10. Information processing equipment 11 Control section 12 Storage section 13 Communications Department 90 terminals (user terminal 2) 91 Control Unit 92 Memory section 93 Communications Department 94 Input section 95 Output section 101 Receiving unit 102 Acquisition Department 103 Calculation Unit 104 Generation part 105 Display processing unit NW Network W10 display screen

Claims

1. A medical support system for providing medical support, the medical support system includes an acquisition unit, a generation unit, and a display processing unit; the acquiring unit acquires questions about diseases posted by a plurality of users and disease response information corresponding to the questions; the generation unit generates answer support for guiding answer candidates to the question based on the question and the disease correspondence information; the display processing unit performs a display process of the answer assistance and transmits a display process result. Medical support system.

2. The acquisition unit acquires the questions and the disease response information posted from a plurality of users in a chronological order in which the questions are displayed in chronological order, and the display processing unit processes the response assistance to be displayed in the time series format and transmits the display processing result. The medical support system according to claim 1 .

3. the generation unit inputs the question and the disease correspondence information into a generation model, and generates answer assistance based on an output of the generation model. The medical support system according to claim 2 .

4. The question and the disease response information include text information, the generation unit inputs the question and the disease correspondence information into a generation model, and generates the answer assistance based on the character information extracted by the generation model. The medical support system according to claim 3 .

5. the medical support system includes a storage unit; the storage unit stores attribute information of the user; the generation unit generates the answer assistance based on the question, the disease response information, and attribute information associated with each user.

3. The medical support system according to claim 1.

6. the generation unit inputs the question, the disease response information, and attribute information associated with each user into the generation model, and selects additional users who can post answer candidates to the question; generating the answer assistant including the selected additional users; the display processing unit displays the reply support including the additional user in a chronological order in which the reply support is displayed in chronological order of posting, and transmits the display processing result. The medical support system according to claim 5 .

7. The medical support system further includes a calculation unit, the calculation unit calculates a score of the user using the attribute information, associates the score with the user, and stores the score in the storage unit; the generation unit inputs the question, the disease correspondence information, and the score into the generation model, and selects additional users who can post answer candidates to the question. The medical support system according to claim 6.

8. The medical support system further includes a storage unit, the storage unit stores test information related to a disease associated with an organization to which the user belongs; the generation unit selects candidate tests for deriving candidate answers to the question based on the question, the disease association information, and the test information; the display processing unit displays the answer support including the test candidates in a chronological order in which the answers are posted, and transmits the display processing result.

3. The medical support system according to claim 1.

9. The medical support system further includes a storage unit, The storage unit stores examination information and map information related to diseases associated with an organization to which the user belongs; the generation unit selects candidate tests for deriving candidate answers to the question based on the question, the disease association information, and the test information; the display processing unit processes the candidate examinations together with the user's location information on a map based on the map information, and transmits a display processing result.

3. The medical support system according to claim 1.

10. the generation unit generates a follow-up question for deriving answer candidates to the question based on the question and the disease correspondence information; the display processing unit displays the answer support including the follow-up question in a chronological order in which the answers are posted, and transmits the display processing result.

3. The medical support system according to claim 1.

11. the generation unit has a plurality of support modes; The normal medical consultation mode, the rare disease response mode, and the infectious disease epidemiology mode can be switched by automatic determination in the generation process of the answer support or manual operation by the user, Change the search target according to a predetermined support mode; 3. The medical support system according to claim 1.

12. A medical support method executed by a medical support system for providing medical support, comprising: the medical support system includes an acquisition unit, a generation unit, and a display processing unit; The acquisition unit acquires questions about diseases posted by a plurality of users and disease response information corresponding to the questions; generating answer support for guiding answer candidates to the question based on the question and the disease correspondence information by the generating unit; the display processing unit performs a display process on the answer assistance and transmits a display process result. Medical support methods.

13. A medical assistance program for providing medical-related assistance, causing a computer to function as an acquisition unit, a generation unit, and a display processing unit; The acquisition unit acquires questions about diseases posted by a plurality of users and disease response information corresponding to the questions, the generation unit generates answer support for guiding answer candidates to the question based on the question and the disease correspondence information; the display processing unit performs a display process of the answer assistance and transmits a display process result. Medical assistance program.

Images